Comprehensive Utilization Beneficiation Process of Lithium Pegmatite Ore: A Pilot-Scale Study

Abstract

Pegmatite ores, the primary and technologically advanced lithium (Li)-bearing minerals, comprise various rare metal-based elements, including niobium (Nb), tantalum (Ta), tin (Sn), and beryllium. With increasing Li demand, global exploitation of pegmatite ores has generated vast tailings, mainly comprising quartz and feldspar. However, the process for comprehensively utilizing valuable minerals from pegmatite ores remains undeveloped, and the persistent gap between laboratory studies and industrial practice hinders the sustainable advancement of the pegmatite mineral processing industry. Herein, a comprehensive utilization beneficiation process was designed and validated at both laboratory- and pilot-scale levels. Locked-circuit flotation tests at the laboratory-scale on spodumene and feldspar yielded (i) an Li concentrate with an Li₂O grade of 5.80% and recovery of 88.62%, and (ii) a feldspar concentrate with a (K₂O + Na₂O) grade of 11.41% and good recoveries of K₂O (81.30%) and Na₂O (84.81%). In a 72 h continuous pilot-scale test, an Li flotation concentrate with an Li₂O grade of 5.72% and recovery of 86.78%, and a final Li concentrate with an Li₂O grade of 5.89% and recovery of 86.56% were obtained. Using Li flotation tailings as feed, a feldspar concentrate with a (K₂O + Na₂O) grade of 11.41% was obtained, achieving K₂O and Na₂O recoveries of >75%. The proposed process realizes nearly overall mineral recovery from the pegmatite ores, producing qualified concentrates of Li, Nb–Ta, Sn, feldspar, and quartz. In water reuse feasibility tests, ferrous sulfate (FeSO₄) was identified as the optimum flocculant at a dosage of 1000 g m⁻³. In the locked-circuit test with returned water, the consumption of sodium hydroxide (NaOH), sodium carbonate (Na₂CO₃), and EMT-12 (collector) was reduced by 18.75%, 3.33%, and 3.45%, respectively, while the flotation indices of the Li concentrate (Li₂O grade of 5.77% and recovery of 86.47%) were slightly lower than those in freshwater. In addition to increasing economic benefits, the process offers considerable reductions in tailings disposal, full utilization of multiple elements, and a potential decrease in water and reagent consumption. This study provides important guidelines for the mineral processing of Li pegmatite and other associated multimetallic ores.

1. Introduction

The global demand for clean and sustainable energy is increasing. Hence, lithium (Li) has become essential in several fields, such as electric transportation, mobile devices, and energy storage systems, because it has a higher electrochemical equivalent, higher heat capacity, and lower environmental hazards compared to fossil fuels [1,2,3]. Data from the United States of America Geological Survey revealed that Li production increased by 192.7% (from 82,000 to 240,000 metric tons of Li carbonate equivalent) from 2020 to 2024 [4,5]. The global Li demand is expected to continue increasing in the future [6,7].

Pegmatite ores are widely known as the primary and technologically advanced Li-bearing minerals [8]. With the increase in Li demand, the exploitation of pegmatite ores has significantly increased in recent years. However, a huge volume of tailings is generated during spodumene concentrate production owing to the low Li content in LiO₂ [9]. Long-term storage of tailings causes serious environmental impacts, including metal ion leaching [10,11,12], water contamination [13], and soil occupation and pollution [14]. Meanwhile, treating tailings require advanced technologies—such as alkali-activated methods for geopolymer production and microorganisms to induce carbonate precipitation [9]—and incurs high costs.

The associated minerals of pegmatite ores commonly comprise quartz, feldspar, and accessory minerals containing rare metal elements such as niobium (Nb), tantalum (Ta), tin (Sn), beryllium (Be), rubidium (Rb), and cesium (Cs) [15,16]. Recently, Nb, Ta, Be, and Sn have attracted considerable interest because of their superior performance in aerospace [17], high-end manufacturing [18], electrochemical energy storage, superconducting materials [19], and alloy production [20]. Rb and Cs exhibit unique advantages in high-tech industrial applications including thermoelectric generators, atomic clocks, laser cooling, and medicine [21,22]. Notably, Rb does not form minerals in nature but can be enriched during the beneficiation of potassium (K) minerals such as K-feldspar and mica [23]. Feldspar and quartz, generally regarded as gangue minerals in the beneficiation of the pegmatite ores [24], are also raw materials used in manufacturing ceramic materials [25] and glass [26]. From the perspective of sustainable development, the comprehensive utilization of these associated minerals is of great significance during the beneficiation of pegmatite ores, as it alleviates shortages of rare metals and reduces the burden of tailings disposal, improving the utilization efficiency of mineral resources.

In recent decades, numerous studies have focused on the separation of spodumene from pegmatite ores, typically using gravity separation, froth flotation, and magnetic separation [27]. Dense or heavy media separation is effective when high liberation occurs at coarse grain sizes [28] or for removing heavy minerals such as Nb–Ta and cassiterite [29]. Magnetic separation is used to remove iron-bearing gangue minerals [30]. Froth flotation is the primary method for separating spodumene from aluminosilicate minerals, including feldspar and mica. Collector selectivity is a key factor in spodumene separation owing to the presence of similar Al³⁺ sites on mineral surfaces. A mixture of anionic and cationic collectors has been used to improve selectivity in spodumene flotation from feldspar [31,32,33]. Spodumene in pegmatite ores is currently well recovered in production, while the recoveries of minerals bearing rare metal elements remain low [34]. The main issue in the comprehensive utilization of pegmatite ores lies in effectively separating associated minerals. However, it is generally believed that the grades of the associated elements must exceed industrial thresholds—Nb₂O₅ ≥ 0.022%, Ta₂O₅ ≥ 0.022%, Sn ≥ 0.2%, and BeO ≥ 0.04%—otherwise, their recovery is not economical [35].

Limited studies have reported the integrated recovery of multiple valuable minerals from pegmatite ores. The process flow of coarse grinding–gravity pre-enrichment–high magnetic separation–centrifugal separation [36] was reported to recover Ta and Nb from the pegmatite ore before spodumene beneficiation, yielding a Ta–Nb concentrate with tantalum oxide (Ta₂O₅) and niobium oxide (Nb₂O₅) grades of 13.90% and 29.14% and recoveries of 49.50% and 58.37%, respectively. Moreover, a tin concentrate with a Sn grade of 41.45% and recovery of 54.39% was obtained. High-gradient magnetic separation coupled with magnetic fluid was also reported to enhance Ta–Nb recovery associated with spodumene, producing a final concentrate with (Ta, Nb)₂O₅ grade of 39.70% and overall recoveries of 51.05% (Nb₂O₅) and 45.36% (Ta₂O₅) [37]. The process of primary grinding–desliming–coarse mica flotation–spodumene flotation was proposed to achieve flotation separation of mica and spodumene, yielding a spodumene concentrate with Li₂O grade and recovery of 6.02% and 87.34%, respectively. Recycling feldspar from spodumene tailings has also been researched, producing a (K₂O + Na₂O) grade and recovery of 11.33% and 61.33%, respectively, through magnetic separation and froth flotation [24]. However, a systematic process for the comprehensive utilization of valuable minerals in pegmatite ores remains undeveloped, and the persistent gap between laboratory studies and industrial practice continues to hinder sustainable development of the pegmatite mineral processing industry.

In addition to the issue of comprehensive utilization of associated minerals, substantial water requirement and polluted water discharge are other concerns in the beneficiation of pegmatite ores [38]. The feasibility of water reuse, along with its effects on and mechanisms in the beneficiation performance of multiple minerals, has been studied [39,40]. Ma et al. [41] introduced an electrocatalytic iron/carbon microelectrolysis system to achieve 97.59% removal of lead (Pb²⁺) and zinc (Zn²⁺) ions, butyl xanthate, and dianilino dithiophosphoric acid from Pb–Zn flotation wastewater, enabling water reuse at appropriate ratios. Dissolved air flotation combined with coagulation/flocculation treatment was also studied for wastewater reuse in apatite flotation [42]. An industrial-scale test of wastewater reuse at the Shizhuyuan Polymetallic Dressing Plant, China, reported a 34.62% reduction in freshwater consumption and an 18.56% reduction in reagent consumption [43]. However, the water reuse during Li pegmatite ore beneficiation process has rarely been studied.

To solve the problem of effective recovery of valuable minerals in the pegmatite ore and narrow the gap between laboratory studies and industrial practice, a comprehensive utilization beneficiation process was designed and proposed in this study, validated through both laboratory- and pilot-scale tests. The feasibility of water reuse from the slurry of Li flotation tailings was also investigated, and a locked-circuit test was conducted to evaluate the effect of returned water on spodumene flotation indices. The proposed process realizes overall mineral recovery from the pegmatite ore, separating concentrates of Li, Nb–Ta, Sn, feldspar, and quartz. In addition to enhancing economic benefits, the proposed process significantly decreases tailings disposal, recovers multiple elements, and reduces water and reagent consumption.

2. Experimental Methods

2.1. Materials and Reagents

2.1.1. Ore Samples

The Li pegmatite ore used in laboratory- and pilot-scale beneficiation studies was collected from the Keeryin area of western Sichuan Province. Multielemental characterization of the raw ore was conducted after acid dissolution via atomic absorption spectroscopy and inductively coupled plasma mass spectrometry.

Table 1. Chemical composition analyses of the Li pegmatite ore (wt.%).

Element	Li2O	BeO	Nb2O5	Ta2O5	Sn	SiO2	MgO	Al2O3	Fe2O3
Content (%)	1.18	0.0612	0.0124	0.0105	0.0479	72.35	0.031	15.60	1.13
Element	K2O	Na2O	CaO	MnO	P2O5	Rb2O	Cs2O	TiO2
Content (%)	2.60	4.45	0.56	0.13	0.33	0.11	0.017	0.030

presents the analysis results, showing that the ore sample primarily comprises 1.18% Li₂O, 72.35% SiO₂, and 15.60% Al₂O₃. In addition to Li₂O as the primary recoverable component, Nb₂O₅, Ta₂O₅, Rb₂O, and BeO also meet the thresholds for comprehensive utilization. However, the Sn content is below the threshold.

The X-ray diffraction (XRD) pattern of the ore sample is presented in Figure 1. The contents of the main minerals in the ore sample were determined using optical microscopy, energy-dispersive micro-area analysis, and the advanced mineral identification and characterization system (

Table 2. Mineral compositions of the lithium pegmatite ore (wt. %).

Mineral	Spodumene	Albite	Quartz	Potassium Feldspar	Muscovite	Bioclase	Chlorite
Content (%)	14.09	31.74	27.87	15.49	4.28	1.36	1.27
Mineral	Bauxite	Kaolinite	Amphibole	Nb–Ta–Fe Minerals	Cassiterite	Other Trace Minerals
Content (%)	0.73	0.71	0.51	0.01	0.01	1.93

). Based on the characterizations, spodumene was identified as the main recoverable mineral, along with Nb–Ta-bearing minerals. No independent carrier minerals for Be and Rb were identified. Be was associated with albite, and Rb with potassium feldspar, which makes their recovery difficult. Moreover, the high contents of quartz (27.87%) and feldspar (47.23%) indicate potential for comprehensive recovery.

2.1.2. Reagents

In the flotation tests, sodium carbonate (Na₂CO₃), sodium hydroxide (NaOH), sulfuric acid (H₂SO₄), and calcium chloride (CaCl₂) were used as regulators. EMT-12, the cationic–anionic combined collector for spodumene, was optimized by mixing laurylamine, oxidized paraffin wax soap, and sodium oleate [24]. The collector EMS-9, a novel cationic–anionic combined collector primarily comprising dodecyl trimethyl ammonium chloride (DTAC) and sodium dodecyl sulfonate, was used for the flotation of feldspar. The reagents used in lab- and pilot-scale flotation were of analytical (≥97%) and industrial grades (≥80%), respectively. Tap water was used for all flotation tests.

2.2. Processes and Devices

Before pilot-scale tests, laboratory-scale tests on reagent conditions and locked-circuit flotation were conducted to provide fundamental research data.

2.2.1. Laboratory-Scale Tests

Condition Flotation Tests

A 500 g sample of Li pegmatite ore used in laboratory-scale batch flotation tests was ground to 75.6% passing 0.074 mm using a ball mill (XMQ; 240 mm × 90 mm, Wuhan Exploration Machinery Co., Ltd., Wuhan, China) with Na₂CO₃. The target fineness was selected based on previous flotation tests assessing the effect of particle size on flotation indices as well as liberation analysis. Li tailings served as the feed for feldspar flotation. Flotation tests were performed using XFD-type flotation machines with 3.0, 1.5, 1.0, 0.75, or 0.5 L cell. The pulp density was set at 30–35%, and cell selection was based on the weight of the sample. The flowsheets of spodumene and feldspar are shown in Figure 2 and Figure 3, respectively. The detailed reagent dosages are presented in Section 3.1. Each test was repeated three times, and the average value was reported, with error bars representing the standard deviation.

Locked-Circuit Tests

• Locked-Circuit Test of Spodumene

The flowsheet and flotation conditions for the locked-cycle test of spodumene are illustrated in Figure 4. The laboratory-scale locked-circuit test, with one rougher and stage scavenger and three-stage cleaners, was conducted. The grinding conditions matched those used in the flotation condition tests. Reagents included NaOH, CaCl₂, and EMT-12 in the rougher; Na₂CO₃ in the cleaner; and NaOH with EMT-12 in the scavenger stages. All reagents were prepared as 5 wt.% solutions. Each middling was recycled to the preceding beneficiation stage, ultimately producing Li concentrate and tailings. Locked-circuit flotation tests of spodumene and feldspar were conducted with five recycles of the middlings.

• Locked-Circuit Test of Feldspar

The flowsheet and flotation conditions for the feldspar locked-cycle test are presented in Figure 5. Using Li tailings obtained from the previous locked-circuit test as the feed, the laboratory-scale locked-circuit test was performed using a flowsheet comprising one rougher and scavenger stage and three cleaner stages. During the process, H₂SO₄ and EMS-9 were used as the regulator and collector, respectively. The recycling of middlings followed a procedure similar to that used for spodumene flotation.

2.2.2. Pilot-Scale Flotation Tests

The pilot-scale flotation tests were continued for 72 h after the stabilization of indices to test the stability of flotation indices.

Pilot-Scale Flotation Tests of Spodumene

The appropriate processing capacity for the pilot-scale continuous test was set at 1.0 ton per day (t d⁻¹), based on the capacity of the selected equipment and the required grinding fineness. To ensure that the grade of the concentrate met the desired specifications, a fourth-stage cleaner was added to the pilot-scale flotation process of spodumene based on the locked-circuit test, while the rest remained the same as the locked-circuit test. The diagram showing the equipment interconnections is presented in Figure 6, with the volume of each flotation tank labeled.

To monitor indices and calculate mass–quality flowsheet data, grab (random sampling) and shift (composite sample per shift, i.e., 8 h) samples of the Li flotation concentrate, tailings, and raw ore were analyzed for Li₂O grade. The standard deviations of the Li flotation concentrate and tailings grades were calculated from shift-sample data to evaluate flotation stability using Equation (1):

$$\sigma =\sqrt{{\displaystyle \frac{1}{N}}{\sum }_{i=1}^{\mathrm{N}}{\left(x_i-\mu \right)}^2},$$

(1)

where $\sigma$ is the standard deviation of the Li₂O grade of the products, N is the number of samples, $x_i$ is the Li₂O grade data, and μ is the average Li₂O grade.

The mass–quality balance flowsheet calculations were based on the average results of shift samples and process stream samples. The yields of the concentrate and tailings were determined from the measured Li₂O grades. Li₂O recovery, defined as the proportion of Li₂O in the products, was then calculated based on the metal mass. The Li₂O grade and recovery of the middlings were calculated from the sum of the separated products using the metal mass. The detailed calculation procedure is provided in Equations (2)–(7):

Mass balance: γ_α = γ_β + γ_θ

(2)

Metallurgical balance: γ_αα_Li = γ_ββ_Li + γ_θθ_Li

(3)

given γ_α = 100%.

Hence,

$${\mathsf{\gamma }}_{\mathsf{\beta }}={\displaystyle \frac{{\mathsf{\alpha }}_{\mathrm{L}\mathrm{i}}-{\mathsf{\theta }}_{\mathrm{L}\mathrm{i}}}{{\mathsf{\beta }}_{\mathrm{L}\mathrm{i}}-{\mathsf{\theta }}_{\mathrm{L}\mathrm{i}}}}\times 100(\%)\mathrm{and}$$

(4)

γ_θ = 1 − γ_β

(5)

As a result,

$${\mathsf{\epsilon }}_{\mathsf{\beta }}={\mathsf{\gamma }}_{\mathsf{\beta }}{\displaystyle \frac{{\mathsf{\beta }}_{\mathrm{L}\mathrm{i}}}{{\mathsf{\alpha }}_{\mathrm{L}\mathrm{i}}}}\times 100(\%)\mathsf{and}$$

(6)

ε_θ = 1 − ε_β,

(7)

where γ_α is the yield of the raw ore, γ_β is the yield of the concentrate, γ_θ is the yield of tailings, α_Li is the Li₂O grade of the raw ore, β_Li is the Li₂O grade of the concentrate, θ_Li is the Li₂O grade of tailings, ε_β is the recovery of the concentrate, and ε_θ is the recovery of tailings.

Pilot-Scale Flotation Tests of Feldspar

Li flotation tailings were used as the feed for the recovery of feldspar, with a thickener to increase the pulp concentration. To increase the recovery of feldspar, the flowsheet added a second-stage scavenger based on the laboratory-scale test, i.e., a one-stage rougher–two-stage scavenger–three-stage cleaner flowsheet in the pilot-scale test. The diagram of the equipment interconnections is presented in Figure 7. The sampling method and mass–quality balance flowsheet calculations for feldspar are identical to those for spodumene.

Devices

The equipment used in the pilot-scale tests are listed in

Table 3. Devices used in the pilot-scale flotation test.

Device	Model	Manufacturers
Overflow continuous rod mill	XMBL-Ф420 × 600	Wuhan Exploration Machinery Co., Ltd., Wuhan, China
Single-spiral classifier	Ф150 × 1000	Wuhan Exploration Machinery Co., Ltd., Wuhan, China
Mixing tank	XDC-30L	Wuhan Exploration Machinery Co., Ltd., Wuhan, China
Mechanical agitated continuous flotation machine	FX2-24L/12L	Jilin Jitan Machinery Co., Ltd., Changchun, China
Thickener	50L	Self-made
Low-intensity drum magnetic separator	XCRS-Φ400 × 240	Wuhan Exploration Machinery Co., Ltd., Wuhan, China
High-intensity magnetic separator	SLon-500	Slon Magnetic Separator Co., Ltd., Ganzhou, China

.

3. Results and Discussion

3.1. Laboratory-Scale Flotation Tests

3.1.1. Laboratory-Scale Flotation of Spodumene

Effect of Collector Species and Dosage

To investigate the flotation performance of the mixed collector EMT-12, comparative tests were conducted on collector types and dosages under the following conditions: (i) grinding fineness of −0.074 mm at 75.6%, (ii) 1500 g/t Na₂CO₃ added to the mill, and (iii) NaOH and CaCl₂ dosages of 500 and 200 g/t, respectively. The grinding fineness was set at 75.6% based on previous experiments, which achieved optimal liberation and maximum Li₂O recovery, consistent with the results of other studies [24]. The flotation performance of EMT-12 and an oxidized paraffin wax soap (OPWX) at different dosages was compared on spodumene (Figure 8). The yield and Li₂O recovery of the rough concentrate gradually increased with increasing dosages of both collectors, while the Li₂O grade decreased. However, Li₂O recovery and grade were consistently lower with OPWX as the collector than those achieved using EMT-12. Notably, when a dosage of EMT-12 exceeded 3000 g/t, the increase in Li₂O recovery becomes less significant. At a dosage of 2500 g/t EMT-12, the Li₂O grade and recovery reached 3.97% and 89.63%, respectively, with 0.61% and 10.85% more than those obtained using OPWX at the same dosage. Thus, EMT-12 is a better collector for spodumene flotation, which is attributed to the synergistic effect and co-adsorption of the mixed collectors [44].

Locked-Circuit Test

Table 4. Results of locked-circuit test for spodumene flotation.

Product	Yield (%)	Li2O Grade (%)	Li2O Recovery (%)
Li flotation concentrate	17.68	5.80	88.62
Li tailings	82.32	0.16	11.38
Feed	100.00	1.16	100.00

shows that the nondesliming process yields an Li flotation concentrate with an Li₂O grade of 5.80% and a recovery of 88.62%. The mass–quality balance flowsheet is shown in Figure 9. The distribution of other valuable elements in the products is summarized in

Table 5. Distribution of other valuable elements in spodumene flotation products of the locked-circuit test.

Product	Grade (%)				Recovery (%)
	Nb2O5	Ta2O5	BeO	Sn	Nb2O5	Ta2O5	BeO	Sn
Li flotation concentrate	0.0573	0.0475	0.235	0.215	85.42	84.30	67.50	79.10
Li tailings	0.0021	0.0019	0.0243	0.0122	14.58	15.70	32.50	20.90
Feed	0.0119	0.0100	0.0616	0.0481	100.00	100.00	100.00	100.00

. These results indicate that 85.42% Nb and 84.30% Ta are concentrated in the Li concentrate with spodumene, creating favorable conditions for their comprehensive recovery in subsequent processes. Furthermore, 79.10% of Sn is enriched in the concentrate with a grade of 0.215%, warranting comprehensive recovery experiments for Sn to enhance its utilization. Notably, 67.50% of Be is enriched in the concentrate. However, its concentration (0.235%) remains low, and no independent Be-bearing mineral is present; therefore, Be separation was not considered.

3.1.2. Laboratory-Scale Flotation of Feldspar

Effect of Collector Species and Dosage

The flotation recovery of feldspar from Li tailings was investigated under the condition of 3000 g/t H₂SO₄. Herein, H₂SO₄ was used as an alternative to the commonly used hydrofluoric acid (HF) owing to environmental concerns. The performances of DTAC and EMS-9 as collectors were evaluated. Figure 10 shows that increasing the dosage of the collectors gradually increases the yield and recovery of (K₂O + Na₂O) in the rough concentrate. The grade and recovery indices of the concentrate are higher with EMS-9 than with DTAC. However, when the dosage of EMS-9 is >1500 g/t, the increase in (K₂O + Na₂O) recovery becomes marginal. Excessive dosages of EMS-9 cause disordered flotation, forming sticky and unstable foam. Thus, a dosage of 1500 g/t of EMS-9 is considered optimal for the rough stage of feldspar flotation, obtaining a 10.13% (K₂O + Na₂O) grade and >86% recovery in the rough concentrate.

Locked-Circuit Test

The results of the locked-circuit test for feldspar recovery in Li tailings are presented in

Table 6. Results of the locked-circuit test for feldspar flotation.

Product	Yield (%)	Grade (%)	Recovery (%)
		K2O + Na2O	K2O	Na2O
Feldspar concentrate	60.45	11.41	81.30	84.81
Tailings	39.55	3.45	18.70	15.19
Feed	100.00	8.26	100.00	100.00

. A qualified feldspar concentrate with an 11.41% (K₂O + Na₂O) grade was obtained. Operational recoveries of K₂O and Na₂O reach 81.30% and 84.81%, respectively, while their recoveries relative to the raw ore were 80.00% and 80.85%, demonstrating efficient feldspar recovery compared with other studies [24].

3.2. Pilot-Scale Tests

According to the grinding fineness of −0.074 mm at 75.6% in the laboratory-scale test, the feed capacity was set at 42 kg h⁻¹, equivalent to 1.01 t d⁻¹. During the 72 h continuous test, the fineness of the grinding product was sampled and sieved per hour, stabilizing in the range of 75–80%, with an average fineness of −0.074 mm at 77.1%.

3.2.1. Pilot-Scale Flotation Test of Spodumene

Owing to the return of the slurry, the reagent dosages used in the pilot-scale test were adjusted based on the laboratory-scale test during the debugging period, which were checked every 30 min. The results were averaged as the reported dosages. The reagent scheme is shown in

Table 7. Reagent scheme for the pilot-scale flotation test of spodumene.

		Reagent	Na2CO3	NaOH	CaCl2	EMT-12
	Dosage (g/t)
Stage
Rougher			3000	300	80	1500
Scavenger I			-	100	-	400
Scavenger II			-	-	40	250
Cleaner I			1200	-	-	-
Cleaner II			750	-	-	-
Cleaner III			550	-	-	-
Cleaner IV			350	-	-	-
Total			5850	400	120	2100

.

The Li₂O grade data of grab and shift samples of the Li flotation concentrate and tailings are shown in Figure 11. The average Li₂O grades of the concentrate from grab and shift samples are 5.75% and 5.72%, respectively, meeting the requirement of the qualified spodumene concentrate (≥5.5%). The Li₂O grades of the flotation tailings from both samples are effectively controlled at ~0.20%. These results closely match those of the Li concentrate grade obtained from the laboratory-scale locked-circuit test. The standard deviation of Li₂O grade of shift samples are 0.144 and 0.024 for Li flotation concentrate and tailings, respectively. Overall, the pilot-scale flotation test of spodumene demonstrates consistent stability, which is significant for pilot-scale tests and subsequent industrial production.

The results of mass–quality balance flowsheet calculations are shown in

Table 8. Calculations of indices in the pilot-scale test of spodumene flotation.

Product	Yield (%)	Li2O Grade (%)	Li2O Recovery (%)
Li flotation concentrate	17.90	5.72	86.78
Li tailings	82.10	0.19	13.22
Feed	100.00	1.18	100.00

, with the corresponding flowsheet presented in Figure 12. The distribution of Nb, Ta, Be, and Sn in the products is shown in

Table 9. Distribution of other valuable elements in the spodumene flotation products of pilot-scale test.

Product	Grade (%)					Recovery (%)
	Nb2O5	Ta2O5	BeO	Sn	Fe2O3	Nb2O5	Ta2O5	BeO	Sn	Fe2O3
Li flotation concentrate	0.0562	0.0461	0.2270	0.2030	4.30	83.05	82.72	65.56	78.11	69.06
Li tailings	0.0025	0.0021	0.0260	0.0124	0.42	16.95	17.28	34.44	21.89	30.94
Feed	0.0121	0.0100	0.0620	0.0465	1.10	100.00	100.00	100.00	100.00	100.00

. Table 8 shows that an Li₂O concentrate with a recovery of 86.78% and a grade of 5.72% is obtained in the pilot-scale test, meeting the requirements for subsequent metallurgical processing (YS/T 2016-2011). Table 9 shows that 83.05% Nb₂O₅, 82.72% Ta₂O₅, and 78.11% Sn from the raw ore were enriched in the Li flotation concentrate, with grades of 0.05620%, 0.0461%, and 0.2030%, respectively. As stated in Section “Locked-Circuit Test”, the separation of BeO from this ore via beneficiation is infeasible and was not considered. The indices agree well with the laboratory-scale locked-circuit results.

3.2.2. Pilot-Scale Flotation Test of Feldspar

Similar to the pilot-scale test of spodumene, reagent dosages were modulated based on the flotation phenomenon and indices. The specific reagent scheme is illustrated in

Table 10. Reagent scheme of the pilot-scale flotation test of feldspar.

		Reagent	H2SO4	EMS-9
	Dosage (g/t)
Stage
Rougher			3600	1400
Scavenger I			650	200
Scavenger II			300	150
Cleaner I			1200	200
Cleaner II			800	150
Cleaner III			450	-
Total			7000	2100

.

The K₂O and Na₂O grades of grab and shift samples of feldspar concentrate and tailings are presented in Figure 13. The average (K₂O + Na₂O) grades of the concentrate from grab and shift samples are 11.37% and 11.41%, respectively, meeting the requirement for the feldspar concentrate (≥11.00%). The standard deviation of (K₂O + Na₂O) grades of shift samples are 0.063 and 0.082 for feldspar flotation concentrate and tailings, respectively. The results exhibit stable indices, consistent with those from the laboratory-scale locked-circuit test.

The mass–quality balance flowsheet calculation results are shown in

Table 11. Calculations of indices in the pilot-scale test of feldspar flotation.

Product	Yield (%)	Grade (%)	Li2O Recovery (%)
		K2O + Na2O	K2O	Na2O
Feldspar concentrate	55.80	11.41	75.46	77.87
Feldspar tailings	44.20	4.31	24.54	22.13
Feed	100.00	8.27	100.00	100.00

, with the corresponding flowsheet presented in Figure 14. The (K₂O + Na₂O) grade of the concentrate is 11.41%, with K₂O and Na₂O recoveries of >75% relative to the Li flotation tailings. The K₂O and Na₂O recoveries of feldspar concentrate are 73.83% and 74.33%, respectively, relative to the raw ore. The main content of the feldspar tailings is quartz, with an SiO₂ grade of 91.16%, meeting the standards required for Grade II quartz sand in glass manufacturing (JC/T 529-2000 [45]). The SiO₂ recovery of feldspar tailings is 45.72% relative to the raw ore. This demonstrates the potential for full utilization of Li flotation tailings. Future studies will focus on further purifying quartz in feldspar tailings to produce high-quality quartz concentrates.

3.2.3. Comprehensive Recovery of Nb–Ta Minerals and Cassiterite

Section 3.2.2 indicates that ~80% of Nb₂O₅, Ta₂O₅, and Sn are enriched in the Li flotation concentrate. Thus, the comprehensive utilization of these minerals in the Li flotation concentrate is feasible. The main impurity in the Li flotation concentrate is 4.30 wt.% Fe₂O₃, which is commonly removed via low-intensity magnetic separation (LIMS) [46]. Owing to the weak magnetism of Nb–Ta minerals and the nonmagnetism of spodumene and cassiterite, Nb–Ta minerals can be separated from spodumene and cassiterite via high-intensity magnetic separation. The densities of Nb–Ta minerals, cassiterite, and spodumene are 5.00–7.90, 6.80–7.10, and 3.13–3.20, respectively [35]. Based on the density differences among these minerals, gravity separation can be an effective method. Through multiple stages of cleaner, recleaner, and scavenger, Nb–Ta and Sn concentrates were simultaneously obtained while producing a high-grade Li concentrate. The detailed flowsheet and indices of the comprehensive utilization are shown in Figure 15 and

Table 12. Indices of comprehensive recovery of Nb–Ta minerals and cassiterite.

Product	Yield (%)	Grade (%)				Recovery (Relative to the Li Flotation Concentrate, %)				Recovery (Relative to the Raw Ore, %)
		Li2O	Nb2O5	Ta2O5	Sn	Li2O	Nb2O5	Ta2O5	Sn	Li2O	Nb2O5	Ta2O5	Sn
Nb–Ta concentrate	0.09	0.01	32.790	27.410	7.12	0.00	52.52	52.07	3.14	0.00	43.62	43.27	2.45
Li concentrate	97.38	5.89	0.027	0.023	0.06	99.75	46.79	47.28	26.71	86.62	38.86	39.29	20.89
Sn concentrate	0.23	0.30	0.007	0.004	62.17	0.01	0.03	0.02	70.04	0.01	0.02	0.02	54.78
Fe impurities	2.30	0.59	0.016	0.013	0.01	0.24	0.65	0.63	0.11	0.20	0.54	0.52	0.09
Feed (Li flotation concentrate)	100.00	5.75	0.056	0.047	0.20	100.00	100.00	100.00	100.00	-	-	-	-

, respectively. The Nb–Ta concentrate achieves a total Nb₂O₅ + Ta₂O₅ grade of 60.20%, with 52.52% Nb₂O₅ and 52.07% Ta₂O₅ recovered from the Li flotation concentrate. A qualified Sn concentrate with a Sn grade of 62.17% and recovery of 70.04% is obtained, yielding a final Li concentrate with an Li₂O grade of 5.89% and a recovery of 99.75%. Relative to the raw ore, the recoveries are 43.62% Nb₂O₅, 43.27% Ta₂O₅, 54.78% Sn, and 86.62% Li₂O, indicating the high quality of the Li concentrate and the efficient, comprehensive recovery of associated Nb, Ta, and Sn.

3.3. Water Reuse from Tailing Slurry

To investigate the feasibility of water reuse from the slurry of the Li flotation tailings, the influence of returned water on the flotation indices was studied. The returned water, obtained from the supernatant after 1 d of settlement of Li flotation tailings, was subjected to flotation and grinding. The spodumene flotation procedure is identical to that of the laboratory-scale flotation experiment.

3.3.1. Effect of the Ratio of Water Reuse on Spodumene Flotation

Flotation tests were conducted using varying ratios of returned water to assess its effect on the flotation indices of spodumene. Figure 16 shows that a 10% returned water ratio causes a lower change in the indices compared with freshwater. However, increasing the ratio of the returned water from 10% to 40% reduces the Li₂O grade by 0.68% and recovery by 10.96% in the rough concentrate. At a 40% returned water ratio, the Li₂O recovery of the rough concentrate decreases by 11.32% compared with freshwater. Therefore, direct reuse of returned water is not recommended. Treatment is needed to reduce its adverse effect on flotation.

3.3.2. Effect of Different Treatments on Spodumene Flotation

To mitigate the detrimental effect of returned water on spodumene flotation, returned water treated with different flocculants was tested for its impact on flotation indices [47]. Four water treatment reagents—H₂SO₄, lime (CaO), alum (KAl(SO₄)₂·12H₂O), and FeSO₄—were selected to treat the Li flotation tailing slurry at a dosage of 500 g m⁻³. The resulting pH values of the returned water after treatment were 4.95, 11.80, 7.48, and 8.38, respectively. Figure 17 shows that the LiO₂ grade and recovery of the rough concentrate decreased by 0.50% and 19.08%, respectively, when using H₂SO₄-treated returned water compared with freshwater-treated rough concentrate. CaO- and KAl(SO₄)₂·12H₂O-treated returned water samples show Li₂O recoveries of 24.53% and 51.77%, respectively, mainly due to residual calcium and aluminum ions in the wastewater, which hinder the adsorption of flotation reagents on the spodumene surface [40]. FeSO₄-treated returned water exhibits an Li₂O grade of 3.60% and a recovery of 85.55%, close to freshwater performance. Polyferric sulfate has been widely used in wastewater treatment, including woolen wastewater [48] and mine wastewater [49]. The primary mechanism of FeSO₄ as a flocculant is likely the charge neutralization of negatively charged colloids and mineral particles in the returned water by hydrolysis products of Fe ions, accelerating water clarification [50]. Herein, considering that the pH of FeSO₄-treated water is close to that of freshwater and its minimal impact on flotation, FeSO₄ is recommended as the optimal reagent for water reuse.

Subsequently, the effect of FeSO₄ dosage on flotation indices was further studied, as shown in Figure 18. As the FeSO₄ dosage increased from 500 g m⁻³ to 1500 g m⁻³, the pH of the returned water gradually decreased from 8.4 to 7.0, and the recovery of Li₂O in the rough concentrate initially increased and then decreased. At a dosage of 1000 g m⁻³, Li₂O recovery reaches a maximum (88.53%), with an Li₂O grade of 3.59% in the concentrate. Compared to freshwater, Li₂O grade and recovery decreased by 0.38% and 1.10%, respectively, while the pH of the returned water was nearly neutral. Therefore, the optimal FeSO₄ dosage was selected as 1000 g m⁻³.

3.3.3. Locked-Circuit Test

To assess flotation performance after water reuse, a laboratory-scale locked-circuit test was conducted. The process was identical to the freshwater test described in Section Locked-Circuit Test of Spodumene, with minor adjustments to the reagent scheme. The locked-circuit test produced a concentrate with an Li₂O grade of 5.77% and recovery of 86.47% (

Table 13. Results of the locked-circuit test of spodumene flotation with water reuse.

Product	Yield (%)	Li2O Grade (%)	Li2O Recovery (%)
Li flotation concentrate	17.39	5.77	86.47
Li tailings	82.61	0.19	13.53
Feed	100.00	1.16	100.00

). Although these indices of the concentrate in the locked-circuit test with returned water were slightly lower than those in the absence of returned water (Table 4), it is still an encouraging result because of water conservation and low reagent consumption.

Table 14. Reagent consumption of the locked-circuit flotation test of spodumene with and without water reuse.

	Reagent	Na2CO3	NaOH	CaCl2	EMT-12
Reagent Consumption (g/t)
Total with water reuse		2900	650	180	2800
Total without water reuse		3000	800	180	2900
Saving ratio (%)		3.33	18.75	-	3.45

shows that the consumption of NaOH, Na₂CO₃, and EMT-12 is reduced by 18.75%, 3.33%, and 3.45%, respectively. Although FeSO₄ treatment of returned water is economically and technically feasible for industrial production, the impact of settled tailings on feldspar flotation performance still requires further investigation.

3.4. Economic Evaluation and Environmental Perspective

To evaluate the economic benefits of the comprehensive utilization process of the Li pegmatite ore, the economic returns from Li and byproducts, including Nb–Ta, Sn, feldspar, and quartz concentrates, were approximately estimated via comprehensive utilization and summarized in

Table 15. Economic returns of the comprehensive utilization of the Li pegmatite ore.

Product	Yield (Relative to Raw Ore, %)	Price	Economic Returns
Li concentrate	17.43	CNY 7855/t	1369.21
Nb–Ta concentrate	0.016	CNY 199,800/t	32.19
Sn concentrate	0.04	CNY 94,800/t	39.03
Feldspar concentrate	45.81	CNY 200/t	91.62
Quartz concentrate (Feldspar tailings)	36.29	CNY 100/t	36.29
Total	99.59	-	1568.34

Measurement unit (CNY/t): the economic benefit for processing per ton of raw ore.

. Yield data were calculated based on pilot-scale tests, providing great representativeness for production practice. The process provides an economic benefit of CNY 1568.34 per ton of raw ore, with an additional CNY 199.13 per ton from the comprehensive utilization of byproducts. Furthermore, the cost of freshwater is reduced by ~30% using returned water [43], and reagent consumption is also reduced.

Table 16. Economic analysis of the comprehensive utilization process of an Li pegmatite ore.

Item	Expense/Income
Comprehensive utilization of Nb–Ta, Sn, feldspar, and quartz	+199.13
Cost of comprehensive utilization process	−38.00
Reagents reduction	+1.32
Freshwater reduction	+0.50
Total	+162.95

Measurement unit (CNY/t): the economic benefit for processing per ton of raw ore.

presents an approximate cost–benefit analysis of the process, indicating that the comprehensive recovery of multiple concentrates from the Li flotation concentrate and tailings provides benefits of CNY 162.95 per ton. It should be noted that this cost estimate is preliminary, primarily including equipment acquisition, routine wear, and maintenance. For industrial production, precise calculations are necessary. Compared with traditional spodumene flotation, the proposed process offers reduced tailings disposal, full utilization of multiple elements, and lower water and reagent consumption.

4. Conclusions

Herein, laboratory- and pilot-scale tests were conducted to develop a comprehensive utilization process for Li pegmatite ore, effectively combining flotation, magnetic, and gravity separation techniques. Compared with conventional methods, the process produces qualified concentrates of Ni–Ta, Sn, feldspar, and quartz, in addition to Li concentrate, achieving stepwise utilization of the minerals in the ore. Based on the results, the primary findings are summarized below:

• Laboratory-scale locked-circuit flotation tests of spodumene and feldspar yielded an Li concentrate with an Li₂O grade of 5.80% and recovery of 88.62%, and a feldspar concentrate with a (K₂O + Na₂O) grade of 11.41%, with recoveries of K₂O and Na₂O of 81.30% and 84.81%, respectively.
• In a 72 h continuous pilot-scale test, Li flotation yielded a concentrate with an Li₂O grade of 5.72% and recovery of 86.78%, enriching 83.05% Nb₂O₅, 82.72% Ta₂O₅, and 78.11% Sn. Using Li flotation tailings as feed, a feldspar concentrate with a (K₂O + Na₂O) grade of 11.41% was obtained, achieving K₂O and Na₂O recoveries of >75%.
• Comprehensive utilization of Li flotation concentrate via a combined magnetic and gravity separation process produced an Nb–Ta concentrate with a total Nb₂O₅ + Ta₂O₅ grade of 60.20% and a Sn concentrate with a Sn grade of 62.17%. Recoveries relative to the raw ore were 43.62% Nb₂O₅, 43.27% Ta₂O₅, and 54.78% Sn, demonstrating efficient recovery of multiple valuable elements. Simultaneously, the main content of feldspar tailings was quartz, with an SiO₂ grade of 91.16%, satisfying the standards required for Grade II quartz sand in glass manufacturing.
• The feasibility of water reuse from the slurry of Li flotation tailings was initially researched. FeSO₄ with a dosage of 1000 g m⁻³ was identified as the optimum flocculant, with flotation indices of the Li concentrate (Li₂O grade of 5.77% and recovery of 86.47%) slightly lower than those using freshwater. Consumption of NaOH, Na₂CO₃, and EMT-12 was reduced by 18.75%, 3.33%, and 3.45%, respectively.
• The proposed process provides economic benefits and offers remarkable advantages, including marked reduction in tailings disposal, full utilization of multiple elements, and low water and reagent consumption. This study provides valuable guidelines for the mineral processing of an Li pegmatite ore and other associated multimetallic ores.