Combined Physicochemical and Energy Methods to Improve the Recovery of Rare Earth Elements from Eudialyte Concentrate

Abstract

The parameters for efficient nitric acid leaching were experimentally determined, which ensured the recoveries of Zr and REEs from eudialyte concentrate up to 87.0%–91.7% and 76.0%–81.1%, respectively. The possibility was shown of intensifying the leaching process through preliminary energy treatments to ensure the intensive breakdown of mineral complexes and grains; as a result, the recovery of Zr and REEs increased by more than 10%. A process was developed for the selective recovery of up to 91.5% of zirconium and up to 71.2% of REEs in the form of carbonate compounds from the pregnant solution of nitric acid leaching by chemical precipitation as well as up to 81.1% REEs and up to 91.7% zirconium on hypercrosslinked polystyrene sorbents.

1. Introduction

REEs are indispensable elements in the modern automotive, aerospace, metallurgical, chemical, electronic, energy and nuclear industries. It is impossible to produce high-tech communications, modern magnets, special steels, batteries, and computer chips without use of the REEs. In some sources, rare earth elements are called “metals of a new technological order [1]”.

However, the reserves of rich mineral resources (bastnasite, monazite, and xenotime) are limited, which is why there is a need for the development of new technology for processing the secondary REE resources. These resources include mineral feeds with a low content of REEs (apatites, eudialyte ores, and ocean bottom sediments) and waste (electronic scrap, phosphors, lighting bulbs, coal fly ash, etc.) [2,3,4,5,6,7,8,9,10,11].

A promising source of REE in Russia is the eudialyte ores (Lovozero deposit, Kola Peninsula). The resources of eudialyte ores of the Lovozero massif are huge (ZrO₂: 300 million ton;, REE oxides: 60–70 million tons). The content of the valuable components in eudialyte ores is quite low (0.5 g/t), but the content of the most valuable heavy REE (HREE) is sufficiently high (20%). Eudialyte is a potential source of Hf, and it usually contains very little thorium and uranium. This is why the development of technology for the recovery and separation of REE from the eudialyte is of current interest, taking into account the depletion of rich REE resources [12,13].

Various methods for the processing of eudialyte concentrates have been proposed in the literature [14,15,16,17]. The main problems of hydrometallurgical eudialyte processing is the formation silicic acid. It undergoes a polycondensation process with the formation of polymers of various compositions and structures up to stable gels. The latter leads to a drop in the recovery of valuable components. The various methods of silicic acid polymerization (silicate gel formation) are reported in the literature. Various solutions the silica gel formation problem have been proposed: silica gel formation can be avoided by applying a combination of mechanochemical and extraction with nitric acid/tri-n-butylphosphate [18], the introduction of F⁻ ions during leaching [19], and a two-stage hydrometallurgical process [17], including preliminary acid treatment of the heated concentrate followed by aqueous leaching. In this paper, the use of energy effects on mineral pulp (ultrasound and electrochemical processing) is suggested. The experimental results show an increase in leaching efficiency and the opening of the silicate matrix [20,21].

Another significant problem with eudialyte processing is the recovery and separation of the valuable components (Zr and REE) from the pregnant solution of the eudialyte concentrate through leaching. However, the pregnant solution of the eudialyte concentrate lichate is quite complex in terms of the high concentration of the leaching agents, the low concentration of the valuable components, and the high concentration of the accompanying elements, most significantly alkaline earth. Extractive and sorption technologies are used for the selective recovery of the REEs from the leaching solutions. Among the human-made selective sorption materials, sorbents based on hypercrosslinked polystyrene are important [22]. The chemical inertness of hypercrosslinked polystyrene allows its use in aggressive media such as concentrated solutions of HNO₃, HF, and H₂SO₄. The important advantage of this sorbent is the stability of the volume of the sorbent layer in the column during contact with electrolyte solutions of different concentrations; additionally, the method is simple and can be easily automated. It is also important that all used sorbents are produced on an industrial scale [23]. This paper presents the results of the application of sorbents based on hypercrosslinked polystyrene for the processing of a real pregnant solution of eudialyte acid leaching.

2. Materials and Methods

Eudialyte concentrate from the Lovozero field was used in leaching experiments (

Table 1. Chemical composition of eudialyte concentrate (wt %).

Component	Na2O	MgO	Al2O3	SiO2	K2O	CaO	TiO2	MnO	Fe2O3
Content	10.14	0.18	6.90	44.96	1.37	5.42	3.29	1.66	5.20
Component	Cl	SO3	HfO2	Nb2O5	Rb2O	SrO	Ba	U	Th
Content	0.97	0.20	0.14	0.81	0.01	2.63	0.09	<0.01	<0.02
Component	ZrO2	Ta2O5	As	CeO2	Y2O3	Nd2O3	La2O3	Other REE oxides	LOI
Content	11.50	0.05	0.03	0.68	0.46	0.32	0.29	0.75	1.97

Note: LOI, ignition loss.

).

Leaching conditions were varied depending on the aims of the experiments. These conditions are carefully detailed in [24,25] and are presented in

Table 2. Leaching conditions.

Autoclave Name	Sample Mass, G	Acid Concentration, M	Temperature Range, °C	Residence Time, min	S:L Ratio	Agitation Speed, min–1
TOP 120 series (TOP INDUSTRY)	60 g	2.4–9.6	80–140 °C	60–300	1:10	550
IPKON RAS Equipment *

* autoclave equipped with the fitting for US and EC treatment of mineral pulp during leaching.

.

Preliminary ultrasonic (US) treatment in water (5, 10, 20, and 30 min) and in a 7.2 M acid solution (10 min) was carried out with a dispersion machine (MEF-15, MELFIZ, manufacturer—MELFIZ, Moscow, Russia) at a power of 600 W and an operating frequency of 22 kHz at S:L = 1:10. Electrochemical (EC) treatment conditions were as follows: anode current density of 200 A/m² and volume current density of 1.5 A/L.

After the leaching, the insoluble mineral residue was separated from the pregnant solution with centrifugation (UC-1536 E, ULAB, China) then it was washed with distilled H₂O and dried. The chemical composition of the solid residues was analyzed by ICP-MS (ICPE-9000, Shimadzu, Kyoto, Japan) after digestion with HNO₃ solution and XRF (ARL ADVANT’X, Thermo Scientific, Waltham, MA, USA). The chemical composition of the liquid phases was analyzed using ISP-AES (Elan-6100, PerkinElmer, Waltham, MA, USA) and MS.

The surface transformation of the eudialyte grains was examined with XPS (Versa Probe II spectrometer, ULVAC-PHI, Chigasaki, Japan) and DRIFT (IR-Affinity, Shimadzu, Kyoto, Japan).

The micromorphology and chemical composition of the eudialyte grains were determined using an analytical scanning electron microscope (ASEM) LEO 1420VP (Carl Zeiss, Cambridge, UK), equipped with an energy-dispersive X-ray INCA 350 spectrometer (Oxford Instruments Ltd., Oxford, UK) at an accelerating voltage of 20 kV in high-vacuum mode. The samples were fixed on a carbon conductive tape and coated with a thin layer of carbon.

Chemical precipitation of Zr and REE was performed in two steps. Zirconium hydroxides were recovered during the 1st step, when the leach solution was alkalized with CaCO₃ to pH 3.8–4.0. The REEs were recovered during the second stage in the form of hydroxides and hydrocarbonates, with further solution alkalization with caustic soda (Na₂CO₃) to pH 6.0–6.4. Precipitates were separated from the solution. The final precipitates were washed with distilled H₂O and dried to constant weight.

Highly crosslinked polystyrene sorbents DOWEX 1–X–8 (Acros Organics), PCA-433 (−1.2 + 0.3 mm), and Puromet MTS 9500, Purolite C 160 (Purolite) sized −63 + 43 µm were used in the study. The specifications of the sorbents used are presented in

Table 3. Ion exchange resins used in the study.

Resin	Dowex 1 × 8 Strong Base Anion Exchange Resin Strongly Basic	Purolite PCA433 Strong Base Anion Exchange Resin	Purolite C160 Strong Acidic Exchange Resin	Puromet MTS 9500 Chelating Resin
Polymeric matrix	Styren divinylbenzene copolymer
Functional group	Trimethyl ammonium	Quaternary ammonium type I	Sulfonic/aminophosphonic	Aminophosphonic
Exchange capacity	1.2 eq/L	1.3 eq/L	2.4 eq/L	1.3 eq/L Ca
Other properties	Gel polymer Crosslinking 8% Operating pH 0–14	Gel polymer Operating pH 0–14	Macroporous Operating pH 0–14	Macroporous chelating Operating pH 0–14

.

The preparation of polystyrene sorbents for the experiment was as follows: the sorbent sample was kept in a 50% HNO₃ solution for 24 h, then washed with distilled H₂O until the neutral pH of the wash was reached.

In the sorption experiment, a pregnant acid leach solution was used (line 1 in

Table 4. Composition of the pregnant leach solution and refined REE solution.

	Na	Mg	Al	K	Ca	Ti	Si	Sc	Cr	Ni
		mg/L
1	2272.0	16.1	520.7	115.7	1121.0	233.7	0.1	0.1	0.5	1.6
2	187.65	5.7	288.4	43.9	263.0	89.6	0.1	0.1	0.4	1.5
	Zn	Cu	Sr	Mn	Fe	Zr	Ba	Hf	Nb	ΣREE
1	13.6	0.8	580.0	527.0	580.0	2657.0	36.9	29.8	0.7	596.0
2	10.2	0.7	34.0	87.7	398.0	n/d	n/d	29.8	0.6	571.6

Note: n/d—not detected, below detection limits.

and

Table 5. REE concentration.

	Sc	Y	La	Ce	Pr	Nd	Sm	Eu	Gd	Tb	Dy	Ho	Er	Tm	Yb	Lu
		mg/L
1	0.4	138.4	76.7	18.6	22.5	66.1	20.1	6.7	23.5	4.1	23.2	4.8	12.9	2.0	11.6	1.6
2	0.3	135.4	68.4	17.4	21.8	63.3	19.7	6.5	22.7	4.0	23.2	4.7	12.8	1.9	11.0	1.5

) as well as a refined REE solution obtained by dissolution (50% HNO₃) of REE precipitate isolated from the initial pregnant solution during its neutralization with calcium carbonate, according to the method described above (line 2 in Table 4 and Table 5).

The sorption experiment involved the following protocol: a 30 mL glass column was filled with water and sorbent granules, then a solution (1 mL/min) was fed through the column until the concentration of La at the column outlet was equal to the concentration at the column inlet. La was used as indicative element to control the column saturation with REEs as one of the representatives of the REE group.

After the column with the initial solution was equilibrated, the electrolyte was displaced by a top-to-bottom flow of the mobile phase. In both cases, the eluate was collected in 2.0 mL fractions, the content of metals in which was measured via ISP-MS, AES, and MS.

3. Results

3.1. Effect of Acids on the Leaching Efficiency of Eudialyte Concentrate

SPECIAL FEATURES OF EUDIALYTE SURFACE WHEN EXPOSED TO ACIDS. The microscopic investigations revealed the morphological features of eudialyte particles after chemical interaction with acid solutions. We found that a main feature of the eudialyte particles after interaction with H₂SO₄ and HCl solutions was the appearance of deep, vertical cracks (Figure 1b,c). It can be assumed that, in these acids, the mineral normal dissolution rate is comparable to and/or exceeds the tangential (layer-b-layer) rate, which contributes to a more active dissolution of the entire volume of the particle and not just its surface. A specific feature of the interaction of eudialyte with a HNO₃ solution is the layer-by-layer mechanism of mineral dissolution (Figure 1a).

Using the ASEM technique, it was determined that the degree of eudialyte grain destruction varied during the leaching process, while the grains of potassium feldspar, loparite, and aegirine were morphologically unchanged; the particles of lamprophyllite demonstrated the initial stage of dissolution.

Some secondary phases were found among the solid materials after the leaching of concentrate with H₂SO₄: the intergrowth of elongated gypsum and celestine crystals formed due to the interaction of Ca²⁺ and Sr²⁺ ions passing into the solution upon dissolution of lamprophyllite and the anionic group [SO₄]^2- (Figure 2).

Most mineral particles after interaction with the HNO₃ solution were considerably changed and covered with a dense layer of redeposited leaching products. The exception was individual grains of aegirine, which retained their initial needle-prismatic habitus. In addition, the formation of secondary phases was not excluded. Apparently, they were represented by a thin mixture of silicate gels with various degrees of hydration, with captured micron particles of acid-resistant minerals (egirine, loparite, feldspars), hydrolysis products of aluminum salts, as well as various ions of rare, alkaline, and alkaline earth elements sorbed from solution and complex ionic groups, presumably [Nb(H₂O)₆](OH)₂}³⁺, [Zr(NO₃)₆]₂, [Zr(OH)₂]²⁺, [NaSiO₃]⁻, etc.

The influence of various inorganic acids (hydrochloric, sulfuric, and nitric) on the eudialyte surface state was also studied by IR spectroscopy. The IR spectra of the sample both initially and after the interaction with HCl, H₂SO₄, and HNO₃ are given in Figure 3. The analysis of the obtained results led us to the following conclusion: the most considerable effect on the structural and chemical properties of the mineral was exerted by the HCl solution. Figure 3 shows that the spectra profile in the region from 800 cm⁻¹ to 1350 cm⁻¹ significantly changed compared with that of the initial (untreated) sample. For the sample after the interaction with HCl solution. A change in the band symmetry with an obvious broadening was observed: the shoulder at 1135 cm⁻¹ transformed into a distinct band at 1145 cm⁻¹; the peaks at 980 cm⁻¹ and 1025 cm⁻¹ disappeared.

A lowering of the mentioned band peak area was also observed. In the initial sample, the peak area value was about 585 rel. units, while for the sample treated with HCl, it was about 132 rel. units, i.e., a change by a factor of 4.4.

The action of the aqueous solutions of H₂SO₄ and HNO₃ on eudialyte modified its surface to a lesser extent. The spectra in the region of Si-O-Si bond-stretching vibrations maintained the symmetry; the area under the spectra in the discussed range dropped from 585 rel. units (initial sample) to 355 rel. units and 350 rel. units, respectively.

The XPS results generally agreed with the IR spectroscopy data. The results of the XPS study of the elemental composition (at.%) of eudialyte particles after the interaction with the solutions of HNO₃, H₂SO₄, and HCl are presented in

Table 6. Atomic concentration (at.%) of eudialyte surface according to XPS data.

ACID	Atomic Concentration, at.%																ΣMe/Si
	C	N	O	S	F	Na	Mg	Ca	Al	Si	K	Ti	Mn	Fe	Sr	Zr
No treatment	4.7	0.6	60.0	n.d.	0.6	12.2	0.5	0.5	2.8	14.8	0.4	0.9	0.5	0.7	0.4	1.2	1.36
NHO3	4.8	1.9	66.2	n.d.	0.1	3.4	0.2	0.2	1.4	19.2	0.2	0.6	0.2	0.5	0.2	1.2	0.42
H2SO4	2.5	0.3	69.2	2.2	<0.1	0.9	0.1	0.1	0.1	23.1	0.1	0.3	0.2	0.2	0.5	0.1	0.11
HCl	3.9	0.2	70.2	n.d.	0.1	0.2	<0.1	<0.1	<0.1	24.6	n.d.	0.1	<0.1	0.1	n.d..	0.4	0.05

Note: n.d.—not detected.

. The surface of the treated samples was depleted in metal content as a result of the intensive removal of cations from the surface into the leaching solution. The surface concentration of aluminum and sodium dropped by ~98%; calcium by 80%; magnesium, manganese, and potassium by 75%; titanium by 70%; and iron by 70%. The surface concentration of Si enhanced by 1.3–1.7 times.

Through this process, the ratio of Me to Si significantly decreased, as shown in Table 6. The largest decrease in the Me/Si ratio was observed for the sample after the interaction with HCl (from 1.36 to 0.05), which confirmed that HCl was the most efficient agent for the leaching.

LEACH CAKES. The highest yield of metals from the eudialyte concentrate was observed when HCl was used as a leaching agent (

Table 7. XRF results of the concentrate before the treatment and after leaching with various acids.

ACID	Chemical Composition, at.%
	SO3	Na2O	MgO	CaO	Al2O3	K2O	TiO2	MnO	Fe2O3	SrO	ZrO2	CeO2	La2O3	Nd2O3	Y2O3	ΣMeO	SiO2	Me/Si
Without treatment	0.20	10.1	0.18	5.42	6.9	1.37	3.29	1.66	5.2	2.6	11.5	0.67	0.29	0.32	0.46	50.16	44.96	1.14
NHO3	<0.02	1.6	0.08	1.60	1.6	0.99	2.39	0.45	3.0	1.6	5.9	0.35	0.18	0.14	0.12	20.02	64.61	0.30
H2SO4	2.39	2.0	0.07	2.54	1.7	1.06	2.96	0.51	3.2	2.7	4.1	0.60	0.30	0.22	0.17	24.52	63.62	0.35
HCl	0.02	0.8	0.04	1.11	1.1	0.72	2.05	0.27	2.4	1.0	3.7	0.27	0.12	0.10	0.08	13.78	70.57	0.20

). The ratio of ΣMe/Si in the cakes dropped by ~8 times (from ~1.7 to 0.2). When using the solutions of HNO₃ and H₂SO₄ as the solvent, the obtained ratio of ΣMe to Si in the cakes was about 0.3.

It should be noted that in the cakes, in comparison with the initial concentrate, there was a significant increase in the concentration of gel, for example, SiO₂ from 44.96% to 64.61% (HNO₃), 63.62% (H₂SO₄), and 70.57% (HCl), which apparently caused minimal gelation when using HCl as the leaching agent (Table 7).

GEL. The effect of the acids on the content (vol.%) of the gel-like silicate was investigated. As shown, the use of H₂SO₄ resulted in a considerable volume of gel. The percentage of gel-like silicate was ~63%, 49%, and 48% for H₂SO₄, HNO₃, and HCl leaching agents, respectively. The centrifugation of the gel-like silicate reduce the volume of the gel by 1.5 times.

Regarding the occurrence form of the valuable components in the silicate gel, it was found that during sulfuric acid leaching, the loss of REEs with the gel was due to the REE sulfate phases precipitating from the pregnant solution (Figure 4) as well as the precipitation of CaTi (SO₄)₃, whose crystals precipitate in a cubic or pseudocubic form (Figure 5).

DISTRIBUTION OF REEs TO THE LEACHING PRODUCTS. The usage of H₂SO₄ provided the most effective leaching of the concentrate: the Zr recovery was about 92%, the Σ REE recovery was about 82%, and the loss of mass was about 78%. In the cases of HNO₃ and HCl, the recovery of Zr was 76% and 84%, respectively, and the Σ REE recovery was 79% and 83%, respectively (

Table 8. Recovery of Zr and REE and losses with silicate gel under various leaching conditions and their concentrations in the pregnant solutions after the gel-like phase separation.

ACID	REE Recovery, %		Zr Recovery, %		Zr Loss with Gel, %	ΣREE Loss with the Gel,%	CZr in Pregnant Solution, g/dm3	CΣREE in Pregnant Solution, mg/dm3
	Gel	Solution	Gel	Solution
H2SO4	70.22	12.34	26.90	64.65	30.0	85.0	4.50	221.6
HCl	18.76	64.74	8.34	75.57	10.0	23.0	5.00	999.3
HNO3	44.32	35.38	41.04	35.83	54.0	56.0	2.40	494.2
HNO3(US treatment)	29.39	54.52	29.39	57.87	32.25	35.03	3.10	637.1

).

However, sulfuric acid leaching provided the most significant losses of valuable components with gel-like silicate: the losses of Zr and Σ REE were about 30% and 85%, respectively, while in the HNO₃ and HCl solutions, the losses of Zr and REE were about 54% and 56%, and 10% and 23%, respectively (Table 8). At the same time, it should be noted that the US treatment of the mineral pulp produced an increase in the recovery and a drop in the loss of valuable components. The increase in the REE recovery was about 4% and the recovery of Zr rose to nearly 10% (Table 8) compared with those of the experiment without US treatment (Table 8). The reduction in the loss of Zr and REEs with the gel -like silicate was ~21% (from 54.0% to 32.25%).

Based on the above, in purified pregnant solutions (after removing the silicate gel), the maximum concentration of Zr and REE cations was observed with HCl leaching, at ~5.04 and 1.0 g/dm³, respectively (Table 8).

However, despite the good indicators of the leaching with the usage of HCl, this technology has no future in industrial processes for many practical reasons. This is why HNO₃ was chosen as the leaching agent in further studies.

3.2. Effect of Preliminary EC and US Treatments on the Nitric Acid Leaching

The utilization of various kinds of energy [20,26] is the alternative way of the hydrometallurgical process intensification. In this study, the effect of energy (US and EC) treatments and their combination on the efficiency of eudialyte concentrate leaching was examined.

EFFECT OF EC AND US TREATMENTS ON EUDIALYTE SURFACE EC treatment led to a significant reduction in the areas of particles covered by gel-like silica (Figure 6a,b) and to an increase in the areas composed of parallel teeth. The trigonal symmetry of the parallel-oriented teeth inherited the symmetry of the three- and nine-membered rings of silicon–oxygen tetrahedra located along the c axis of the eudialyte crystal structure. This indicated that leaching processes proceeded most intensively along this direction. The qualitative elemental composition of these regions was close to that of the original eudialyte, but the content of the rare earth elements significantly decreased and became below the detection limit of the energy-dispersive spectrometer.

The eudialyte particles after US treatment during the leaching demonstrated intensive surface fracturing until complete breakdown and the absence of silicate gel deposition (Figure 6c).

The results clearly demonstrated that the mineral surface layer was noticeably depleted in Al, Na, and Mg (% at.). The surface concentration of Al, Na, and Mg decreased by 2.0–4.5, 3.0–5.5, and 2.5–5.0 times, respectively. These data support the conclusion that the concentrations of Al, Na, and Mg in a sequence of studied samples consistently decreased. This shows that US treatment and its combination with EC treatment is more effective for breaking down eudialyte silicate matrices.

Table 9. XPS results of the eudialyte surface layer.

TREATMENT	Atomic Concentration, at.%
	C	N	O	F	Na	Mg	Al	Si	K	Ti	Mn	Fe	Sr	Zr
1. None	4.7	0.6	60.0	0.6	12.2	0.5	2.8	14.8	0.4	0.9	0.5	0.7	0.4	1.2
2. HNO3	4.8	1.9	66.2	0.1	3.4	0.2	1.4	19.2	0.2	0.6	0.2	0.5	0.2	1.2
3. HNO3 + EC treat.—200 A/m2	6.2	1.9	63.8	0.2	4.1	0.2	0.9	18.3	0.2	1.3	0.5	0.5	0.3	1.2
4. HNO3 + EC treat. 200 A/m2 + US 20 kHz	6.1	0.9	66.6	0.1	2.2	0.2	0.6	21.0	0.1	0.6	0.2	0.3	0.2	1.0
5. HNO3 + US 20 kHz	4.9	1.2	67.7	0.2	2.4	0.1	0.7	20.2	0.1	0.7	0.2	0.5	0.1	1.1

presents XPS results obtained for the initial (untreated) eudialyte sample and for the samples after the various treatments.

The XPS and IR spectroscopy [24,25] data indicated that the eudialyte decomposition mechanism during the interaction with the leaching agent could be explained by the following processes: The first is associated with the selective extraction of alumina, sodium, and manganese. The second is associated with the breakdown of the silicate structure. At the same time, the concentration (at.%) of metal cations reduced, and the destruction of the silicate matrix was accompanied by the transformation of the silicate matrix into a gel-like phase. US and EC treatments significantly promoted the breakdown of the silicate matrix and improved the efficiency of eudialyte leaching.

3.3. Validation of the Optimized Process Modes and Process Flows for Nitric Acid Leaching of Eudialyte Concentrate

Special methods have been developed for identifying the dominant factors affecting various processes, which are called screening experiments. A mathematical model of a screening experiment with the preparation of a Plackett–Berman design was used in the study for the separation of the factors that insignificantly influenced the recovery of the valuable components with a minimum number of experiments, while ensuring the reliability of the results obtained (

Table 10. Screening experiment plan.

Factor	UStreat.	T, °C	CHNO3, g/dm3	EC Treat.	t, h	Fictive Factors				$\mathit{y}$,%
Base level		60	450		2
Variability interval		20	100		1
Lower level (−)	no	40	350	no	1
Higher level (+)	yes	80	550	yes	3
code	${\mathit{x}}_{\mathbf{1}}$	${\mathit{x}}_{\mathbf{2}}$	${\mathit{x}}_{\mathbf{3}}$	${\mathit{x}}_{\mathbf{4}}$	${\mathit{x}}_{\mathbf{5}}$	${\mathit{x}}_{\mathbf{6}}$	${\mathit{x}}_{\mathbf{7}}$	${\mathit{x}}_{\mathbf{8}}$	${\mathit{x}}_{\mathbf{9}}$	$\mathit{y}$
1	+	+	+	+	–	–	–	+	–	64.0
2	–	–	+	+	+	–	–	–	+	54.0
3	+	+	–	+	+	+	–	–	–	93.0
4	–	+	+	–	+	+	+	–	–	55.0
5	–	–	+	+	–	+	+	+	–	25.0
6	–	+	–	+	+	–	+	+	+	46.0
7	+	–	+	–	+	+	–	+	+	56.0
8	+	–	–	+	–	+	+	–	+	36.0
9	+	–	–	–	+	–	+	+	–	65.0
10	–	+	–	–	–	+	–	+	+	54.0
11	+	+	+	–	–	–	+	–	+	68.0
12	–	–	–	–	–	–	–	–	–	12.5
bi	11.29	10.96	1.29	0.62	9.12	0.79	−3.21	−0.71	−0.045

). The value of Zr recovery from the pregnant solution during the leaching of eudialyte concentrate was taken as a response. The influencing parameters (x_n) were: ultrasonic treatment (x₁), mineral pulp temperature in the leaching process (x₂), the concentration of nitric acid (x₃), electrochemical treatment (x₄), and leaching duration (x₅) (Table 10).

The most valuable coefficients were determined: $S_y^2$ = 29.91, $S_{b_i}^2$ = 2.8612, and ${∆}_{b_i}$ = 4.702, from the results of the 12 experiments [27,28,29,30].

The analysis of the obtained results proved that the major parameters influencing recovery were US treatment, temperature of the pulp, and duration of the process.

The experimental studies of the influencing parameters were performed for the selection of the optimal parameters. The analysis of the obtained results showed that the correlation between the Zr recovery and US treatment followed a simple equation. The most significant values were achieved during the stable US treatment of the pulp. This experiment condition improved the Zr recovery by 2.8 times. This increase in the Zr recovery under the US treatment of the pulp could be explained by the fracturing of the eudialyte grains [27]. US treatment inhibited the formation of gel-like silica on the grains surfaces.

The results in

Table 11. Effect of the number and duration of leaching stages on the recovery of valuable components.

N	Acid/Treatment	n	t, min	ε,%		ν,%
				Zr	REE	Zr	REE
1	HNO3	1	60	76.9	79.5	41	44.2
2	HNO3 (US treatment)	1	60	87.3	83.5	32.3	35.0
3	HNO3	2	60/60	86.5	81.4	51.9	50.8
4	HNO3	3	60/60/60	91.1	83.2	60.3	59.9
5	HNO3	3	40/40/100	96.2	92.8	41.5	44.7
6	HNO3 (1st-stage US treatment)	3	40/40/100	97.1	94.5	31.4	33.8
7	HNO3 (H2SO4 at 3rd stage)	3	40/40/100	95.8	91	41.5	44.7

Note: n—number of the stages; t—stage duration; ε—recovery (total); ν—loss of valuable component with gel.

illustrate the influence of the leaching stage number (stage duration was 1 h) on Zr and REEs recovery. Two stages of the process increased the Zr recovery by approximately 10% and the REE recovery by 2%. In this case, the volume of gel-like silicate increased by 11% (from ~55% to ~66%), which led to increases in the Zr and REEs loss by ~10% and ~6%, respectively (Table 11, lines 1 and 3). A further increase in the number of stages from two to three led to the growth in the Zr recovery by 4.5% and in the REEs recovery by 2% (Table 11, lines 3–4). The total volume of gel-like silicate increased to 70% and, as a consequence, the loss of Zr and REE of with three stages of leaching was 60%.

The decreases in REEs and Zr losses with by the lowering of the duration of the first and second leaching stages with an increase in the duration of the third stage (while the total process duration was unchanged) were investigated. Reducing the durations of the first and second stages and the lengthening of the third-stage during provided gains in the Zr recovery by approximately 5% (Table 11, lines 4 and 5) and in REE recovery by 10% (Table 11, lines 4 and 5).

At the same time, the volume of the gel-like silica insignificantly increased, while the loss of the valuable components was the same as that of the one-stage leaching process (Table 11, lines 1 and 5). The application of US treatment during the first leaching stage increased the total Zr and REEs recovery by 1% and approximately 1.5%, respectively (Table 11, lines 5 and 6), and reduced the volume of gel-like silica, which led to a drop in the loss of valuable components by 1.3 times.

The use of H₂SO₄ acid during the third leaching stage had a little impact on the leaching efficiency: Zr and REEs recovery slightly dropped (Table 11, line 7).

Thus, a three-stage leaching process is recommended with the following conditions: duration of the first stage, 40 min, second stage, 40 min; and third stage, 100 mi;, nitric acid concentration, 450 g/L; mineral pulp temperature, 80 °C; S:L ratio, 1:20; continuous US treatment of the mineral pulp in the first stage of leaching. The application of the mentioned procedure enabled a Zr recovery of 97.1% and a REE recovery of 94.5%

The technological scheme of eudialyte concentrate nitric acid leaching is depicted in Figure 7.

3.4. Recovery of Valuable Components from Pregnant Eudialyte Acid Leach Solution

3.4.1. Recovery by Chemical Precipitation

The process of neutralizing the pregnant solution to pH 4.0 using calcium carbonate is accompanied by the chemical precipitation of zirconium and metal impurities [31,32]. The precipitate forms hydroxides (zirconium aluminum, and iron), carbonates, and hydroxides (REEs). The real solution has high mineralization, which is why the actual processes and chemical reactions may be more complicated than the theoretical reactions and require additional study.

Figure 8 a shows the dependence of the recoveries of Zr, Al, and Fe into a solid residue as a function of pH at 20 °C.

The efficiency of element precipitation increase with the rise in pH. At pH 4.0, the recovery of Zr and Al was about of 100% and that of Fe was 40%. At the same time, recovery of sodium, manganese, strontium, titan and potassium was 15–35% and REEs loss was ~9%. The sharp increase in REE loss up to 40.0 % (due to co-precipitation with zirconium) was observed during the further pH rising. At that time, the CaCO₃ consumption raised from 375 to 600 g/dm³ due to the growth in solubility.

The separation of Zr and REEs must take place at pH 4.0–4.2 in order to achieve entire recovery of Zr, Al, and Fe. Further chemical precipitation of REE compounds should be carried out using Na₂CO₃ and NaOH. The obtained results indicate the highest REEs recovery (96%) was achieved at a Na₂CO₃ consumption of 6 g/dm³ (pH 6.0). With NaOH application, the highest REEs recovery was 59%.

Small rounded particles were found among the main mass of solid precipitate. They contained rare earth elements of the cerium group (La, Ce, and Nd), yttrium, and synthetic hydroxyl analogs of bastnasite-group minerals: hydroxylbastnesite (Ce, La, Y)CO₃(OH, F) or hydroxylsynchysite CaREE(CO₃)₂OH. At the same time, we could not exclude the deposition of aqueous rare earth carbonates calcinsite (Ce, La)₂(CO₃)₃·4H₂O, lanthanite (Ce, La, Nd)₂(CO₃)₃·8H₂O, or calcioanquilite CeCa(CO₃)₂(OH)·H₂O. Because of the micron size of these particles, it was impossible to correctly determine their composition.

It was determined that the Zr recovery was independent of the initial concentration; it was 99.6%–99.9%. However, the increase in the initial REE concentration from 150 to 1200 mg/dm³ led to increased loss with the residue from 0.3% to 8.5% due to the coprecipitation of rare earth elements with impurities [33,34].

3.4.2. Sorption Method for the Recovery of REEs and Zirconium from Pregnant Eudialyte Acid Leach Solution

Three types of sorbents were used in this study: a strongly basic anion exchange resin (Dowex 1 × −8 and Purolite PCA 433), a strongly basic cation resin (Purolite C 160), and a chelate resin (Puromet MTS 9500). As expected, the two grades of highly basic anion resins produced very similar results. Although they differed somewhat due to the fact that Purolite PCA 433 carries a quaternary ammonium base as a functional group, which is known to have more pronounced basic properties than its tertiary forms. Figure 9 and Figure 10 show the concentration curves obtained by the elution of a pregnant acid leach solution of eudialyte through the strongly basic polystyrene anion resins Purolite PCA 433 and Dowex 1 × −8.

The analysis of the obtained experimental data and the calculation of the material balance when using Purolite PCA 433 showed a sorption of 63.5% of Σ REE into the pregnant solution of the following elements: cerium (79.8%), lanthanum (67.5%), neodymium (99.5%), and praseodymium (81.1%) when eluate fractions 22 to 30 (0.88 to 1.20 of the column volume) were selected.

It was found that when eluting with water, the recovery of zirconium, lanthanum, cerium, and praseodymium from the resin was about 30%, 68%, 80%, and 20%, respectively. Thus, it was found that approximately 70% zirconium, 32% lanthanum, 20% cerium, and 80% praseodymium remained associated with the sorbent and could not be recovered in the reverse experiment (

Table 12. REE and Zr recovery with elution with various eluents.

	Recovery in H2Odistill., %	Recovery in 0.1 N HNO3, %	Recovery in 1.0 N HNO3, %
Zr	29.9	35.6	51.9
La	67.5	69.7	72.5
Ce	79.8	82.2	85.4
Pr	81.1	83.2	84.6
Nd	99.5	99.5	99.5

, column 1).

In order to desorb these elements, the sorption column was washed with 10 column volumes of the following solutions: 0.1 N and 0.5 N hydrochloric acid; 0.1 N, 0.5 N, and 8 N nitric acid; and 0.1 N sodium hydroxide at a flow rate of 1.0 mL/min. The results obtained are presented in Table 12, from which it can be seen that when washed with 0.1 N HNO₃, the recovery of zirconium increased by 5.7% (from 29.9% to 35.6%); those of lanthanum, cerium, and praseodymium increased by 2.1%–2.3%.

Elution with 1 N HNO₃ was found to perform better and made it possible to increase the recovery of zirconium from 29.9% to 51.9%, i.e., by 22%. At the same time, the recovery of lanthanum increased from 67.5% to 72.5%, that of cerium from 79.8% to 85.4%, and that of preseodymium from 81.1% to 83.5%, which led to the total recovery of REEs increasing to 73.7%.

Under the same conditions, when passing the refined REE solution through the strongly basic anion resin Purolite PCA 433, the recovery of REEs increased to 75.3% (Σ REE content) in the initial solution (75.8% Ce, 80.6% La, 64.5% Pr, and 46.5% Nd) and to 59.5% Zr. When passing the initial pregnant solution through a strongly basic anion exchange resin Dowex 1 × −8 in a single pass, 61.3% Σ REE was extracted into the bulk selective fraction: cerium (84.1%), lanthanum (87.9%), neodymium (44.1%), and praseodymium (28.5%), when selecting eluate fractions 22 to 30 (0.88 to 1.20 of the column volume). As can be seen from the data presented in the Figure 9, it is important that when using DOWEX X 8, in contrast to PCA433, zirconium is recovered along with heavy REEs and alkaline earth metals in the first three-quarters of the column volume.

Passing a refined REE solution through a sorbent improved the process performance and made it possible to recover 70% of Σ REE in the solution in the bulk selective fraction due to an increase in the recovery of cerium by 1.1% (from 86.9 to 97.9), lanthanum by 9.1% (from 75.0 to 84.1), praseodymium by 9.3% (from 57.1% to 66.4%), and neodymium by 8.8% (from 35.3% to 44.1%), with additional recoveries of 28.5% of europium, 18.1% of samarium, and 35% of holmium and lutetium.

The analysis of the experimental data obtained for the strongly acidic cation exchanger Purolite C 160 and the chelate resin Puromet MTS 9500 revealed the priority of the sorption of total REEs. The recovery on the cation exchanger C 160 was 95% REEs, 83% Zr, 72% Sr, 80% Fe, 85% Mn, 83% Ca, 55% Al, and 67% Na. For the Puromet MTS 9500 sorbent, the recovery of total REEs and Zr into the resin was 99%; that of the associated elements ranged from 20% to 95% (Na 21%, Al 95%, Ca 28%, Mn 38%, Fe 95%, and Sr 17%).

Because the use of Puromet MTS 9500 made it possible to achieve higher recoveries of valuable components into the resin, while reducing the sorption of accompanying impurities, further research was aimed at finding effective desorbing agents (eluents) for the selective recovery of REEs and other sorbed elements.

During desorption, the feed method and feed rate of eluents were similar to those used in the process of feeding solutions to the sorbent; the volume of the passed desorption solution was 5–10 times resin volume (column volume). The list of desorption solutions is given in

Table 13. Eluents and molarities.

Compound	Concentration, mol/L
H2SO4	0.1	0.25	0.5	1.0	2.0	5.0	10.0
HNO3	0.25	0.5	1.0	2.0	3.0	5.0	6.0	8.0
HCl	0.25	0.5	1.0	2.0	4.0	6.0
NH4HF2	0.25	0.5	1.0	2.0	3.0	5.0
EDTA	0.125 (pH 5.8)	0.125 (pH 4.5)	0.125 (pH 4.0)
Citric acid	0.25	0.4	0.8	1.5	2.0
Oxalic acid	0.20	0.4	0.8	1.5	2.0

.

From our analysis of the experimental data (Figure 11) and calculation of the material balances, we drew the following histograms regarding the performance of the eluents used. Citric acid and EDTA showed poor performance in the process of desorption of the elements adsorbed by Puromet MTS 9500. The use of a 0.25 M solution of ammonium bifluoride (NH₄HF₂) made it possible to recover 98.9% of the zirconium, without any loss of REEs.

The use of 0.2 M oxalic acid as an eluate made it possible to recover 97.2% of the Na, while the concentration of rare earth elements in the eluate was below the detection threshold with the 0.05 ppm method, in other words, without the loss of REEs.

Sulfuric acid (H₂SO₄) at a concentration of 0.1 mol/L was an effective elution solution for Ca and Mn: when passing five column volumes (150 mL), 34% and 33% passed into the solution, respectively. Under the same conditions, sulfuric acid at a concentration of 10 mol/L caused the transition into the solution of 35% Na, 39% Ca, 37% Fe, and 38% Al. At the same time, the loss of REEs with the eluate of a 0.1 M H₂ SO₄ solution was 0.5% adsorbed on resin and 1.7% in the eluate of a 10 M solution (Figure 11).

Hydrochloric acid at concentrations of 0.5–2.0 mol/L, when passing fives column volumes (150 mL), caused the transition to the elution solution of 22.54% to 36.3% Fe; at the same time, 2.31% to 9.29% of the REEs passed into the eluate (Figure 11b).

Finally, concentrated (5 and 8 M) solutions of nitric acid (HNO₃) were found to be effective eluting agents for REEs. The 8 M solution performed better: when passing five column volumes, a recovery of 83.8% REEs was achieved; for 5 M, the recovery was 9.8% REEs. At the same time, 13.5% Al and 8.5% Ca passed into the 10 M nitric acid. The composition and concentration of REEs in the final product of REE desorption are shown in

Table 14. REE concentration in the effluent (8 M HNO₃).

La	Ce	Pr	Nd	Sm	Eu	Gd	Tb	Dy	Ho	Er	Tm	Yb	Lu	ΣREE
mg/L
90.54	218.03	26.88	78.33	24.20	7.99	27.85	5.04	28.14	5.87	15.77	2.39	13.57	1.92	715.4

.

Based on the data obtained, two processes for the selective recovery of Zr and REEs from pregnant solutions of eudialyte concentrate leaching were proposed: sorption and precipitation–sorption (Figure 12).

The sorption extraction process involved feeding the pregnant solution in its native form to the Puromet MTS 9500 sorbent and the subsequent recovery from the sorbent by stepwise elution with mineral acid solutions (10–15 column volumes each, which ensured 100% removal of impurities), first of sodium; second of calcium, manganese, iron, and aluminum; third of REEs; and, finally, of zirconium into a solution of ammonium bifluoride.

The precipitation–sorption process involves sequential alkalization of the initial pregnant solution with calcium carbonate in order to isolate the zirconium product (pH 4) and REE carbonates with further introduction of sodium carbonate to pH 6.1. In order to remove impurities (Fe, Al, Ca, etc.), the resulting REE carbonates are proposed to be dissolved in 50% nitric acid and then REEs can be recovered into sorbents, followed by the selection of impurities and valuable REEs by elution with mineral acid solutions.

4. Conclusions

The mechanism of the effect of various acids on the structural and chemical properties and dissolution of eudialyte concentrate was revealed. It was determined that during the interaction of eudialyte with HCl and H₂SO₄, the mineral surface is described by the appearance of a series of deep, parallel-oriented fractures because of the predominance of the normal dissolution rate. When interacting with HNO₃, the tangential (layered) mechanism of dissolution dominates.

In the nitrogen–sulfuric–hydrochloric acid sequence, a consistent decrease in the content (at. %) on the eudialyte surface was identified for Al by 90%; Na by 98.4%; for Ca, Mg, and Mn by 75%–80%; for Ti by 88%, and for Zr by approximately 92%. The phase composition of the silicate gel for all acids was characterized by Zr compounds, REEs, and nanofragments of the main concentrate minerals. According to the silicate gel formation efficiency and, accordingly, Zr and REE loss, the acids could be ranked as follows: sulfuric acid, nitric acid, and hydrochloric acid.

The most promising leaching results were obtained with HCl and HNO₃, which provided comparable qualitative and quantitative process indicators. Because the application of HCl is difficult because of many practical reasons, it cannot be recommended for industrial applications of the eudialyte concentrate leaching process.

The influence of combined energy treatment (thermal, electrochemical, and ultrasonic treatment) was determined. It consisted of dispersing the colloidal silicate gel, preventing the formation of its precipitates on the surface of mineral particles, and the formation of numerous defects and microcracks on the surface of eudialyte grains up to breakage during the ultrasonic treatment. It led to increases in the intensity and rate of the leaching process.

The greatest intensifying effect on the process of concentrate acid leaching was exerted by ultrasonic treatment of the mineral pulp. It enabled the highest Zr and REE recovery at a coarser feed size, lower slurry temperatures, and shorter leaching duration, as well as a reduction in REE loss to the gel by a factor of 2–2.5.

A highly efficient leaching process was developed for eudialyte concentrate, the practical implementation of which ensures a Zr recovery of 97.1% and a REE recovery of 94.5%, with significant reductions in silicate gel-like formation and valuable components loss by 21.0%, as well as increases in Zr and REE recovery by 10 % and 4%, respectively.

Our study demonstrated the high performance of Zr and REEs chemical precipitation. It was established that an increase in the pH of the initial pregnant solution during the first stage of precipitation with CaCO₃ to 4.0 provided 99.95% Zr recovery into the solid phase and the associated recoveries of impurities: 100% of Al and 43.8% of Fe. The total recovery of REEs into the solid phase was as high as 74.6%, and the total loss of REEs with the product of the precipitation was 9%. A further increase in the pH of the solution using Na₂CO₃ to 6.1 provided an REE recovery into the solid phase of 87.5%, with a total recovery from solution of 95.5%.

It was found that the use of the Puromet chelate resin MTS 9500 as a sorbent provided the highest (99%) recovery of REEs and Zr from the productive solutions. As a result of the subsequent stepwise gradient elution of the saturated sorbent, the following results were obtained: a Zr solution with a concentration of 4.1 g/L (Zr recovery into the solution was as high as 91.7%) and a REE solution with a concentration of 0.75 g/L (REE recovery 81.1%).

A process for the selective recovery of Zr and REEs from the pregnant leaching solutions of eudialyte concentrate based on a combination of chemical precipitation, sorption, and elution methods was proposed. Two process flows for treating pregnant solutions of eudialyte acid leaching were developed: sorption and precipitation–sorption.