Molybdenum Recovery from the Copper Hydrometallurgical Extraction Route with High Content of Chloride Ions Using the Ion Exchange Technique

Abstract

This manuscript describes molybdenum recovery from copper Pregnant Leaching Solutions (PLSs) in the copper oxide mining industry with high contents of chloride ions. This product was recovered from the copper leaching pond solutions of the Chilean National Copper Mining Corporation (CODELCO) using the ion exchange process. This process recovered molybdenum from initial Mo concentrations of 50 mg/L using two different anion−exchange resins. The first experiments, with 1 g/L Cl⁻, achieved recovery yields of 90% and molybdenum concentrates as CaMoO₄. However, the process was permanently halted because of the early saturation of the ion exchange resins given by high chloride concentrations (10 g/L Cl⁻) of the original copper PLS and the significant reagent consumption given by the low adsorption capacity. Static studies were developed to determine the adsorption isotherms, followed by continuous molybdenum recovery tests. The Langmuir adsorption parameters were determined as a function of the chloride concentration, giving absorption capacities from 180 to 250 mg Mo/gram of resin and recoveries from 63% to 90%. The breakthrough values for the DOWEX and Lewatit resins for chloride contents of 10 g/L were 180 and 245 BV, respectively, while for 1 g/L, these values were 620 and 890 BV. This allowed obtaining elution solutions of 890 mg Mo/L for the DOWEX resin and 1000 mg Mo/L for the Lewatit resin.

1. Introduction

In 2023, the estimated average molybdenum production increased to 287,000 T, a 9% increase compared with that in 2022. In descending order of production, China, Chile, the United States, Peru, and Mexico provided 93% of total global production [1]. Maintaining a constant production level of this metal is challenging because its main source is as a by−product of other resources, typically copper mines. Molybdenite (MoS₂) is the most common molybdenum ore found around the world. It is usually found in association with other sulfide minerals, most commonly as a byproduct of copper production [2,3,4,5]. This source accounts for over 60% of global molybdenum output. Molybdenum, as a metal alloy, improves steel toughness and heat resistance while also adding hardening ability and corrosion resistance to cast iron. Molybdenum alloys retain their strength at high temperatures, which is important in the aerospace, electrical, industrial motor, and nuclear production industries [6,7]. Molybdenum is found in low−grade porphyry molybdenum deposits as the primary metal sulfide and as a secondary metal sulfide in low−grade porphyry copper deposits. Currently, in 2024, there are around 260,000 tons of identified molybdenum deposits globally, with small declining trends in most of the nations that generate this resource.

Table 1. Molybdenum world mine production and reserves.

Mine Production (Tons)			Reserves
	2022	2023	(thousand MTs)
USA	34,600	34,000	3500
Argentina	−	−	100
Armenia	7800	7800	150
Australia	277	500	690
Canada	1390	970	72
Chile	45,600	46,000	1400
China	106,000	110,000	5800
Iran	3700	3700	43
Korea, North	700	700	NA
Korea, Republic of	367	400	8
Mexico	15,500	15,000	130
Mongolia	3000	3100	NA
Peru	31,600	37,000	1500
Russia	1700	1700	1100
Turkey	−	−	52
Uzbekistan	1700	1700	21
World total	253,000	260,000	15,000

illustrates the global mine output and reserves of molybdenum [8].

Molybdenum production procedures primarily involve the production of molybdenum powder and molybdenum byproducts from chemical purification operations. Figure 1 depicts the stages of molybdenum powder production, which include roasted molybdenite to obtain molybdic acid and ammonium molybdate as raw materials after calcination to obtain molybdenum trioxide (MoO₃). Then, molybdenum powders are obtained through a reducing agent, usually hydrogen [9,10,11]. Figure 1 depicts the specific production processes involved in the production of molybdenum. After the dissolution of molybdenum concentrates, ion exchange and solvent extraction can be employed to separate or purify metal ions present in aqueous solution [12,13,14,15,16]. In these two processes, each has its advantages and disadvantages over the other. Solvent extraction is highly selective and efficient for extracting Mo from sulfuric acid leach solution. However, owing to the co−extraction of impurities, such as Co, Ni, and Fe, Mo separation is inefficient in extraction alone, requiring in some cases previous acid washings [17]. Although the loading capacity of ion exchange for metal ions is low, the operation of ion exchange is simple and producing Mo with high purity is possible [18]. Hence, when the concentration of Mo in sulfuric acid solution is not high, ion exchange resins can selectively load Mo from the solution in the presence of other metals. Therefore, the ion exchange processes are the best way to recover molybdenum from aqueous solutions. In the pH range 1–6, the molybdenum ions exist as polyanions, which are bigger than the monomeric anions. That is why it is convenient to use highly permeable anion exchangers for the ion exchange recovery of molybdenum. Such sorbents are the porous and microporous resins [19,20]. In general, quaternary amine−based resins are used to recover anionic species and they have the following selectivity: SO₄²⁻ > NO₃⁻ > Cl⁻ > HCO₃⁻ > OH⁻ > F⁻ [21,22].

On the other hand, the main source for secondary molybdenum recovery is the catalysts used to desulfurize petroleum, petrochemicals, and coal−derived products [26,27,28]. Several hydro− and pyrometallurgical techniques have been proposed to extract molybdenum from low−grade concentrates. These techniques include hypochlorite, electro−oxidation, nitric acid leaching, alkali and pressure leaching, chlorination, and/or roasting with lime or soda ash. While hypochlorite and electro−oxidation procedures are efficient, they require significant amounts of chemical reagents, making this alternative economically inefficient [29,30,31,32]. Other alternatives include nitric acid digestion, while alkali leaching injects oxygen gas. The high installation costs hinder this process’s commercial viability. The chlorination process uses Cl₂ gas or a combination of chlorine and oxygen to convert molybdenum to chloride, oxychloride, and SO₂. Chlorine gas is poisonous and corrosive, which makes this alternative unfeasible for current industrial standards [33,34,35]. Finally, roasting molybdenum in the presence of lime or soda ash yields molybdenum to molybdate. Lime roasting produces insoluble water molybdate, which requires subsequent leaching, while soda−ash roasting produces water−soluble molybdate in a single stage. The soda ash method yields sodium molybdate [35]. Although this method is preferred because of its high efficiency and low gas emissions, high reagent consumption can be a significant problem for an eventual scaling−up of the processes [36,37,38]. For these reasons, from 2008 to 2012, the Chilean Nuclear Energy Commission (CChEN) worked jointly with CODELCO to develop metallurgical processes to recover byproducts from copper mining operations to lower overall copper extraction operating costs [39]. As a result, CODELCO discovered average molybdenum concentrations of 50 mg/L in their copper leach solutions and concentrated them through the ion exchange technique, giving as a result molybdenum concentrates such as CaMoO₄. A speciation analysis of the conditions under which molybdenum is present in the aqueous solution is required to extract molybdenum from these solutions. Figure 2 shows that the major species for molybdenum at pH values greater than one are from the anion spectrum, and more precisely, the predominant species is Mo₇O₂₁(OH)₃³⁻. However, it is crucial to note that molybdenum is extremely sensitive to pH fluctuations. If the pH of the copper leaching solution falls below one, the major molybdenum species becomes cation charged, reducing the efficacy of the chemical recovery process.

Although the process was successful at the bench scale, the copper PLS contained high amounts of chloride in solution because of the presence of atacamite, Cu₂Cl(OH)₃. This mineral, when leached, releases chloride ions into the solution, which compete with the molybdenum recovery process during adsorption. This resulted in decreased process performance and increased reagent consumption while recovering molybdenum from anion−exchange resins. As a result, the method became commercially unsustainable, and the molybdenum production plant ceased functioning. This manuscript shows the methodology and the results obtained during laboratory− and pilot−scale tests. Since the main source of molybdenum production is the secondary ore molybdenite, there are few alternative sources to obtain this metal. Furthermore, there is limited information on the recovery of this metal from sources other than molybdenite, making the scaling up of this class of methods problematic.

2. Materials and Methods

2.1. Experimental Development—Laboratory Tests

Table 2. Composition of the copper leaching solution.

Cu (g/L)	Mo (mg/L)	Fe (g/L)	Zn (g/L)	Cl (g/L)	SO42− (g/L)	pH	Eh
2.5	50	5	0.5	1.0−5.0−10.0	91.8	1.0–1.5	0.3–0.5

shows the chemical analysis of the synthetic copper PLS used in the assays, which was carried out at room temperature (25 °C). The chemical composition of the PLS solutions used for all the tests was determined using inductively coupled plasma mass spectrometry (ICP−MS). The molybdenum adsorbed at the resin’s surface was determined using the difference in molybdenum concentrations from the initial and final solutions. Since the chloride concentration in the copper PLS solution changes substantially during the industrial operation, it is necessary to study its influence on the molybdenum adsorption results, and to a lesser degree, the process’s performance may be hindered by the presence of ferrous ions and zinc. For this reason, the effect of three different concentrations of chloride ions in synthetic solution was also investigated: 1, 5, and 10 g/L.

Since copper is the foundation of Chile’s mining industry, it is critical to underline that the fundamental condition for completing this project is that the composition of the copper−bearing solution cannot be changed. For this reason, the ion exchange method was the best choice. The experimental development of this work comprised two stages: static− and continuous−type tests. Figure 3 illustrates the static tests [43]:

The adsorption properties of the anion−exchange resins available for use were determined using static ion−exchange tests. Two ion exchange resins were available for the experiments: DOWEX M−43 (Delft, The Netherlands), a strongly basic resin, and LEWATIT MP 62 WS (Cologne, Germany), a weakly basic resin. Both resins are quaternary amine based. The goal of these tests was to identify the key adsorption parameters for each resin using the copper leaching solution from Table 2 based on the Langmuir theory. The resins were pre−conditioned with a 10% hydrochloric acid solution prior to the ion exchange tests to boost the resin affinity for the molybdenum anion of interest. These tests were made using a constant volume of 6000 cm³ copper leaching solution and 5 different quantities of the resins under study, 0.1, 0.25, 0.5, 0.75, and 1.0 g, for 24 h and at room temperature.

The molybdenum recovery process was as follows [44]:

(a)

Resin preconditioning: Both resins need a HCl preconditioning to increase their affinity for molybdenum anions:

$$R^{+}{\left(OH\right)}^{-}+H^{+}C{l}^{-}\to R^{+}C{l}^{-}+H^{+}O{H}^{-}$$

(1)

(b)

Molybdenum recovery from the copper PLS: Molybdenum can be recovered and exchanged during the elution stage. According to the Figure 2b, molybdenum, at pH values 1.0–2.5, was present mostly as Mo₇O₂₁(OH)₃³⁻. Equation (2) illustrates the ion exchange mechanism:

$$3R^{+}C{l}^{-}+M{o}_7{O}_{21}{\left(OH\right)}_3^{3-}\to R_3^{+}M{o}_7{O}_{21}{\left(OH\right)}_3^{3-}+3C{l}^{-}$$

(2)

(c)

Molybdenum elution: According to the Pourbaix diagram from Figure 2a, molybdenum can be effectively recovered from the resin phase using a NaOH solution, because in these conditions, the prevalent species are the anions. Equation (3) shows the ion exchange:

$$R_3^{+}M{o}_7{O}_{21}{\left(OH\right)}_3^{3-}+O{H}^{-}\to 3R^{+}O{H}^{-}+M{o}_7{O}_{21}{\left(OH\right)}_3^{3-}$$

(3)

For pH values higher than 7, molybdenum will be present as MoO₄⁻.

(d)

Resin regeneration: To carry out the process continuously, the resin conditioning must start again, as described in Equation (1).

The above process description shows that the concentration of chlorides is closely related to the total process efficiency. If this ion is present in high concentrations, the adsorption mechanism stated in Equation (2) will be severely constrained. As a result, it is critical to understand the effect of chloride ions on the overall process. Given that the copper PLS solution had a natural concentration of around 10 g/L, it is critical to compare the performance of the continuous adsorption curves for real and synthetic PLS solutions without chlorides. Finally, the best performing resin was tested with temperatures of 40 °C and 55 °C to increase the molybdenum yield. Although the molar concentration of sulfates in the copper PLS was higher than the chlorides, their presence is not as detrimental as the chloride ions. This was because of the sulfate speciation shown in Figure 4. According to this figure, SO₄²⁻ ions are mostly present as HSO₄⁻ ions. At pH values of 1.0–1.5, the ion exchange resin loses selectivity towards sulfates because the anion−exchange resins are less selective towards the ion HSO₄⁻ than sulfates [45,46,47].

2.2. Langmuir Adsorption Analysis

The Langmuir model is an analytical technique to evaluate the resin’s sorption capacity for the molybdenum ion recovery from aqueous solutions as adsorbent. This model’s basic assumption model is the formation of a monolayer throughout the adsorbent’s surface, assuming that only one molecule could be adsorbed on one adsorption site and the intermolecular forces decrease with the distance. The adsorbent surface is homogeneous in character and possesses identical and energetically equivalent adsorption sites. The Langmuir equation is presented as Equation (4) [48,49,50]:

$$q=q_{MAX}*{\displaystyle \frac{C*K_L}{1+C_e{K}_L}}$$

(4)

“C” is the molybdenum concentration at equilibrium, and “q” is the quantity of total ions adsorbed at the resin’s surface divided by the total mass of resin. “q_MAX” is the resin’s maximum adsorption capacity to form a single layer and K_L is the Langmuir adsorption isotherm’s equilibrium constant. All the Langmuir adsorption parameters can be calculated from the slopes and intercept of the linear plot of 1/q versus 1/C, according to Equation (5).

$${\displaystyle \frac{1}{q}}={\displaystyle \frac{1}{q_{MAX}}}+{\displaystyle \frac{1}{K_L*q_{MAX}}}*{\displaystyle \frac{1}{C}}$$

(5)

The trend on the adsorption process, whether it is favorable or not, can be determined using the dimensionless constant “R_L”, given by Equation (6).

$$R_L={\displaystyle \frac{1}{1+K_L{C}_0}}$$

(6)

where C₀ is the initial molybdenum concentration. Typical values of R_L are located within the range 0–1. This parameter indicates whether the adsorption process is unfavorable (R_L > 1), favorable (0 < R_L < 1), or irreversible (R_L = 0). Both resins were studied using the lineal regression from Equation (5) under static conditions.

2.3. Experimental Development—Continuous Tests

Following the static tests, both resins underwent continuous−type tests, as shown in Figure 5. The continuous tests were carried out in a glass column.

Table 3. Dimensions of the column used for continuous tests.

Ion Exchange Column
Diameter of the column (D)	2 cm
Height of the column (h)	100 cm
Time (t)	2 h
Volumetric flow rate (f)	5 BV/h
Linear flow rate (u)	2 × 10−3 m/s
Resin porosity (e)	85%

provides the column´s measurements. To control the solution feed flow to the column, continuous tests were performed utilizing a 500 series cased peristaltic pump and the copper leach solution pond. Samples were taken every 20 min at the end of the column to determine the molybdenum concentration (C) relative to the original concentration (C₀). When this parameter C/C₀ reaches the value 1%, this point is called the breakthrough point, and when this value reaches 99%, this is called the saturation curve.

3. Results and Discussions

The experimental results were the following: Figure 6 depicts the obtained adsorption isotherms, while

Table 4. Langmuir adsorption parameters and recoveries.

Resin	Cl− (g/L)	${\mathbf{q}}_{\mathbf{MAX}}, \frac{\mathbf{m}\mathbf{g} \mathbf{M}\mathbf{o}}{\mathbf{g}\mathbf{r}\mathbf{a}\mathbf{m} \mathbf{o}\mathbf{f} \mathbf{r}\mathbf{e}\mathbf{s}\mathbf{i}\mathbf{n}}$	RL	KL, L/mg	Mo Yields	Zn Yields	Fe Yields
DOWEX M43	1	250	2.4E−03	8.2	83%	47%	19%
DOWEX M43	5	190	4.4E−03	4.5	63%	36%	13%
DOWEX M43	10	180	6.2E−03	3.2	60%	22%	7%
Lewatit MP 62−WS	1	270	2.0E−03	10.0	90%	53%	15%
Lewatit MP 62−WS	5	210	3.1E−03	6.5	70%	42%	8%
Lewatit MP 62−WS	10	190	4.7E−03	4.2	63%	30%	2%

illustrates the Langmuir parameters and recovery yields for various chloride concentrations in the copper PLS:

The adsorption process starts once the resin comes into contact with the molybdenum carrier solution point, and this adsorbate can be eluted from the resin by one or more of the following mechanisms: (1) continued elution with the initial mobile phase (adsorbate bound mildly to the column); (2) pH gradient; or (3) ionic strength gradient [51]. The results show that the role of chloride ions is a crucial component of these findings. The first working condition evaluated was with synthetic PLS, 1 g/L Cl⁻, for both resins. In both cases, for DOWEX and Lewatit resins, Mo recoveries reached maximum recoveries of 83% and 90%, while the maximum loading capacities were 250 and 270 mg Mo/L of resin, respectively. This process is highly favorable towards the molybdenum ion, as shown by the low R_L values from Table 4, thanks to the molybdenum’s polymeric anionic forms of the type [(MoO₃)_x*(HMoO₄)_y ]^−y, where “x” and “y” may have values from 2 to 24, also called isopoly−acids [20]. However, since the actual working solutions contained high concentrations of chlorides, these adsorption conditions were also evaluated. These ions were incorporated by the leaching of atacamite ores (Cu₂Cl(OH)₃), which enhanced the concentration of chloride in solution. In these cases, it is possible to see a decrease in the loading capacity of both DOWEX and Lewatit resins to 180 and 190 mg Mo/L of resin, respectively, and an increase in the K_L parameter, which implies a loss of selectivity of the process towards molybdenum. High chloride ion concentrations, according to Equation (2), decrease the maximum adsorption capacity of molybdenum ions across the resin’s surface and increase the PLS’s ionic strength. In this situation, high chloride concentrations reduce the selectivity towards the Mo₇O₂₁(OH)₃³⁻ ion, lowering the total yield of the process. The leaching solutions under study possessed low levels of ionic strength. At low ionic strengths, the selectivity for the present ions is at a maximum and the adsorbates are bounded strongly. Increasing the electrolyte’s ionic strength enhances competition and reduces the interaction between the resin and the species of interest, resulting in their elution during the adsorption stage [52].

Influence of the impurities: Another consideration is the concentration of contaminants in the solution, such as iron or zinc. In solution, both species are present in low amounts. However, it is vital to investigate their impact on this process. The speciations in Figure 7a,b show how these impurities behave, respectively. For the adsorption tests, iron reached a maximum recovery yield of 19% for the DOWEX resin, while zinc reached a maximum of 53% for the Lewatit resin, both at 1 g/L Cl⁻. Given that the Fe²⁺ ion does not act like an anion in this scenario, iron does not pose severe issues under the potential circumstances of the solution. However, because of its complexation with sulfate ions, the Zn²⁺ ion behaves like an anion, in the form of ZnSO₄²⁻, and can contaminate the final samples. Nevertheless, since both ions behave as hydroxides at basic pH values, both impurities can be purged from the molybdenum solution during the elution stage.

Figure 8a,b show the results of the continuous ion adsorption tests. Because of their affinity for anion−exchange resins and their abundance in solution, chloride ions occupy the bulk of anionic resin active sites. At low concentrations of chlorides in solution (1 g/L Cl⁻), the resins reached breaking points were 620 BV for the DOWEX M43 resin and 891 BV for the Lewatit MP 62 WS resin. However, as the concentration of chlorides in solution increases, for 5 g/L Cl⁻, the breakpoints dropped to 250 BV for DOWEX resin and 405 BV for the Lewatit resin, and at high chloride concentrations (10 g/L), they dropped to 185 BV and 245 BV, respectively. These drops in the breakthrough points of the continuous ion exchange tests caused the original process to recover molybdenum to become economically unprofitable due to the high consumption of chemical reagents to carry out the subsequent molybdenum elution from the column.

Figure 9 show the influence of the temperature on the continuous ion adsorption tests. The diffusion coefficient and the equilibrium conditions attained for the isotherms are both affected by temperature. Although working at room temperature is the most usual procedure in ion exchange, these two methods can boost molybdenum recovery yield. Though increasing this parameter can enhance the results, the rise in the breakthrough point from 890 BV at 25 °C to 907 BV at 40 °C and 940 BV at 55 °C was insufficient to justify the practical application. As a result, the process was carried out at room temperature.

Finally, Figure 10 illustrates the elution curves for both resins using a 5% NaOH solution. Ionic species are usually eluted from the column by increasing the ionic strength of the buffer or by increasing the salt concentration or a pH gradient to gradually increase the species concentration. Given the low order of magnitude of the R_L affinity coefficient, a high pH gradient is necessary to reverse the molybdenum desorption process from the ion exchange resin. In this case, the pH gradient from the acidic solution to the NaOH solution allowed effective molybdenum desorption. The pH of the elution solution also affects the molybdenum speciation and, therefore, the selectivity and the ionic strength of the buffer, decreasing its retention from the resin [53,54]. Decreasing the chloride ions improves the resin’s adsorption capacity, but it also increases the amount of the reagents required to recover molybdate ion MoO₄²⁻ from the resin’s surface. Thus, the molybdenum eluate solutions reached concentrations of 1000 mg/L for Lewatit resin and 890 mg/L for DOWEX resin.

Table 5. Elution recoveries.

Resin	Cl− (g/L)	Mo Yields	Zn Yields	Fe Yields
DOWEX M43	1	89%	0.3%	0%
DOWEX M43	5	85%	0.2%	0%
DOWEX M43	10	78%	0.15%	0%
Lewatit MP 62−WS	1	95%	0.32%	0%
Lewatit MP 62−WS	5	89%	0.23%	0%
Lewatit MP 62−WS	10	83%	0.2%	0%

illustrates the recovery efficiency of molybdenum and Fe and Zn impurities carried over during elution. These results show that Zn impurities can slightly contaminate these solutions, with a maximum of 0.32% for the Lewatit resin. If they were to accumulate as a result of multiple cycles, Zn may be a problem for the elution, but iron ions are not transferred to the basic NaOH solution, so they do not contaminate the final molybdenum product.

4. Conclusions

The main conclusions of this work are the following:

Molybdenum can be recovered as a byproduct of the copper oxide industry from PLSs to boost the production of valuable metals. This was supported by the anion−exchange resin‘s affinity R_L towards the molybdenum‘s polyvalent polymeric form.

This process’s technical feasibility was highly tied to the copper PLS’s chloride ion concentration. For the Lewatit MP 62 WS resin, at 10 g/L Cl⁻, the breakthrough point was 240 BV. For 5 g/L Cl⁻, it was 405 BV, and for 1 g/L Cl⁻, it was 890 BV. For the DOWEX M43 resin, these values were 185 BV, 250 BV, and 620 BV for 10, 5, and 1 g/L Cl⁻, respectively. Despite the favorable behavior of the resin towards the molybdenum given by the R_L factor, during the adsorption process, there was an important loss of selectivity towards the molybdenum to the chloride in solution. This was because the adsorption mechanisms are highly dependent on the diffusion in high concentration gradients of chloride ions.

The presence of the impurities affects the process in different ways. In the case of iron, its effect depends on the solution’s potential. In ferrous form, it is not capable of complexing to an anionic form, slightly affecting the process. On the other hand, it is very important to control the Zn concentration because of its high capacity for forming anionic complexes in chloride solutions for both adsorption and elution.

Even though temperature enhances the diffusion processes and the resin’s adsorption capacity, the improvement margin over room temperature is insufficient to justify its use in the process. Thanks to the stability of the MoO₄^2- ion at basic pH, the elution process could be carried out selectively, recovering molybdenum and isolating it from the Fe and Zn impurities.