Recovery of Tungsten from Raw and Secondary Materials Using Hydrometallurgical Processing

Abstract

As in the case with other metals, tungsten is an element with a number of uses in different fields, which is why its recovery from both primary and secondary materials continues to be of great interest. Various hydrometallurgical processes, considered as unit operations, can be used for the recovery, separation and concentration of tungsten from any source, with ease of scaling-up a potential factor when considering the best process for practical use. The present work reviewed investigations into the use of such unit operations for the recovery of tungsten which were published during 2024 and the first half of 2025. Because most if not all of these investigations were conducted on a laboratory scale, there is still much room for improvement before deciding on the best option for tungsten recovery. In all cases, however, this recovery is based on a series of steps from leaching to separation technologies (ion exchange resins, liquid–liquid extraction, etc.) to the tungsten end-product.

1. Introduction

The European Union has listed tungsten in its inventory of critical raw materials continuously since 2010, with China being the largest producer of this metal during this period [1]. In the year 2024 the average price of tungsten was 250 USD/ton, much lower than that of other materials also considered as critical and strategic (lithium carbonate 14,000 USD/ton, dysprosium oxide 420 USD/kg, neodymium oxide 56,500 USD/ton, gold 2350 USD/troy ounce, aluminum 2475 USD/ton) [2].

However, according to data presented by the MSM (Mines Material Systems Model) [3], the future for tungsten seems to be not as impressive as that for other materials. While in 2021 tungsten ranked 30th (out of 33 materials) in terms of tonnage produced, and 28th in terms of market size, this latter rank falls to 30th in a projection for the year 2050.

Tungsten is widely used in mining, machining, electrical, aerospace, defense, nuclear and chemical industries. Typical applications in these fields include tunnel-boring machines, cutting tools, electric-lamp filaments, aircraft engines, bullets, radiation shielding, and catalysts.

While production of this metal continues to rely on the processing of raw materials (scheelite, wolframite, etc.), new methods for the recovery of tungsten from secondary resources are gaining in importance. These secondary resources include tungsten-bearing cement carbide, alloys and scrap. With a −4.8% change in quantity demanded for the 2021–2050 time period [3], it is estimated that the amount obtained from recycling will increase while the amount recovered from raw materials will decline.

Broadly speaking, tungsten recovery from any source is carried out by pyro- or hydrometallurgical procedures, and sometimes by a combination of both. Researchers have reviewed flow-sheets for a series of tungsten extractive-metallurgy processes [4] and for treatment of W-bearing wastewaters [5], with more specific investigations including the treatment of wolframite ore from the Akchatau deposit [6]. This treatment involves sintering of the concentrate with soda, followed by a leaching step and treatment of the leachate with a cation-exchange membrane. In [7], a combination of pyro- and hydrometallurgical processing of spent selective catalytic reduction catalysts was conducted to recover tungsten, vanadium and titanium. In this process, waste material was first roasted in sulfuric acid medium. The next step involved water-leaching of the roasted material. The leachate was subjected to a liquid–liquid extraction procedure using the amine N1923 as extractant. While vanadium was not extracted, both titanium and tungsten were loaded onto the organic phase. Finally, and most importantly, titanium was stripped with sulfuric acid and H₂O₂, and at a later stage tungsten was stripped from the organic phase in alkaline medium.

The authors of [8] described a process based on the pyrometallurgical treatment of wolframite which yielded sodium tungstate as the final product. Utilizing tungsten alloy waste as a tungsten source, the authors of [9] investigated a pyrometallurgical procedure to separate the metal from alkali molten slag. In [10], this metal was recovered from tungsten-filled vinyl-methyl-silicone-based flexible shielding materials by pyrolysis in an argon atmosphere, and further treatment of the pyrolyzed samples was carried out using the ultrasonic cleaning procedure.

Bioleaching of ores has been widely investigated, and its application in tungsten recovery from semiconductor waste was described in [11]. In addition, a tungsten-rich solution was treated via activated carbon adsorption–desorption or ammonium paratungstate precipitation. A preliminary economic study indicated that precipitation required 7% more capital investment, though the bioleaching step was responsible for most of the operational costs.

Due to their high heavy-metal content, waste tungsten mine tailings were the subject of an investigation to recover tungsten, lead and cadmium [12]. The waste was treated using two leaching steps. In the first—neutral—step, tungsten is recovered; in the second, use of acidic medium allows the dissolution of lead and cadmium. As a variation of the above, phytoremediation of tungsten tailings has also been conducted [13,14].

In [15], both wolframite and scheelite tailings were used to fabricate ceramic foams for industrial applications via a pyrometallurgical procedure. Also in the pyrometallurgical world, hazardous tungsten leaching residues were subjected to treatment with photovoltaic silicon kerf waste to enable recovery of a cobalt-rich alloy, on the one hand, and, on the other, of the metals (tungsten, nickel, niobium and tantalum) contained in the starting material [16,17].

In [18], tailings from tungsten mining waste at the abandoned Wolfram Camp mine in northern Queensland, Australia were treated via gravity separation and bioleaching to allow recovery of this metal.

The authors of [19] also investigated the use of secondary wastes, in this case, W-Mo alloy scrap, as starting material in a hydrothermal process to produce W-Mo bimetallic sulfides (Mo_1−xW_xS₂ nanoflowers).

The use of organic acids in the leaching of metals is increasingly being used for the recovery of metals. Tungsten is no exception to this, which is why a process using lactic acid to leach oxidized powders composed of CoWO₄ and WO₃ was described in [20].

In [21], a hybrid membrane electrolysis process coupled with nanofiltration was used in the treatment of welding electrodes to recover tungsten and thorium. The best conditions for the electrolysis and nanofiltration steps involved use of solutions of pH 6–9. In the first nanofiltration step, tungsten was separated in the permeate. Next, a further shift in the pH value of the concentrate to 13 allowed the separation of aluminum (permeate) and thorium dioxide.

Because coal ash produced during electricity generation contains valuable metals (germanium and tungsten), its recycling is of interest. The authors of [22] described how such recycling may be made affordable by use of a sequential vacuum distillation method with disulfide. Hazardous arsenic is first volatilized at temperatures below 550 °C, and germanium and tungsten are volatilized in the form of sulfides at 1050 °C.

Simulations of the adsorption of tungsten onto goethite using a first-principles molecular dynamics methodology were carried out in [23]. However, it should be noted that, regardless of the pH value of the solution, polymeric tungsten species were not considered, only tungsten-mononuclear species.

Since tungsten was first identified as an emerging contaminant in the 2000s, major progress has been made regarding awareness of its toxicity; however, our knowledge of its impact on human health remains somewhat limited [24].

A more comprehensive research of the literature indicated that in the recovery of tungsten from any of these sources, hydrometallurgy plays an important role due to the diversity of the unit operations involved in the use of this branch of extractive metallurgy. Thus, and in contrast to recent publications [4,5] which reviewed the state of the art up to and including 2023, the present work reviewed and focused on the most recent reported advances (2024–mid-2025) in the recovery of tungsten from primary and secondary materials using the various unit operations involved in the hydrometallurgical processing of these sources.

2. Unit Operations in Hydrometallurgy

As is well known, hydrometallurgy involves the use of aqueous media and low and/or moderate temperatures. This type of processing consists of unit operations that allow sequential treatment of a starting material, either an ore or a secondary residue, until the metal or a product of that metal that is of practical interest, either to be sold or to be recycled, is obtained. Sometimes, prior to hydrometallurgical processing, the starting material may require some pre-treatment (concentration, thermal treatment, etc.). Once the starting material is prepared for treatment, the first unit operation is the dissolution of the metal of interest, or the leaching process. Sometimes, this first operation is also used to remove impurities, keeping the metal of interest in the solid residue. Then, after a filtration stage, the solution containing the metal of interest is subjected to a series of separation and/or concentration operations to obtain a product, either solid or liquid, which allows the final pure metal or metallic product of interest to be obtained. The advantage of all these unit operations is that they can be scaled up according to the needs of the process, and with little or no modification they can be used for the treatment of both raw and secondary materials.

In the case of tungsten specifically, the most-used unit operation is leaching, followed by ion exchange with resins, adsorption, liquid–liquid extraction (solvent extraction), precipitation, and evaporation/crystallization.

2.1. Unit Operation: Leaching

As stated above, after the preparation or pretreatment of the metal-bearing material, the next stage is the dissolution of the constituents of the starting material; thus, this leaching operation is a solid–liquid mass transfer process. The conditions for this leaching step vary in terms of temperature, pressure, type of leachant, etc. These parameters need to be fixed in line with the requirements of the given process. The final objective of this step is to produce metal ions or metal complexes in solution in the most selective form, or to enable their further selective recovery from the solution. In the case of tungsten, the leaching step is normally conducted in acidic or alkaline media, or even in water. This section describes the most recent advances in the dissolution of tungsten contained in different types of such raw or secondary materials.

The authors of [25] carried out a series of investigations into the processing of Akchatau ores, with the aim of improving the treatment of wolframite concentrates. Concentrates (60–62% WO₃) was first reacted with sodium carbonate at 520–550 °C. Temperature was then increased further, to 750–850 °C. This sintering process was conducted to yield sodium tungstate (Na₂WO₄), which was leached at temperatures around 95 °C with water (98.6% leaching effectiveness). Purification of the leachates was carried out via electrodialysis in a two-chamber cell with separation of anode and cathode spaces by a cation exchange membrane, the end product being marketable Na₂WO₄.

Using a secondary source as hydrocracking catalyst (102-HC W/Ni/Al₂O₃, SiO₂) a process to recover tungsten was developed by the authors of [26]. Firstly, material with a pulp density of 10 was roasted at 798 °C for two hours. Next, the roasted material was leached in alkaline medium (sodium hydroxide) at 70 °C for one hour, resulting in a tungsten recovery rate of 93%. The investigation did not present data about the further processing of this tungsten-bearing solution.

Another secondary material, specifically, spent tungsten-containing catalysts of organic production, formed the object of another investigation [27]. In this case, the leaching process consisted of two stages. In the first stage, the starting material was leached with water. In the second stage, the material from the first operation was leached with sodium carbonate. During the first step, cesium alum was precipitated. Calcium tungstate (KVGF grade) was the product of the second leaching step. This calcium salt was treated with hydrochloric acid to yield tungstic acid (WO₃ content exceeding 80%).

A procedure which aimed to recover tungsten from scheelite (Chinese origin) using a mixture of H₃AsO₄ and H₂SO₄ solution was developed in [28]. Under leaching conditions of 70 °C and eight hours, arsenotungstic acid (H₃AsW₁₂O₄₀) was formed with a near-quantitative (99.9%) tungsten leaching rate. This acid was purified by application of a solvent-extraction operation, with sec-octanol found to be the extractant presenting the best extraction rates (

Table 1. Extraction of arsenotungstic acid using various extractants.

Extractant	%W Extraction
Sec-octanol	50
Dodecanol	55
Hexadecanol	47
Octadecanol	42
Methyl benzoate	35
Benzophenine	25
Diphenyl ether	1

Temperature: 65 °C. O/A: 1. Time: two hours. Adapted from [28].

); however, due to the volatility associated with this chemical, dodecanol was selected as the most promising extractant. Tungsten was stripped from the loaded organic phase by water at 70 °C, with 30 min of contact time and an O/A ratio of 1. After five consecutive stages, the accumulated stripping rate was 99.9%. The purified arsenotungstic acid solution was precipitated in ammoniacal medium to yield solid (NH₃)₃AsW₁₂O₄₀·xH₂O. This chemical was then subjected to a reductive step using hydrogen to yield tungsten powder, while arsenic was volatilized and condensed. The most curious aspect of this work is the use as a leachant of an arsenic product which is considered toxic by many standards.

In [29], scheelite tailings (Canadian origin) were subjected to a leaching step in carbonate solutions (2.5 M) at temperatures in the 25–75 °C range [29]. The best dissolution results (near-99% efficiency) were obtained with a temperature of 75 °C and 15–20 min of reaction time. Further processing of the Na₂WO₄ solution was not described in the published manuscript.

In [30], in order to recover tungsten, diamond core drilling crowns (Serbian origin) were recycled by aqua regia leaching and further alkali leaching [30]. After the first step, cobalt, iron and nickel all dissolved, whereas tungstic acid and diamonds remained undissolved. This solid was leached (about 98% efficiency) in ammoniacal medium (20% ammonium hydroxide, three hours at 89 °C, S/L ratio of 1/10), resulting in the dissolution of tungsten as ammonium tungstate while diamonds remained undissolved. Once these diamonds were filtered, the tungsten-bearing solution was evaporated, and ammonium paratungstate ((NH₄)₁₀H₂W₁₂O₄₂·xH₂O) crystallized. This salt was used as a precursor to yield metallic tungsten. Processing of the solution obtained from the first leaching step yielded cobalt powder.

In the next reviewed study, the authors investigated the treatment of wolframite (Chinese origin) in sulfuric–phosphoric acid mixtures under pressure conditions [31].

Table 2. Effect of temperature on wolframite leaching.

Temperature, °C	% Efficiency
150	95
140	90
130	85
120	75
110	65

Time: two hours. Particle size: 58–75 µm. Leachant: 2.5 M sulfuric acid +2 M phosphoric acid. Adapted from [31].

shows that an increase in the temperature increased the percentage of the leaching efficiency. It was proposed that tungsten was dissolved as an H₃PW₁₂O₄₀ species; however, no further data was provided about the processing of the as-obtained solutions.

Several issues relating to the leaching of scheelite concentrates with sulfur–phosphorus mixed acid have led to the development of a new procedure involving the inhibition effects of a calcium sulfate-blocking membrane, the selective extraction of tungsten by ion exchange, and the removal of phosphorus by the magnesium ammonium salt procedure [32]. Scheelite concentrate (Chinese origin) was leached in a H₂SO₄-H₃PO₄ mixture at 80 °C and six hours of reaction, with addition of calcium sulfate improving the crystallization environment of gypsum dihydrate. The leachate containing about 90 g/L WO₃ was treated by ion exchange resin technology using D301 resin (weak base resin). In the process, PW₁₂O₄₀³⁻ groups in the leachate were interchanged with HSO₄⁻ from the resin. Once tungsten was loaded onto the resin, the desorption step consisted of its reaction with 4–6 M ammonia at 60 °C. After desorption, PW₁₂O₄₀³⁻ was depolymerized by ammonia into phosphate ions and WO₄²⁻, and ammonium paratungstate was finally crystallized after evaporation of the solution. Phosphorus was removed in the form of magnesium ammonium phosphate after a magnesium salt was used.

Because armor-piercing shells utilize W-Ni-Fe alloys, the recycling of scrap material from these shells is of interest. The authors of the next reviewed investigation dissolved such scrap in a mixture of hydrochloric acid and hydrogen peroxide [33]. The leaching of the scrap was found to be largely dependent on a series of variables, with the best operational leaching parameters being determined as follows: 2 mol/L HCl, 2 mol/L H₂O₂, temperature of −5 °C, one hour of reaction time. The process allowed tungsten powder to be yielded.

In the next reviewed work [34], tungsten was recovered from scheelite ores (Spanish origin) via microwave-assisted alkaline fusion. The process converted scheelite into tungstate salts ([Na₂WO₄, K₂WO₄ and/or K₃Na(WO₄)₂). The use of a low-melting eutectic alkali (NaOH/KOH) system decreased the reaction temperature to 150–200 °C, while the use of microwave technology reduced reaction times to 10–30 min. Leaching of the fused material was carried out with water at temperatures of 60–80 °C and 30 min. The investigation did not mention the subsequent processing of the dissolution. This type of process can be used for high-low grade ores and tailings.

Using synthetic scheelite, a dissolution process using the mixture H₂SO₄/H₂O₂ was developed to recover tungsten in [35]. The leaching conditions were as follows: 1 M H₂SO₄, 1.5 M H₂O₂, 60 °C, two hours of reaction time. The dissolution process yielded gypsum and H₄W₄O₁₂(O₂)₂ in solution. Then, after a filtration step, the solution was thermally decomposed (90 °C and four hours), yielding tungsten in the form of tungstic acid.

In [36], de-nitration processing was carried out using a catalyst of the V₂O₅-WO₃/TiO₂ type so that the spent catalyst could then be used as tungsten source [36]. In this investigation, the spent material was leached in Na₂S medium to yield species containing WO_4-XS_X²⁻ groups (X = 1 or 2). Because these groups cannot be used to produce calcium tungstate, an oxidizing step (30 °C) was necessary to transform these compounds to WO₄²⁻ containing species. This oxidation was performed using ozone (flow rate 2 g/L). After this step, impurities (titanium and silicon) present in the solution were precipitated by shifting the pH of the solution from 10.2 to 9 using HCl. The other experimental conditions for this step were 95 °C and four hours of reaction time. Finally, calcium (CaCl₂) precipitation was used to recover tungsten from the solution. In this operation, the pH of the solution was shifted again to pH 10.9, while a temperature of 85 °C and a 30 min reaction time were used to precipitate calcium tungstate with a near-95% efficiency. The chemical composition of this calcium tungstate fulfilled the GB5192-85 requirements for impurity content of II-class synthetic scheelite.

The recovery process for tungsten–rhenium alloy waste generates calcium tungstate containing rhenium. In light of this, a process to recover these two valuable metals was investigated in [37]. Separation of both metals (dissolution of rhenium and precipitation of tungsten) was carried out under the following conditions: 2.7 M HCl, liquid–solid ratio of 26:1, temperature of 80 °C, processing time of 5 h. Under these conditions, the rhenium leaching efficiency was about 99%, with negligible tungsten dissolution. While the Re-bearing solution was treated by ion exchange resin processing, the precipitated tungstic acid was calcined at 900 °C for 0.5–5 h to produce WO₃. Longer reaction times produced larger oxide particles, with particle size increasing from 195.53 nm (0.5 h) to 540.10 nm (5 h).

Tungsten fine mud is secondary waste which contains quartz and potassium feldspar as well as a small amount of tungsten. In [38], such waste was treated in a caustic soda environment in order to recover tungsten contained within it [38]. Experimentation showed that with a temperature of 160 °C, a caustic soda concentration of 100 g/L, a treatment time of 2 h, and a liquid–solid ratio of 2.5:1 mL/g, the leaching efficiency of WO₃ was about 95%. The caustic concentration in the leaching solution needed to be fixed at 100 g/L. Continuous cycling of the caustic leaching solution allowed the tungsten concentration to increase from 15.9 to almost 120 g/L. Further processing of the leach solution was not reported in the work.

In [39], the previous use of NaHSO₄·H₂O in scheelite processing was extended to the treatment of wolframite. Wolframite was roasted in the presence of this salt. Rates of conversions of the mineral into WO₃ were found to approach 99% under following conditions: temperature of 650 °C, time of two hours, and use of three times the stoichiometric amount of the salt. The leaching of the roasted product to remove manganese and iron impurities resulted in a concentrate with 93% WO₃. The solid was leached with ammonia (4 mol/L) at 80 °C for two hours. The leachate was then evaporated and crystallized to yield ammonium paratungstate.

Because heap leaching has low operational and capital costs, it was used in [40] for processing of a scheelite ore utilizing a mixture of hydrochloric and oxalic acid as leachant. Scheelite was placed on the heap in the form of pellets. The presence of hydrochloric acid helped to leach the ore, reducing oxalic acid consumption during the further leaching of tungsten.

Though it is difficult to compare leaching results, due to variations in the tungsten-bearing materials treated and the experimental conditions used in the leaching experiments, a summary of some of the results derived from this review may be offered (

Table 3. Summary of leaching operations.

W-Material	Thermal Pretreatment	% Leaching	Ref.
Wolframite	Yes	99	[25]
Wolframite	No	65–95	[31]
Wolframite	Yes	98	[39]
Scheelite	No	99	[28]
Scheelite tailings	No	90	[29]
Scheelite	No	96	[32]
Scheelite	Yes	77–97	[34]
Scheelite (synthetic)	No	95	[35]
Spent catalyst	Yes	93	[26]
Spent catalyst	No	82	[36]
Diamond core residue	No	98	[30]
CaWO4 residue	No	W precipitates	[37]
W-fine mud	No	95	[38]

). It can be seen from the table that, with some exceptions, the efficiency of the leaching process is higher than 95%, with efficiency sometimes being dependent on the temperature [31] or the type of material [34]. In the case of [37], tungsten precipitated and was recovered from residue.

There is little to criticize in those investigations in which the leaching step was applied for the recovery of tungsten from tungsten-bearing materials. Probably the only comment required is that processes were only investigated at the laboratory scale. Implementation on a larger scale is required is the advantages or disadvantages of each process are to be fully appreciated. The results of the reviewed studies show little room for improvement, with all the proposals following the same structure of leaching in acidic or alkaline media followed by a precipitation/crystallization step. Only in the case of selected tungsten-bearing materials, and prior to the leaching step, was a thermal pretreatment proposed in order to facilitate this leaching step. With respect to the end products obtained using these treatments, there are also few if any surprises, and little space for improvements, with tungstic acid, APT or tungsten salts being the most common chemicals obtained from the various processes, and only a couple of investigations managing to obtain W-metal powder as a final product. It is also true that, along with the other tungsten products, they pave the way to obtain this metal powder. We may say that what is really new in all this research is the application of known procedures to “new” tungsten-containing materials.

2.2. Separation–Concentration Operations

Because the targeted metal is often accompanied in the solution by undesirable metals or impurities, a series of operations are needed to separate them. Also very commonly, these operations serve to concentrate the metal, improving and facilitating the final stage in the recovery of the metal or any derivative thereof.

2.2.1. Unit Operation: Ion Exchange Resins

Ion exchange resins are synthetic organic polymers which have functional groups responsible for exchanges with the metals or metal complexes from aqueous media. These are thus liquid–solid mass transfer operations. Because tungsten forms anionic compounds in solution, the so-called anionic exchangers, containing quaternary ammonium groups (strong base) or amine groups (weak base), are the type of resins suitable for use in the recovery of this metal. After the metal is loaded onto the resin, a second step, or elution step, is necessary to transfer the metal from the resin to another aqueous phase or eluate, in which the metal is found in a highly concentrated form.

Researchers have used anionic resin to extract tungsten from solutions, applying variations in pH values (3–9) and other experimental conditions [41]. In the latter work, experimental results showed that metal upload onto the resin was independent of temperature in the 6–20 °C range, but increased when the pH of the solution was shifted to lower values. No evidence of the tungsten elution process was presented.

Using synthetic solutions, the authors of [42] investigated the removal of tungsten from aqueous solutions by the use of D318 weakly basic macroporous resin. Maximum metal removal was achieved at pH 4, with lower levels of removal at both lower and higher pH values. An increase in temperature led to an increase in metal loading onto the resin (

Table 4. Influence of temperature on W loading onto D-318 resin.

Temperature, °C	[W], mg/g
25	660
35	675
45	685

Aqueous phase: 10.8 g/L WO₃ at pH 4. Time: three hours. Adapted from [42].

). Dynamic experiments showed that, at pH 4, when the concentration of WO₃ increased from 2 to 10 g/L, the overall loading capacity decreased. The same negative effect was observed when the solution flow rate increased from 15 to 55 mL/min. Dynamic elution experiments using 85 g/L NaOH solutions showed a decrease in the effectiveness of the operation when the flow rate of the eluant increased from 5 to 20 mL/min.

In [43], the removal of isopolytungstate species from aqueous solutions was investigated using different anion exchange resins (201x7, M20, D290, D301, D314). It was found that D301 resin yielded the best removal results (D301: 90%, D290: 80%, D314: 60%, M20: 30%, 201x7: 20%) with 2 g resin dosage; consequently, D301 resin was used in subsequent investigations. Maximum removal efficiency was obtained with the use of solutions of pH 3–4, with decreased efficiency at higher pH values. This is attributable to the predominancy of the different isopolytungstate stoichiometries in solution. The following removal order was established: H₂W₁₂O₄₀⁶⁻ > H₂W₁₂O₄₂¹⁰⁻. The six-charge species was predominant in the 3–4 pH range and the ten-charge tungsten species was predominant in the 5.5–6.5 pH range. Optimum metal elution was obtained using NaOH solutions (4–6 M) and a temperature of 50 °C.

Using synthetic solutions, the authors of [44] investigated the selective separation of molybdenum (Mo(VI)) from tungsten (W(VI)) solutions using 201x7 resin. It was experimentally observed that the acid used to control solution pH had a decisive influence on the values of the Mo/W separation factor. These values followed the following order: tungstic acid>hydrochloric acid>sulfuric acid. A maximum Mo/W separation factor value was obtained when the pH of the solution reached 7; thus, under this condition, MO₄²⁻ species was loaded onto the resin in preference to H₂W₁₂O₄₂¹⁰⁻ species. Column experimentation confirmed the results derived from batch experiments.

2.2.2. Unit Operation: Adsorption

From the point of view of the authors of the present paper, the term adsorption describes a liquid–solid mass transfer process in which metal, in whatever form, is loaded onto a solid matrix but not onto a synthetic resin. Under the adsorption process, therefore, ion exchange and also other physicochemical processes can occur. Very often, the adsorption of a metallic species may be attributed to one or more of these processes. Typical adsorbents include activated carbons, zeolites, metal–organic frameworks (MOFs), etc.

In [45], synthetic Na₂WO₄ solutions were used to investigate the performance of La-doped MgFe-LDH type adsorbent in the recovery of tungsten from aqueous phases. In terms of tungsten loading capacity, the use of La-doped adsorbent presented an advantage over the use of La-undoped material (

Table 5. W loading (mg/g) using MgFe-LDH-type adsorbents.

Adsorbent	pH 6.2	pH 13
MgFe-LDH	41	4
La-MgFe-LDH	56	11

Time: two hours. Adapted from [45].

), whereas the maximum tungsten capacity decreased with increases in pH value to 6, 12 and 13 (59.3, 33.9 and 16.6 mg/g, respectively). Solutions of 1 M NaOH are suitable for tungsten desorption from the metal-loaded adsorbent.

In [46], Zr-based UiO-66 adsorbent was fabricated to investigate its performance in selective separation of Mo(VI) from W(VI) in aqueous media. Maximum Mo/W separation factors were obtained at pH 2, temperature 25 °C, and three hours of reaction time, with metal uptakes of 219.4 mg/g and 11.6 mg/g obtained for molybdenum and tungsten, respectively. NaOH medium was suitable for the desorption step.

An adsorbent (UiO-66-CTAB, CTAB: cetyl trimethyl ammonium bromide) which was similar to that in the above-referenced study but slightly modified was used in [47] for the separation of Mo(VI) and W(VI). The maximum Mo/W separation factor (51.4) was obtained under the very same experimental conditions as those in the above-referenced work, but in this case metal loadings were found to be 355 mg/g (Mo) and 20 mg/g (W). This investigation did not include desorption data.

In [48]. a composite adsorbent (H_xPO₄@Fe₃O₄) was fabricated by loading phosphate ions onto the surface of magnetic Fe₃O₄ nanoparticles via β-particle irradiation, and used for the capture of tungsten from aqueous solutions. The presence of the phosphate ions in the adsorbent greatly increased the adsorption capacity of the adsorbent in comparison with the loading values obtained using pristine Fe₃O₄ nanoparticles (70 mg/g versus 20 mg/g at pH 2). Using the phosphate-modified nanoparticles, tungsten adsorption increased at pH values lower than 3 and when temperature increased from 25 °C to 48 °C. Desorption efficiency increased with an increase in the NaOH concentration from 0.05 to 2 M. The use of this adsorbent (also pristine Fe₃O₄) under continuous adsorption–desorption cycles produced a continuous decrease in the adsorbent loading capacity from about 90% in the first cycle to 70% in the fifth cycle.

Also in the field of Mo(VI)/W(VI) separation, in this case, tungsten loaded selectively over molybdenum, the authors of [49] impregnated D301 resin with chitosan to create CS-D301 adsorbent. A maximum W/Mo separation factor (26.45) was obtained from solutions of pH 6 with metal loadings of about 484 mg/g (W) and 13.8 mg/g (Mo). As desorbent, 0.5 M NaOH solution was used in the study. It was worth mentioning here that the maximum tungsten loading capacity of chitosan (638.53 mg/g) was found to be greater than that of the composite CS-D301 (483.65 mg/g), though the W/Mo separation factor value was greater in the case of the composite than in chitosan (26.45 versus 21.34).

In [50], coating of graphite plate substrate with a nickel hydroxide/carbon black/polyvinylidene fluoride (Ni(OH)₂/CB/PVDF) film was found to promote fabrication of an adsorbent used in the separation of tungsten from aqueous media. Maximum adsorption was reached at pH 3 after 12 h, with metal loading decreasing when the solution pH shifted from 3 to 13. However, considering the dissolution of Ni(OH)₂ from the adsorbent at acidic pH values, it was stated that pH 7 was best for the removal of tungsten from the solution. Desorption was carried out with 0.1 M NaOH solutions.

It is understandable that the results obtained from these two operations are not comparable due to the different conditions employed in each investigation. Nevertheless, an attempt at making a semi-quantitative comparison is presented in

Table 6. Maximum W loadings onto ion exchange resins (IXRs) and adsorbents.

Operation	W In Feed Solution	pH Feed Solution	a W Loading, mg/g	Elution	Ref.
IXRs	Unknown	3–4	347	No	[41]
IXRs	9 g/L	4	660	NaOH	[42]
IXRs	15 g/L	4	unknown	NaOH	[43]
Adsorption	100 mg/L	6	59	NaOH	[45]
Adsorption	74 mg/L	2	88	NaOH	[48]
Adsorption	400 mg/L	3	unknown	NaOH	[50]

^a Based on Langmuir model.

, which summarizes the most noticeable results obtained from this review.

As these two unit operations (ion exchange resins and adsorption) are widely used in industry, their design and use in the recovery of tungsten from aqueous media are perfectly compatible with their intended purpose, i.e., to purify and concentrate, if possible, solutions containing this metal.

The authors of these ion exchange/adsorption investigations seek to make convincing cases that the materials used in their respective works are better than any others used for the removal of tungsten from solutions. Because most of the investigations were carried out on synthetic solutions, the potential of all these products remains under question until they can be tested on real leaching solutions, and the same can be said about the scaling-up potential of some of these proposals. One technical advantage of these processes is that they can be used in the treatment of unclarified solutions. On the other hand, no innovation has yet been proposed with respect to the elution or desorption stage, because in all the above-mentioned proposals, an alkaline medium is used to yield a final solution of sodium tungstate and, as stated above, this salt opens the way to obtain either APT and/or tungsten-metal powder. According to the results presented in Table 6, ion exchange resins have the advantage over adsorptive materials in yielding higher maximum tungsten loadings. Another negative aspect or disadvantage of many of the new adsorbents which have been proposed not only for tungsten but also for other metals is their degradation after continuous use, which in many cases is only a very few adsorption–desorption cycles.

2.2.3. Unit Operation: Liquid–Liquid Extraction

The liquid–liquid extraction operation is based on use of an organic extraction agent which is not soluble in aqueous media and in which the metals of interest have a favorable partitioning with respect to the aqueous medium. Normally the extraction agent is diluted in an organic solvent, which in principle is inert with respect to metal extraction. Sometimes it is also necessary to use a modifier (also an organic chemical) to help in phase disengagement and/or avoid third phase (or second organic phase) formation, this being a phenomenon to be avoided at all costs [51]. In a second step, or stripping, the metal is transferred from the organic phase to a new aqueous phase or strippant.

A leaching solution obtained from treatment of smelting dephosphorization slag (Chinese origin) was used to recover tungsten, molybdenum and phosphorus using liquid–liquid extraction technology [52]. The primary amine (RNH₂) N1923 was used as extractant. This amine was diluted in sulfonated kerosene, and a modifier (isooctanol) was also added to the mixture. Thus, the amine, isooctanol and kerosene formed the organic phase used in the investigation. Experimentation showed that the extraction of the three elements was pH-dependent; at an equilibrium pH (pH value measured after the extraction equilibrium was attained) of about 6.7 the extraction order was P > Mo = WO₃. Shifting the pH value to 7.4, the extraction order was found to be WO₃ > P > Mo, while a further shift to an equilibrium pH value of 8.1 resulted in the following order: WO₃ (80%) > P = Mo (10% each). Stripping was accomplished using 2 M NaOH solutions. This work has a number of inconsistencies. Probably the most important of these concerns the stoichiometry of the tungsten-extracted species (RNH₂)(H₂WO₄). Simply, at the equilibrium pH value of 7.7, the H₂WO₄ species does not exist. It is also rare that the amine acts as a solvation extractant, that is, the metal is extracted as a neutral species. Amines are basic extractants and normally they extract metals by an anion exchange mechanism. Finally, when one considers the W/Mo separation factor value an important discrepancy is observed: in Section 3.1 of the original paper, this is given as 57.08; however, in the Conclusion section of the same publication the value given is 80.34.

The authors of [53] described extraction of Mo from synthetic W-bearing solutions using a TBP-TRPO extractant mixture diluted in sulfonated kerosene (TBP = tributyl phosphate, TRPO = trialkyl phosphine oxide). They found that with a Mo/W separation factor of 400 and an initial pH value of 2, molybdenum could be stripped from the loaded organic phase with a 0.3 M sodium bicarbonate solution.

In [54], the separation of vanadium (V), tungsten (VI) and arsenic (V) from alkaline leachate of spent SCR catalysts was investigated using Aliquat 336 in SO₄²⁻ cycle (MA336-S) dissolved in toluene and an unmentioned modifier. Aliquat 336 extractant is a quaternary ammonium salt in chloride form (R₃CH₃N⁺Cl⁻, R = octyl chain). The extraction of these metals was found to be pH-dependent (

Table 7. Metal extraction percentages achieved using Aliquat 336 (sulfate cycle).

Metal	pH 8	pH 10	pH 12
Vanadium	98	52	30
Tungsten	40	42	45
Arsenic	<5	5	<5

Feed phase: 7.1 g/L V, 7.2 g/L W and 4.6 g/L As, at for the above pH values. Organic phase: 10 wt% Aliquiat 336 and 10 wt% unidentified modifier in toluene. Temperature: 25 °C. Time: two minutes. O/A ratio: 1. Adapted from [54].

); thus, for subsequent investigations the initial pH value of 9 for the aqueous phase was fixed. It was found that vanadium and tungsten were loaded onto the organic phase as HVO₄²⁻ and WO₄²⁻ species via a mechanism of anion exchange with the sulfate group from the extractant. Despite the results shown in Table 7, operational experiments carried out at O/A 2.1 indicated that near-13% arsenic was co-extracted with vanadium and tungsten. The stripping sequence was therefore carried out as follows: The V-W-As-loaded organic phase was first stripped with 0.5 M sodium sulfate to remove arsenic. Next, the V-W-loaded organic phase was stripped with 2.5 M NaOH to strip vanadium. Finally, the resulting organic phase was treated with a mixed 1 M sodium sulfate + 1 M sodium hydroxide solution to strip tungsten. Because the modifier used in the experiments was not known (or not mentioned), reproduction of the results would seem to be a difficult task.

As we have repeatedly mentioned, because liquid–liquid extraction is a separation operation which is widely used industrially, it deserves little criticism. In general terms, we may say that the major disadvantage of liquid–liquid extraction compared with the other unit operations described in this review is the apparent high cost associated with the extractants used in the process. The concerns that some authors have raised about these agents with regard to their toxicity have been disregarded by other authors reporting the use of ionic liquids and deep eutectic solvents as extractants (both of which are labelled as “green solvents” and avoid the use of organic diluents). The use of this technology should avoid the use of unclarified solutions (thus avoiding the formation of sludge, which is associated with entrainment and loss of the organic phase). As stated above, the formation of a third phase is completely undesirable. This phase can be eliminated by using modifiers and/or aromatic diluents. It is the opinion of the authors of this review that many of these negative considerations exist only on paper, because the industrial use of liquid–liquid extraction technology indicates the opposite consideration. As clear advantages over other unit operations, we might mention the short processing times of the aqueous solutions, and the extreme degree of selectivity, i.e., purity, obtained after the extraction-stripping process.

With respect to this review, then, we highlight our major criticism of the results presented in [52] for the reasons given in the text corresponding to this reference. Finally, we may say again that further scaling up of the various extraction systems reported is needed to gain a better knowledge of the potential of each one of these systems.

2.2.4. Unit Operation: Precipitation

The term precipitation refers to a process in which a metal in solution reacts with a chemical in any form, or there is simply a change in the pH of the solution, so that an insoluble metal compound or precipitate is formed. The precipitation of a given metal depends on a series of variables, with the initial concentration of the metal in the solution being a key parameter in allowing this precipitation to occur. Precipitation and crystallization should not be confused, because crystallization (commonly used in the case of tungsten) is a physical process in which a salt in the form of crystals is separated from an aqueous solution. This crystallization is the consequence of an evaporation procedure in which the solution is supersaturated, allowing nucleation and further growth of crystals.

The fixation of NO_x, from flue gas in a coal-fired power plant is effectively carried out using a selective catalytic reduction process. After a period of use, the catalyst (V₂O₅–WO₃/TiO₂) becomes ineffective; thus, the recovery of the metals containing in the spent catalyst is of interest both from economical and environmental points of view. One way to treat these spent catalysts via reducing acid leaching and a further roasting-water leaching operation was describes in [55]. This operation consisted of various steps:

(i)

The spent catalyst was leached with water, in order to remove impurities.

(ii)

The washed residue from step (i) was treated with H₂SO₄ solution and ascorbic acid; from the treated residue, the leachate vanadium was recovered.

(iii)

The residue from step (ii) was mixed with NaOH–Na₂CO₃ and roasted at 750 °C for two hours.

(iv)

The roasted material was leached with water (70 °C and two hours) to yield the tungsten-bearing solution. This solution was treated with HCl in order to shift the pH to the 9–9.5 range, promoting precipitation of silicon (H₂SiO₃) and aluminum (AlOOH and Al(OH)₃). The solution thus obtained was then treated with sulfuric acid to precipitate tungsten. The best conditions were fixed as follows: 30 g/L W, 8 M sulfuric acid, 70 °C, two hours. As final product, H₂WO₄ was obtained. This crude tungstic acid was purified by leaching in concentrated ammonia (twelve hours) to yield a tungsten solution (120 g/L) which was evaporated to crystallize 5(NH₄)₂O·12WO₃·5H₂O.

In [56], tungsten and nickel were recovered from metal-bearing electroplating wastewater (Chinese origin). Firstly, tungsten was recovered from the source material (46.1 g/L W, 39.6 g/L Ni at pH 2.5) by precipitation with HCl (unknown concentration) at 90 °C, with one hour of reaction time and use of 3.5 times the theoretical concentration of the acid. Under these conditions, tungsten precipitated as H₃PW₁₂O₄₀. No further data were given regarding the processing of this solid. It is curious that the authors of this work mention that the product could be refined according to [57] when this formulation did not appear in the referenced work. From the resulting solution, nickel was recovered by thermal oxidation using sodium hypochlorite (NaOCl) and subsequent precipitation in alkaline medium.

It is probable that precipitation is the least selective unit operation of all of these reviewed in the present work. This implies that, before reaching this stage, the solution needs to pass through a series of steps to reach a purity and concentration of the metal which is adequate obtain a pure product. Advantages of this method include its use of relatively inexpensive precipitation agents and, of course, its widespread industrial use.

3. Conclusions

This work demonstrates that the scientific community is constantly searching for new hydrometallurgical processes aimed at the recovery of tungsten from both raw and secondary materials. A general flow-sheet for the treatment of tungsten-bearing primary or secondary sources is shown in Figure 1.

Table 8. Summary of recent relevant operations to recover tungsten from solutions.

Operation	W in Feed Solution	Final Product	Ref.
Leaching + electrodialysis	74–87 g/L	Na2WO4	[25]
Leaching + precipitation	Unknown	H2WO4	[27]
Leaching + LLE + precipitation	18 g/L	W powder	[28]
Leaching + evaporation + crystallization	Unknown	W powder	[30]
Leaching + IXRs + evaporation + crystallization	71 g/L	APT	[32]
Leaching + precipitation	Unknown	H2WO4	[35]
Leaching + precipitation	6 g/L	CaWO4	[36]
Leaching + precipitation	Unknown	WO3	[37]
Leaching + evaporation + crystallization	174 g/L	APT	[39]
IXRs	9 g/L	Na2WO4	[42]
IXRs	16 g/L	Na2WO4	[43]
Adsorption	100 mg/L	Na2WO4	[45]
Adsorption	37 mg/L	Na2WO4	[48]
Adsorption	9 g/L	Na2WO4	[49]
Adsorption	400 mg/L	Na2WO4	[50]
LLE	24 g/L	Na2WO4	[52]
LLE	7 g/L	Na2WO4	[54]
Precipitation	40 g/L	APT	[55]
Precipitation	461 g/L	H3PW12O40	[56]

LLE—liquid–liquid extraction. IXRs—ion exchange resins. APT—ammonium paratungstate.

summarizes results obtained in this review.

As can be seen in Figure 1, the procedure to recover tungsten from any material consists of a series of steps. After a pretreatment step in which undesirable material is separated from tungsten material, the solid is crushed and milled if necessary. This material is subjected to a leaching step; however, in some cases, e.g., in the presence of wolframite, the starting material needs to be subjected to a roasting step in which wolframite is decomposed into a tungstate salt (i.e., Na₂WO₄). The leaching step can be carried out in acidic, alkaline or H₂O media, under atmospheric or pressure conditions, and at various temperatures. The products of this step can be purified by any of the separation technologies described in this review to yield ammonium paratungstate or tungstic acid, although in some studies synthetic scheelite was produced as final product. These chemicals are finally subjected to a reduction operation to yield tungsten metal powder. Thus, the overall flow-sheet consists of three main sequential operations or sequential blocks:

(i)

Decomposition of tungsten source material,

(ii)

Preparation of pure tungsten compounds,

(iii)

Production of tungsten metal powder.

The investigations reviewed in this work undoubtedly contribute to the development of the highly acclaimed circular economy. However, as most if not all of these studies were conducted on a laboratory scale, we will have to wait and see whether any of them have real practical application (including evaluation of economic costs), e.g., scaling up to a pilot-plant or higher level.

Use of the emerging technologies of bioleaching, and phytomining for recycling tungsten-bearing metals from wastes, as well as the use of organic acids (solvometallurgy) in the treatment of hard metal wastes, should open new opportunities in the development of more environmentally friendly processes.

The utilization of tungsten in several components of fusion reactors will serve as the basis for generation of new pollutants such as activated tungsten dust. Thus, new opportunities should arise in the future to investigate the effects of these dusts, and the correct way to treat them, in order to avoid any environmental impact.

During the time of writing this paper (March–April), the news emerged of support by the European Union for a mining project for the recovery of tungsten in Ciudad Real (Spain). The El Moto Project is located in Abenojar village. After the global process (which is also targeted towards gold recovery), a WO₃ concentrate is expected to be yielded.