Thermodynamic Modeling and Experimental Validation for Thermal Beneficiation of Tungsten-Bearing Materials

Abstract

Tungsten (W), a rare metal, is categorized as a Critical and Strategic Raw Material (CRM) by the European Union (EU), with the highest economic importance of all selected CRMs since 2014. Tungsten and its derivatives are extracted from their commercial raw materials, mainly wolframite [(Fe,Mn)WO₄] and scheelite (CaWO₄) ores. Subsequently to mining and mineral processing, the W ore is submitted to thermal treatment and hydrometallurgy under aggressive conditions (high pressure and temperature), which are usually applied for the extraction of tungsten compounds. This paper aims to investigate a thermal route for scheelite processing using various selected chemical agents, resulting in a W-bearing material that is capable of being leached under softer conditions. In this context, a thermodynamic study of the interaction between FeWO₄, MnWO₄ and CaWO₄ and various chemical reagents is described. The thermochemical calculations and data modeling show that, among other considerations, the reaction of CaWO₄ with magnesium chloride (MgCl₂) can lead to the formation of magnesium tungsten oxide (MgWO₄), which appears to be more easily leachable than CaWO₄. Experimental tests of the reaction of scheelite with MgCl₂ appear to validate the thermodynamic predictions with satisfactory process kinetics at temperatures from 725 to 775 °C.

1. Introduction

Energy transition, the manufacturing of smart materials, diverse advanced industries and the development of highly innovative technologies, require increasingly large amounts of several raw materials (mainly metallic raw materials). According to European (EU) assessments [1], a good number of these raw materials are considered to be Critical and Strategic Raw Materials (CRMs). Such a designation results from their great economic importance, combined with a high supply risk. The number of CRMs for the EU has grown from 14 CRMs in 2011 to 27 in 2017, and was augmented thereafter to reach 34 CRMs in 2023. Please note that copper and nickel do not meet the critical thresholds, but they are included in CRMs as Strategic Raw Materials.

France, like other EU countries, is strongly dependent on CRMs for its industrialization and support of emerging technologies. Efforts to enable the sustainable supply of these materials from domestic and international sources are currently ongoing. National scientific programs, including the Priority Research Program and Equipments (PEPR), under the management of the French National Research Agency, ANR, launch initiatives to consolidate the country’s scientific and technological base. One of the PEPR programs (The Subsoil (resources) exploratory PEPR), entitled “Innovative and sustainable Technologies platform—InnovTech”, aims to develop new approaches and technology to reduce energy consumption and waste in different stages of metal extraction. The project is coordinated by the BRGM (The French Geological Survey Office) and the co-leader is the GeoRessources laboratory at the “Université de Lorraine” (France). One of the targeted metals of this program is tungsten (W) contained in scheelite skarns from the “French Massif Central” deposit. Please note that tungsten has the highest economic importance among all of the selected CRMs.

Tungsten (meaning heavy stone in Swedish) is also known as wolfram (from wolframite, said to be named from wolf rahm or spumi lupi, because the ore interfered with the smelting of tin and was supposed to devour the tin) [2]. Tungsten has the highest melting point “m.p.” (3422 °C) of all metals, boils around 5555 °C and is a heavy metal with a specific gravity of 19.3. Tungsten is one of the transition and refractory metals and belongs, along with chromium and molybdenum, to group 6 on the periodic table. All of these elements are known to have diverse valences, which, in common compounds, range mainly from +2 to +6.

Tungsten belongs to the rare elements, with an earth’s crust abundance of 1.25 mg/kg, but it is widespread within all continents. Tungsten occurs in wolframite [(Fe,Mn)WO₄], scheelite (CaWO₄), hubnerite (MnWO₄) and ferberite (FeWO₄) [2].

As described in Reference [3], the amount of tungsten consumed in the EU is used for a wide range of applications, which are as follows: 67% of tungsten is used for the manufacturing of tungsten carbides, 11% is used for the manufacturing of W-metal, 11% is used for the manufacturing of chemicals, 8% is used for the manufacturing of tungsten-based steel applications and the rest (3% of W) is used for the manufacturing of other tungsten alloys. The repartition of tungsten-use in the EU in various applications is shown in Figure 1 [3]. Mill and cutting tools, mining and construction and wear tools, with 33%, 23% and 18%, respectively, share the main applications of tungsten-bearing materials. The remaining material, about a quarter of tungsten’s end uses, is distributed across other applications, as shown in Figure 1.

According to the available data [4], the world mine production of tungsten is estimated at 78,000 tons in 2023 (Figure 2). The mine production of tungsten is dominated by China, providing more than 80% of the global W-production. As far as the European Union is concerned, the main mining production of tungsten is concentrated mostly in Austria, Portugal and Spain.

Although tungsten resources and reserves are scattered across many countries, China remains prevalent [4], with nearly 52% of the world’s reserves, estimated at 4.4 × 10⁶·tons (Figure 3). Australia and Russia hold second and third place, with their W-reserves estimated at around 13% and 9%, respectively.

As mentioned above, the main tungsten ores of industrial importance are scheelite (CaWO₄) and wolframite [(Fe,Mn)WO₄), the latter being a solid solution of the isomorphic minerals, namely ferberite and hubnerite. Scheelite and wolframite account for 70% and 30% of the world resources, respectively. In general, most deposits have low WO₃ content, which can, at most, reach 2% WO₃ for richer W-ores; however, the extractive metallurgy of tungsten requires concentrates with around 65% WO₃ content.

Therefore, the ores must first go through physical enrichment processes that are often complicated due to their complex elemental, chemical and mineralogical compositions [5]. All of the known mineral processing methods, as well as the development of advanced flow-sheets for the treatment of scheelite ores, resulting in a W-concentrate of metallurgical grade, were summarized in Reference [5].

However, the real challenge of the tungsten industry is the metallurgical extraction of tungsten intermediates, such as ammonium paratungstate (APT), with a chemical formula of (NH₄)₁₀[H₂W₁₂O₄₂]·4H₂O, being the most important precursor for the majority of tungsten products. Figure 4 [6] groups simplified schemes for treatments of low- and high-grade scheelite and wolframite concentrates, resulting in APT manufacturing. In general, concentrates undergo a pretreatment (thermal and/or acidic) to remove impurities (S, As, P, organic carbon, etc.) that would be difficult and expensive to remove later. The hydrochloric acid digestion around 90 °C is industrially employed, mainly for the high-grade scheelite concentrates [6,7], producing pure tungstic acid (H₂WO₄ or WO₃·H₂O) and/or APT after dissolving H₂WO₄ into concentrated NH₃·H₂O.

The most current commercial process by soda/caustic soda digestion (high-pressure and high-temperature processes mainly applied), followed by solvent extraction or an ion exchange step, can accommodate a variability of raw materials. However, high reagent consumption, the absence of recycling the aqueous solutions and chemical reagents, as well as discharge of high-salinity wastewater for solvent extraction and ion exchange steps, are some of the process drawbacks [7]. Novel laboratory routes for improving the existing processes, leading to better leaching and efficiency recovery of tungsten from its concentrates, are well-documented in Reference [7].

There are also some thermochemical insights and experimental laboratory works describing new thermal alternative methods to replace the current industrial processes for tungsten extraction from its bearing materials. An early study [8] predicted the reactions of gangue minerals (SiO₂, Ca₅(PO₄)₃F, SnO₂, FeS₂, MoS₂ and FeAsS) with CaCO₃ and Ca₃WO₆, resulting in mixed compounds which nature depends on for the oxygen partial pressure in the selected system. Xu et al. [9] performed a thermodynamic analysis of leaching of the thermally synthetized Ca₃WO₆ in NH₃-(NH₄)₂CO₃-H₂O solution. Their experimental results, in which they simultaneously applied grinding and leaching, indicated an almost full dissolution of Ca₃WO₆ at a temperature close to 40 °C by forming (NH₄)₂WO_4(aq) and solid CaCO₃. Another scientific approach [10] provided a leaching yield of MgWO₄ that was higher than 98% in a solution of NaOH at 90 °C, after thermal synthesis of magnesium tungstate from scheelite using MgCl₂.

Recently [11], a carbo-thermic reduction in the presence of calcium sulfate was applied to a wolframite–scheelite mixed ore containing tin as cassiterite (SnO₂). About 99% of tin is separated by volatilization as SnS. Wolframite is converted to CaWO₄ and Ca₂(Fe,Mn)WO₆, which are easily digested by the mixed sulphuric–phosphoric acid solution with a leaching efficiency of tungsten of around 99%. The kinetics of chlorination of a scheelite–wolframite concentrate with chlorine and sulphur dioxide are investigated early [12]. They identified WO₂Cl₂, FeCl₂, FeCl₃, S₂, FeS, CaSO₄ and CaCl₂ as the main reaction products. In a similar study, Menéndez et al. [13] performed the physicochemical characterization and chlorination of several natural wolframite (Mn_xFe_1−xWO₄) samples with chlorine and sulphur dioxide. Under the best conditions, they announced an extraction extent of tungsten that was close to 86%. Fuga et al. [14] studied the chlorination kinetics of a MnWO₄ reaction with chlorine between 650 and 950 °C, and they identified WO₂Cl₂ and MnCl₂ as the main reaction products.

The aim of this paper is to provide a comprehensive thermochemical assessment of the interaction between tungsten-bearing minerals and diverse chemical reagents. After predicting the thermodynamic reactivity of tungsten compounds towards selected chemical agents, special attention was paid to the modeling of the equilibrium composition of various tungsten-containing systems at chosen temperatures. Furthermore, experimental tests of the reactions of scheelite with calcium carbonate (CaCO₃) and calcium chloride (MgCl₂) are carried out to verify the validity of the thermodynamic predictions.

2. Phase Diagrams of the CaO-WO₃ and MgO-WO₃ Systems

Both calcium and magnesium oxides and their mixture are largely used in the manufacturing of basic refractory bricks due to their chemical and mechanical performance, as well as their high melting points (2613 °C for CaO and 2852 °C for MgO). In addition, the ternary and quaternary phase diagram systems, CaO-MgO-SiO₂, CaO-MgO-SiO₂-Al₂O₃ and many others, are the basis of extractive metallurgy (especially in slag formation) and in the synthesis of glass and glass–ceramic materials.

In the context of this study, it is interesting to examine the available binary phase diagrams of CaO(MgO)-WO₃ systems. In the CaO-WO₃ phase diagram depicted in Figure 5a, there are two well-defined and stable compounds, CaWO₄ and Ca₃WO₆ (3CaO·WO₃), which melt congruently at about 1580 °C and 2250 °C, respectively. According to this diagram, both compounds do not show any crystalline structural change, whilst the tungsten (VI) oxide (WO₃) crystal structure is temperature-dependent. Furthermore, it is orthorhombic at 600–745 °C and it is transformed into a tetrahedral state beyond this temperature.

The MgO-WO₃ phase diagram presented in Figure 5b demonstrates the formation of a single compound, namely magnesium tungstate (MgWO₄), which is stable up to its melting point (ca. 1385 °C); however, it undergoes a reversible structural phase transition at around 1165 °C. The features of this transformation are described elsewhere [16].

These observations prove that, in addition to scheelite, two other compounds that are thermally stable, at least at T < 1350 °C, must be taken into account in the studied systems. Hence, this statement will be used for the thermodynamic investigation of the interaction between scheelite and the selected chemical reagents.

3. Thermodynamic Elements of the Interaction Between (Fe, Mn, Ca) Tungstate and Various Chemical Reagents

As mentioned in the Section 1, one of the aims of this study is to thermally transform tungsten-bearing materials (FeWO₄, MnWO₄ and CaWO₄) into less-common compounds (such as Ca₃WO₆ and MgWO₄) that are more likely to be digested by current chemical reagents under mild conditions. HSC Chemistry Software^® (version 5.1) [17], as well other available data [18], are used for the thermodynamic predictions and assessment of the selected system. Although there are other types of computer software and databases for thermodynamic predictions (such as FactSage, Thermo-Calc Software, etc.), HSC Chemistry appears to be the most widely used. The used methodology, as well as the precautions taken into account during the thermochemical calculations, were mentioned in a recent study [19]. The thermodynamic data (enthalpy (H), entropy (S) and heat capacity ©) of some chemical species are sometimes given in a low and narrow range of temperatures, and HSC data extrapolation at high temperatures may lead to errors. As an example linked to this study showed, the thermodynamic data for CaWO₄ and Ca₃WO₆ are available up to 1000 K (727 °C). However, as shown in Figure 5a, these well-defined compounds are thermally stable at temperatures well above 727 °C and do not undergo any structural and/or transition changes prior to melting. Hence, the thermochemical data for both compounds can be extrapolated without error in the temperature range of interest of this investigation (lower than or equal to 1100 °C).

The first step for an envisaged process is to check the thermodynamic reactivity of input chemicals by evaluating the standard free Gibbs energy changes (ΔG°), computed by the HSC thermochemical database, for foreseen chemical reactions. Figure 6a–c plots the evolution of ΔG° versus temperature for the reactions of FeWO₄, MnWO₄ and CaWO₄ with selected chemical reagents, as described in Equations (1)–(19). There can certainly also be intermediate reactions, such as the formation of FeCO₃, MnCO₃, Fe(HO)₂ and Mn(OH)₂, but in the practical temperature of interest (most likely higher than 600 °C), these compounds are dissociated in respective oxides (FeO and MnO, in absence of oxygen) and their respective gases (CO_2(g) and H₂O_(g)). Likewise, other possible reactions should lead to the formation of more complex intermediate phases, but the thermodynamic data are often lacking. However, as the ΔG° is a state function, the final ΔG° value does not depend on the intermediate paths of the system transformation, but only on the initial and final state of the system.

1/3FeWO_4(s) + CaO_(s) → 1/3Ca₃WO_6(s) + 1/3FeO_(s)

(1)

1/3FeWO_4(s) + Ca(OH)_2(s) → 1/3Ca₃WO_6(s) + 1/3FeO_(s) + H₂O_(g)

(2)

1/3FeWO_4(s) + CaCO_3(s) → 1/3Ca₃WO_6(s) + 1/3FeO_(s) + CO_2(g)

(3)

1/2FeWO_4(s) + NaOH_(s,l) → 1/2Na₂WO_4(s,l) + 1/2FeO_(s) + 1/2H₂O_(g)

(4)

FeWO_4(s) + Na₂CO_3(s,l) → Na₂WO_4(s,l) + FeO_(s) + CO_2(g)

(5)

FeWO_4(s) + MgCl_2(s,l) → MgWO_4(s) + FeCl_2(s,l,g)

(6)

As shown in Figure 6a, the thermodynamic reactivity of FeWO₄, with respect to CaO, Ca(OH)₂, NaOH and MgCl₂, seems to be high (negative values of ΔG°) from around 100 °C. At a higher temperature (T > 500 °C), all of the considered forward reactions (Equations (1)–(6)) are spontaneous. Similarly, the reactions of MnWO₄ (Equations (7)–(12)) with the selected reagents occur spontaneously, starting at least from temperatures higher than 500 °C (Figure 6b).

1/3MnWO_4(s) + CaO_(s) → 1/3Ca₃WO_6(s) + 1/3MnO_(s)

(7)

1/3MnWO_4(s) + Ca(OH)_2(s) → 1/3Ca₃WO_6(s) + 1/3MnO_(s) + H₂O_(g)

(8)

1/3MnWO_4(s) + CaCO_3(s) → 1/3Ca₃WO_6(s) + 1/3MnO_(s) + CO_2(g)

(9)

1/2MnWO_4(s) + NaOH_(s,l) → 1/2Na₂WO_4(s,l) + 1/2MnO_(s) + 1/2H₂O_(g)

(10)

MnWO_4(s) + Na₂CO_3(s,l) → Na₂WO_4(s,l) + MnO_(s) + CO_2(g)

(11)

MnWO_4(s) + MgCl_2(s,l) → MgWO_4(s) + MnCl_2(s,l)

(12)

1/2CaWO_4(s) + CaO_(s) → 1/2Ca₃WO_6(s)

(13)

1/2CaWO_4(s) + Ca(OH)_2(s) → 1/2Ca₃WO_6(s) + H₂O_(g)

(14)

1/2CaWO_4(s) + CaCO_3(s) → 1/2Ca₃WO_6(s) + CO_2(g)

(15)

1/2CaWO_4(s) + NaOH_(s,l) → 1/2Na₂WO_4(s,l) + 1/2CaO_(s) + 1/2H₂O_(g)

(16)

CaWO_4(s) + Na₂CO_3(s,l) → Na₂WO_4(s,l) + CaO_(s) + CO_2(g)

(17)

CaWO_4(s) + MgCl_2(s,l) → MgWO_4(s) + CaCl_2(s,l)

(18)

1/2CaWO_4(s) + NaCl_(s,l) → 1/2Na₂WO_4(s,l) + 1/2CaCl_2(s,l)

(19)

Only the reactions of CaWO₄ with CaO and MgCl₂ (Equations (13) and (18)) are thermodynamically feasible at a low temperature (Figure 6c). For some reactions, such as that with Na₂CO₃ (Equation (17)) and NaCl (Equation (19)), it is necessary to exceed 700 and 950 °C, respectively, in order for the reactions to occur in the desired direction.

Processes based on the roasting of wolframite and scheelite with alkali compounds (especially with Na₂CO₃) are known in the current tungsten extraction processes [6]. Thus, among the chemical processes based on the reactions mentioned above, attempts have been made for those based on the treatment with calcite (CaCO₃) and magnesium chloride (MgCl₂). It is therefore interesting to follow the chemical equilibria between the tungsten compounds and these chosen chemical reagents.

4. Calculation of the Equilibrium Composition in the (Fe,Mn,Ca)-W-C-O Systems

The HSC software® 5.1 also offers the facility to calculate the equilibrium composition of a multicomponent system. The equilibrium composition is calculated using the GIBBS-solver, which uses the Gibbs energy minimization method [17]. Thus, a chemical system seeks the minimum change in Gibbs free energy (ΔG = 0) at the equilibrium state. For a simple system, the equilibrium constant approach is used to find the equilibrium compositions; however, in complex systems with a large number of constituents distributed in different phases (solid–liquid–gas), the calculation of the equilibrium composition is a difficult one. The GIBBS solver module of the HSC allows us to calculate the equilibrium composition as a function of the amount of species introduced at the chosen temperature and pressure. Although there are many options for the result presentation, the obtained data are often plotted as an evolution of the equilibrium composition and as a function of temperature.

Figure 7a shows the evolution of the molar composition of the Fe-Ca-W-C-O system by inputting 1.00 and 3.00 kmol FeWO₄ and CaCO₃, respectively, as starting compounds. To simplify the system, the calculation is made in the absence of oxygen. The species’ equilibrium composition evolves from 100 °C in the following two steps: the first step is up to 450 °C, with a rapid decrease in the amount of FeWO₄ and CaCO₃ and an increase in the amount of CaWO₄, FeO and CO_2(g). Please note that the curves representing CaWO₄, FeO and CO₂ overlap up to 450 °C. This suggests that iron in the ferberite is substituted by calcium-generating scheelite. The second step (T > 450 °C) is characterized by the apparition of a new chemical species (Ca₃WO₆) reaching the value of 1.00 kmol at around 725 °C, while the amount of CaWO₄ tends towards zero at about 725 °C. From this temperature, the equilibrium composition of the system consists of 1.00 kmol Ca₃WO₆, 1.00 kmol FeO and 3.00 kmol CO_2(g), which is consistent with the overall reaction described by Equation (3). Similar observations are valid for the Mn-Ca-W-C-O system (Figure 7b), but the equilibrium evolution is shifted towards higher temperatures. The final equilibrium composition at temperatures near 900 °C is consistent with the stoichiometry of the reaction given in Equation (9).

The simpler presentation is the W-Ca-C-O system, shown in Figure 7c. The evolution of the equilibrium composition starts at temperatures approaching 400 °C and ends at those near 800 °C, with an equilibrium composition of Ca₃WO₆ + CO_2(g) and with a molar ratio of Ca₃WO₆/CO_2(g) = 2, which is in agreement with the stoichiometry of the reaction given in Equation (15).

5. Calculation of the Equilibrium Composition in the Ca-W-Mg-O-Cl System

In the previous section, the equilibrium amount of main constituents in the Ca-W-Mg-O-Cl system was calculated at temperatures up to 1100 °C. Calculations are performed in the absence of oxygen atmosphere. Although the reaction of CaWO₄ with MgCl₂ (Equation (18)) requires an equimolar amount of CaWO₄ and MgCl₂, it was pertinent for modeling the equilibrium amount by varying the MgCl₂ input from 0.50 to 1.50 kmol, while the amount of CaWO₄ was kept constant and equal to 1.00 kmol. The obtained equilibrium diagrams are shown in Figure 8. Using 0.50 kmol MgCl₂ in the starting materials (Figure 8a) and for temperatures lower than 800 °C, the equilibrium amount of the system is composed of equimolar quantities (0.50 kmol) of CaWO₄, MgWO₄ and CaCl₂. Beyond 800 °C, two main species (Ca₃WO₆ and WO₂Cl_2(g)) appear and their amounts increase with the temperature. According to these observations, the likely reaction, thus producing these species, is that described by Equation (20), indicating that CaWO₄ and CaCl₂ are incompatible from a thermodynamic point of view. The calculation with 0.75 kmol MgCl₂ (Figure 8b) also indicates the presence of these two species (Ca₃WO₆ and WO₂Cl_2(g)), but at lower relative amounts. At higher temperatures, magnesium oxide also belongs to the composition of the studied system. The equilibrium composition diagram with 1.00 kmol MgCl₂ is predominated by an equimolar amount of MgWO₄ and CaCl₂ which is consistent with the stoichiometry of Equation (18).

Figure 8c,d shows the equilibrium compositions of the Ca-W-Mg-O-Cl system at 1.25 and 1.50 kmol MgCl₂, respectively. The equilibrium quantities of CaCl₂ remain constant (ca. 1.00 kmol) at all of the considered temperatures. However, the equilibrium amount of MgWO₄ decreases with the temperature, and is more pronounced when 1.50 kmol MgCl₂ is used in the system (Figure 8d). Additionally, equilibrium amounts of WO₂Cl_2(g) and MgO increase significantly at temperatures above 800 °C. From these calculations and according to Equation (21), one may deduce that magnesium chloride can react with scheelite, thus generating calcium chloride and magnesium oxide. It is a spontaneous reaction, as is the case with a ΔG° value at 775 °C of −8.30 kJ/mol. The synthesis of tungsten oxychloride (WO₂Cl_2(g)) can also be described by Equation (22), but this reaction path seems improbable, as its ΔG° value at 775 °C is 72.17 kJ/mol.

The synthesis of metal oxychlorides, such as chromium oxychloride (CrO₂Cl_2(g)), during the oxychlorination of chromite (Fe,Mg)(Cr,Al,Fe)₂O₄ with a gaseous Cl₂ + O₂ mixture, is reported in previous works [20,21,22]; however, due to the high oxidation capacity of Cl_2(g) + O_2(g), chromium(III) is oxidized to chromium(VI), as seen in Equations (23) and (24), while in the case of scheelite, tungsten is already in a hexavalent state. Furthermore, the synthesis of WO₂Cl_2(g), using CaWO₄ and MgCl₂ as starting materials, does not seem to be described in the literature.

2CaWO_4(s) + CaCl_2(s,l) → Ca₃WO_6(s) + WO₂Cl_2(g)

(20)

1/2CaWO_4(s) + MgCl_2(s,l) → 1/2CaCl_2(s,l) + 1/2WO₂Cl_2(g) + MgO_(s)

(21)

MgWO_4(s) + MgCl_2(s,l) → WO₂Cl_2(g) + 2MgO_(s)

(22)

1/2FeCr^III₂O_4(s) + Cl_2(g) + 3/8O_2(g) → 1/4Fe₂O_3(s) + Cr^VIO₂Cl_2(g)

(23)

1/2MgCr^III₂O_4(s) + Cl_2(g) + 1/4O_2(g) → 1/2MgO_(s) + Cr^VIO₂Cl_2(g)

(24)

6. Experimental Insights of Thermal Reactions of CaWO₄ with CaCO₃ and MgCl₂

6.1. Materials and Expermental Procedure

Powdered analytical grade CaCO₃, CaWO₄ and MgCl₂ samples are used is the study. Mixtures of CaWO₄ + CaCO₃ and CaWO₄ + MgCl₂ at defined proportions are produced by dry grinding. The elemental composition and crystallinity of the first two samples, as well as the solid reaction products, are determined using an energy dispersive spectroscopy electron microscopy (EDS-SEM, HITACHI S-4800, Hitachi Ltd., Tokyo, Japan) and X-ray diffraction (XRD, Bruker D8 Advance device, Bruker, Karlsruhe, Germany), respectively. No MgCl₂ analysis was performed due to the highly hygroscopic nature of MgCl₂. The experimental study results were achieved using the setup shown in Figure 9. The main piece of equipment is a tubular resistant furnace with the ability to reach 1500 °C. The working reactor is made of quartz capable of withstanding temperatures as high as 1300 °C in the absence of corrosive atmospheres. When the desired temperature is reached and stabilized under nitrogen flow, the alumina support containing the sample (approximately 1.8 g) is placed in the center of the furnace and the planned treatment time is allowed to elapse.

6.2. Reaction of CaWO₄ with CaCO₃

The formation of Ca₃WO₆ in the W-Ca-C-O system is demonstrated fairly well with thermodynamic assessments. To verify the reliability of thermodynamic calculation, a co-processing of pure CaWO₄ and CaCO₃ that have 15% more CaCO₃ than the stoichiometric molar amount required by Equation (15) is performed in N₂ gas flow at 850 °C, and the solid product obtained after a 120 min of treatment is examined with the XRD technique. Additionally, tests of thermal treatment of only CaCO₃, in N₂ and at various temperatures, show that at 850 °C for a duration time of 30 and 60 min, about 50% and 91% of CaCO₃ are, respectively, decomposed into CaO, according Equation (25).

CaCO_3(s) → CaO _(s) + CO_2(g)

(25)

The analysis of the XRD diffractogram of the (CaWO₄ + CaCO₃) treatment product (Figure 10) confirms the presence of the well-crystallized Ca₃WO₆ with its main XRD peaks (d = 2.84 Å, 4.01 Å and 4.70 Å). However, the initial constituent (CaWO₄) is still the predominant phase (d = 3.11 Å, 4.76 Å and 1.93 Å). In addition, CaO and Ca(OH)₂ (the latter being the product of the interaction of CaO with moisture during sample preparation) are also identified in the thermal treatment sample. This finding indicates that although the reactions of CaCO₃ (CaO) with CaWO₄ are spontaneous processes (at least at T > 600 °C: Equations (13) and (15) and Figure 7c), temperatures higher than 850 °C and/or after an extended time of treatment should be applied for complete conversion of CaWO₃ into Ca₃WO₆. One may conclude that the reactions’ kinetics are the limiting steps of the overall process. Further experimental tests have to be performed to obtain insight into the reaction mechanism and kinetics.

6.3. Reaction of CaWO₄ with MgCl₂

To validate the thermochemical calculation in the CaWO₄-MgCl₂ system, one of the preliminary thermal tests of the CaWO₄ + MgCl₂ mixture (MgCl₂/CaWO₄ = 1.25: molar ratio) is carried out at 775 °C under nitrogen atmosphere for 2 h. This temperature is chosen for at least three reasons, such as the following (i) to follow the thermodynamic predictions; (ii) to insure homogenous mixing between the CaWO₄ solid and MgCl₂ liquid (as the m.p. of MgCl₂ is 714 °C [2]); (iii) to avoid the loss of MgCl_2(l) by evaporation, as this is expected at a temperature higher than 800 °C. The product of the thermal treatment is then washed with deionized water in order to eliminate the chlorides (MgCl₂ and CaCl₂), which cause problems for SEM-EDS and XRD sample preparation and analysis. Chlorides are hygroscopic (easily converted to (Mg,Ca)Cl₂·6H₂O, thus absorbing atmospheric moisture), making their handling and correct consistent analysis difficult. The solid mass after filtration and drying is examined using the XRD and SEM-EDS techniques.

As shown in Figure 11, the magnesium tungstate phase (MgWO₄) is well-crystallized with main peaks (d = 4.69 Å, 3.71 Å, 2.93 Å, 2.90 Å and 3.61 Å) and many other diffraction peaks, which predominate the obtained X-ray diffractogram. CaWO₄ appears to be transformed into MgWO₄. Although some diffraction peaks of CaWO₄ is found overlapping with those of MgWO₄, the presence of CaWO₄ in the product is doubtful or occur as minor phases (i.e., if even present). Several characteristic X-ray peaks of magnesium oxide (MgO) are also revealed in the treatment product, although they are of low intensity. Its presence in the treatment residue may be explained by the reaction given above in Equation (21). Although pure nitrogen is used for experimental testing, traces of oxygen and/or moisture in the system can also result in the conversion of MgCl₂ and MgO.

SEM-EDS results of the treated product are reported in Figure 12a,b. Its morphology shows homogenous particles with sizes less than 10 µm (Figure 12a). Furthermore, another minor particle (spot n° 1) is observed. The general EDS spectrum (red spectrum in Figure 12b) clearly shows the presence of MgWO₄, while the analysis of spot n°1 reveals the presence of MgO. These observations confirm the results obtained by XRD analyses; however, the presence of calcium (as CaWO₄) is not revealed by SEM-EDS, which can be explained as either a complete transformation of CaWO₄ into MgWO₄ or the unreacted CaWO₄ located in the core of the coarse particles, which is not accessible by the EDS probe.

To have an accurate idea of the kinetics of the CaWO₄-MgCl₂ interaction, a set of experimental tests from 725 °C to 775 °C and a 7.5 to 120 min residence time are carried out. As in the previous case, the composition of the obtained products is systematically monitored using the XRD and SEM-EDS techniques.

The XRD patterns for a short-time treatment (7.5 and 30 min) of the resulting products are shown in Figure 13a,b. It is of note that, even at 725 °C, MgWO₄ is the predominant phase when the sample is treated for 7.5 min. Additionally, typical peaks of CaWO₄ are observed, although they are of low intensity. Similar observations are valid for the treatment at 775 °C (Figure 13b). This finding suggests that the kinetic reactivity of MgCl₂-CaWO₄ is high in the temperature range studied, confirming the thermodynamic predictions discussed earlier.

In a recent study [19], it is suggested that MgCl₂ can be used for the extraction of tin from cassiterite (SnO₂) materials via the formation of SnCl_4(g) at around 800 °C. Inferentially, magnesium chloride should also be used for the removal of tin from tungsten ores and concentrates containing SnO₂ as an impurity prior to the hydrometallurgical extraction of tungsten.

7. Conclusions

This study provides some thermodynamic analysis and a modeling approach for the possible conversion of wolframite, and principally scheelite (CaWO₄), into other compounds that are more suitable for subsequent hydrometallurgical processing.

Although treatment of the mineral with CaCO₃ seems to be a feasible process thermodynamically, a satisfactory formation of the reaction products is not observed, which may be attributed to kinetic limitations. Thus, higher temperatures may be required for the treatment of scheelite with this chemical reagent.

Magnesium chloride, on the other hand, appears to be a useful chemical reagent for a satisfactory conversion of scheelite to magnesium tungstate (MgWO₄). Equilibrium-phase diagrams suggest favorable conditions for the transformation of CaWO₄ to MgWO₄. Experimental results of the thermal treatment of scheelite are consistent with the thermodynamic predictions at moderate temperatures with acceptable kinetics. The study reveals that relatively lower temperatures (a little higher than 700 °C) are fine for this process; however, temperatures around 800 °C may be ideal due to the necessity involved in the removal of impurities such as tin (cassiterite), which may be associated with scheelite.

Future work will be devoted to thermochemical investigations on more complex systems, followed by extensive experimental tests on wolframite and scheelite concentrates and their mixtures, thus leading to the synthesis of easily leachable tungsten compounds under mild conditions.