Cobalt Extraction from Sulfate/Chloride Media with Trioctyl(alkyl)phosphonium Chloride Ionic Liquids.

The application of phosphonium-based ionic liquids (ILs) on the selective extraction of cobalt is presented. The extraction mechanism is established, and different parameters of the process are evaluated. It has been found that it is possible to extract cobalt from aqueous solutions in sulfate media, with the addition of sodium chloride, using phosphonium ILs. The cobalt extraction was selective with respect to nickel and strongly dependent on the chloride concentration in the aqueous solution. The cobalt extraction is given by an anion exchange mechanism through an endothermic process. Cobalt extractions greater than 98% were obtained using the proposed methods. Cobalt stripping from the loaded IL phase using water was proved. Therefore, an alternative extraction process to traditional organic solvents is proposed. This alternative has additional advantages such as easy handling, lower costs in reagents and equipment, and risk reduction.

Introduction

The metal processing industry faces critical environmental and economic issues, such as low-concentration metal ores and an increase in the demand for strategic metals, as well as the protection of the environment, reduction in the use of natural resources, and so forth.[1,2] In turn, cobalt is a metal, which is used in many commercial, industrial, and military applications, many of which are strategic and critical.

Acidic solutions that come from the leaching of lateritic nickel minerals, dissolution of metal-rich mattes, or recycling processes are processed in the cobalt and nickel metallurgy. These solutions contain different metallic species, and their separation is a highly challenging problem.[3,4] Traditionally, cobalt and nickel are separated by selective oxidation and precipitation processes which have low efficiency and low selectivity.[5] Extraction by solvents as a final stage to separate the cobalt selectively and with high yields is also used. Organic reagents such as Versatic 10 and LIX63,[6] Cyanex,[7,8] D2EHPA and LIX63,[9,10] Cyanex 272–Cyphos IL 101,[11,12] PRIMENE JMT-Cyanex 272,[13] mixture of PRIMENE JMT-Versatic 10,[14] and extractants derived from D2EHPA[15] have been studied. However, the volatility and flammability of traditional organic solvents usually represent technical problems and risks to the environment and human health.

In recent years, ionic liquids (ILs) have been accepted as an important family of chemical compounds, which have been used for the purpose of both organic diluents and extractive agents instead of traditional or commercial uses. ILs are organic salts or eutectic mixtures of organic and inorganic salts with a melting point below 100 °C. ILs have a structure composed of an organic cation and a polyatomic organic or inorganic anion. Their unique physicochemical properties, such as very high flash point, negligible vapor pressure, high thermal stability, and ability to adjust properties, have led them to an increase in interest for different applications.[16−18] ILs play an active role in the process of partitioning metal ions, with exchange of a metal complex for an ionic or cationic constituent of the IL, which represent an important mode of ionic transfer.[19] ILs could be used to extract metal ions from an aqueous solution phase with relatively high efficiency in a short period of time.[20] For example, Visser et al.[21] designed a series of task-specific ILs based on imidazole for the Hg(II) and Cd(II) extraction. The Nd(III) extraction with 1-hexyl-3-methylimidazolium bis(2-ethylhexyl)phosphate, [C6mim][DEHP], 1-hexyl-1-methylpyrrolidinium bis(2-ethylhexyl)phosphate, [C6mpyr][DEHP], and tetrabutylammonium bis(2-ethylhexyl)phosphate, [N4444][DEHP], were also studied.[22] An aqueous two-phase system formed by tetrabutylphosphonium nitrate, [P4444][NO3], was developed for the selective separation of Nd(III) from Co(II) and Ni(II).[23] Novel thiosalicylate-based ILs were synthesized and investigated for Cu(II), Zn(II), and Cd(II) extraction.[24] Extraction and separation of Pt(IV), Pd(II), Ru(III), and Rh(III) with imidazolium-based ILs were evaluated.[25] The mostly studied ILs for the extraction of metal ions are those of imidazolium derivatives with different extractants, ligands, and metal chelators.[7−9,26−33] However, they present two great disadvantages that cannot be overcome in metal extraction. First, these are hydrolyzed in an acid medium.[34,35] Second, many extraction mechanisms are cation exchange[36,37] or anion exchange,[38,39] which indicates the release of 1-alkyl-3-methyl-imidazolium cations, Cmim+, and bis(trifluoromethanesulfonyl)imide anions, Tf2N–, to the aqueous phase.[40] These compounds can become persistent pollutants in wastewater because of their high stability in water.[41]

Phosphonium salts constitute an important subclass of ILs, which are known to have important properties, sometimes superior in comparison with those of the widely investigated nitrogen-based ILs.[42] Some detailed studies about the synthesis, physicochemical, and thermal properties of tetraalkylphosphonium chlorides have been recently published.[43,44] Phosphonium ILs, compared to 1,3-dialkylimidazolium salts, offer important advantages in applications such as their low cost of preparation and low potential for interaction with various solutes.[45] The extraction of different metallic species with these ILs has been reported: extraction, separation, and recovery of Mo(VI) with Cyphos IL 102,[46] Pd(II) extraction with Cyphos IL 101 and Cyphos IL 104,[47−49] separation of Fe(III) from Ni(II) with Cyphos IL 101,[50] U(VI) extraction with Cyphos IL 101,[51] extraction of Eu(III) and other rare-earth elements with trihexyl(tetradecyl)phosphonium N,N,N′,N′-tetra(2-ethyl-hexyl)malonate, [P66614][MA], diluted in trihexyl(tetradecyl)phosphonium nitrate, [P66614][NO3],[52] Rh(III) and Ru(III) extraction with Cyphos IL 101 and Cyphos IL 104,[53] and extraction of Pt(IV), Pd(II), and Rh(III) with trioctyl(dodecyl)phosphonium chloride, [P88812][Cl].[54]

Regarding the use of ILs for the Co(II) extraction, there are several studies: Co(II) extraction with Cyphos IL 101,[1,55−57] Cyphos IL 102,[56] Cyphos IL 104,[1] Cyphos IL 167,[56] trihexyl(tetradecyl)phosphonium dicyanamide, [P66614][DCA],[58] tetraoctylphosphonium bromide, [P8888][Br],[56] and tetraoctylphosphonium oleate, [P8888][Oleate].[59] These investigations basically show the effectiveness of ILs to extract mainly Co(II) and study the dependence of variables such as IL concentration, pH, and acid concentration in the aqueous solution.

This paper aimed to study the cobalt extraction from sulfate/chloride media by solvent extraction with different ILs with trioctyl(alkyl)phosphonium cation, P888+, for the first time, as a contribution to knowledge about the use of quaternary phosphonium salts in metal extraction. They are compared with commercial ILs such as tricaprylmethylammonium chloride, A336, and trihexyl(tetradecyl)phosphonium chloride, [P66614][Cl].

Materials and Methods

Chemicals

For the preparation of aqueous solutions, the following salts were purchased from Merck: CoSO4·7H2O, NiSO4·6H2O, and NaCl. Aqueous solutions at different concentrations of chloride and sulfuric acid were prepared in Milli-Q water (Milli-Q Direct). The concentration of chloride in the aqueous solutions was achieved by adding NaCl or HCl. The A336 and [P66614][Cl] ILs were purchased from Sigma-Aldrich. Toluene purchased from Merck was used for the dilution of the ILs.

Experimental Apparatus and Procedures

The hydrophobic [P888][Cl] ILs were studied in cobalt and nickel extraction. Their structure is shown in Figure . They were synthesized by the addition of trioctylphosphine to chloroalkanes according to the proposal by Bradaric et al.[60] A description of the synthesis is presented in the Supporting Information file.

Figure 1

Structure of trioctyl(alkyl)phosphonium chloride ILs, where n = 5, 8, 14, and 16 were investigated.

Extraction and Stripping Processes

The extraction experiments were carried out in a 50 mL double-jacketed glass reactor with a hot water flow from a PolyScience thermostatic bath. The water flow allowed maintaining a constant temperature during the contact of both phases. The reactor was placed on a plate with magnetic stirring. The experiments were performed at 1200 rpm and 25 °C by mixing 1 mL of pure IL and 2 mL of aqueous solution. Phase separation which took less than 2 min to complete was carried out in glass separatory funnels. The determination coefficient determination and extraction percentage were calculated according to eqs and 2, respectively.where [Mz+] is the concentration of the metal ion in the aqueous phase, A, and the IL. The subscripts “in” and “eq” refer to the initial and equilibrium concentration of the metal ion, respectively. The metal concentration in the IL phase was calculated as the difference between the initial and equilibrium concentration in the aqueous phase. The stripping was carried out in the same reactor. The loaded IL (1 mL) from 3 M HCl and 1 g/L Co(II) aqueous solution was mixed with 2 mL of different aqueous solutions (pure H2O, 0.1 M and 0.2 M HCl, and 0.2 M H2SO4) at 25 °C within 20 min. The stripping percentage was calculated according to eq .where [Mz+] is the concentration of the metal ion in the stripping phase, st, and the IL and V is the volume of the stripping phase, st, and the IL.

Chemical Composition of the Aqueous Solution Measurement

Co(II) and Ni(II) concentrations in the aqueous solutions were measured by mean of an Agilent 200 series AA spectrometer. The metal concentration in the aqueous solution was measured before and after contact with the ILs. For solutions with only cobalt content, its concentration was measured by volumetric titration using ethylenediaminetetraacetic acid (EDTA). An aliquot of 1.6 mL of aqueous solution was taken and diluted to 10 mL using deionized water. Two aliquots of 4 mL were taken and titrated with EDTA using murexide as indicator. This method was previously standardized using a 1000 mg/L Co(II) standard solution where a maximum relative deviation of 1% was obtained.

Viscosity Measurement

The viscosity was measured with a ViscoSystem AVS 370 viscometer controlled by the WinVisco 370 software using # 200 and 450 Cannon-Fenske Routine Capillaries. The viscosity measurements were made by the pressure method. The parameters included in the software were a preheating time of 15 min, five measurements, and a maximum relative deviation of 1%.

UV/Vis Spectroscopy

The UV/vis OPTIZEN POP spectrophotometer was used using optical polystyrene cuvettes (Deltalab) with an optical path (metric) of 10 mm, having a capacity of 1.5 mL. The UV/vis absorption spectra were recorded in a range of 350–750 nm. As a blank, ultrapure water and pure ILs were used.

Results and Discussion

In most industrial processes for obtaining nickel and cobalt, aqueous solutions of sulfuric acid are used as a leaching agent. This is mainly seen when nickel and cobalt are primarily obtained from the ores of nickel sulfides and nickel laterites. Also, studies with phosphonium salts have shown that it is possible to extract Co(II) ions from chloride media. With this in mind, [P888][Cl] ILs were studied and compared with A336 and [P66614][Cl] ILs in the Co(II) extraction from aqueous solutions of sulfate media with the addition of chloride ions. The selectivity for the Co(II) extraction with respect to Ni(II), time, concentration of reagents, extraction mechanism, and temperature was evaluated.

Selectivity in the Extraction of Cobalt versus Nickel

Co(II) and Ni(II) extraction experiments were performed in an aqueous solution of 1 g/L of Co(II) and 1 g/L of Ni(II), 1 M H2SO4 (0.5 M for A336), and 2 M HCl and 5 M HCl. The contact time of the phases was 20 min. The Ni(II) extraction was negligible for all ILs. It will be shown that the ILs studied extract the metallic species through an anion exchange, and as Ni(II) does not form anionic complexes in sulfate and chloride media, it is not possible to extract it with the ILs studied here. Table shows that the most effective IL for the extraction of Co(II) is [P8888][Cl], while A336 is the least effective. However, all the ILs were studied in order to evaluate the effect of the parameters considered on cobalt extraction.

Table 1

Co(II) and Ni(II) Extraction Percentage with Several ILsa

	extraction (%)
	[P8885][Cl]		[P8888][Cl]		[P88814][Cl]		[P88816][Cl]		A336
	Co(II)	Ni(II)	Co(II)	Ni(II)	Co(II)	Ni(II)	Co(II)	Ni(II)	Co(II)	Ni(II)
2 M HCl	26.7	∼0	71.2	∼0	57.6	∼0	50.0	∼0	5.0	∼0
5 M HCl	89.8		98.3		96.6		95.8		78.8

Aqueous phase = 1 M H2SO4 (0.5 M for A336), 1 g/L of Co(II), 1 g/L of Ni(II). Equilibration time = 20 min, temperature = 25 °C, A/O ratio = 2:1.

Table shows that pure water is the most effective stripping phase. The stripping percentage decreases with increasing chloride concentration in the stripping phase. The discharge of H+ ions affects the discharge of Co(II). Cobalt stripping in sulfuric acid solution is less favorable in comparison with that in the other stripping solutions.

Table 2

Stripping Percentagesa

	stripping solution
	H2O	HCl (0.1 M)	HCl (0.2 M)	H2SO4 (0.2 M)
S (%)	100	99.4	97.7	93.6

Stripping phase = 2 mL, loaded IL = 1 mL. Contact time = 20 min, temperature = 25 °C.

Effect of Phase Contact Time

Co(II) extraction experiments were carried out with [P88814][Cl] and A336 at different contact times. An aqueous solution of 0.5 M H2SO4, 3 M HCl, and 1 g/L of Co(II) was used. The results are shown in Figure . For [P88814][Cl], the extraction increases rapidly, having 50% extraction after 1 min of contact and 70% after 2 min, reaching equilibrium at 10 min of contact of both phases. The extraction kinetics is slower for A336 reaching equilibrium after 20 min. The constant rate was determined through the linear adjustment of the initial extraction stage according to eq .where [Co2+]A,in and [Co2+]A,f refer to the Co(II) concentrations (g/L) in the aqueous phase before and after the extraction, respectively, k (min–1) is the rate constant, and t (min) is the contact phase time. A constant rate of 0.1043 min–1 was calculated for A336, while for [P88814][Cl], it was 0.618 min–1. [P88814][Cl] shows a cobalt extraction rate considerably higher than A336. In addition, [P88814][Cl] had a viscosity of 49.79 mm2/s at 25 °C, much lower compared to the value of 1152.86 mm2/s for A336. Hence, it is concluded that the difference in viscosity between the two ILs is reflected in the extraction rate. The viscosity of ILs can affect the diffusion rate.

Figure 2

Effects of contact time on Co(II) extraction with the [P88814][Cl] and A336 ILs. Aqueous phase = 0.5 M H2SO4, 3 M HCl, and 1 g/L of Co(II). Temperature = 25 °C, A/O ratio = 2:1.

Effect of Chloride Concentration

Extraction tests were carried out for different concentrations of chloride, with an initial concentration of 1 g/L of Co(II) and 0.5 M H2SO4. It is observed in Figure how the Co(II) extraction depends strongly on the initial concentration of chloride in the aqueous solution for all ILs studied. The Co(II) extraction is quite similar with the [P888][Cl] IL extraction except that n = 5. The extraction with the A336 IL is the lowest. The Co(II) extraction with the [P8888][Cl] IL seems to be slightly higher at low chloride concentrations. The [P8888][Cl] IL is an interesting option to develop the Co(II) extraction process, while A336 is not efficient for Co(II) extraction because it needs high chloride acid concentrations.

Figure 3

Co(II) extraction percentage as a function of the initial concentration of HCl in the aqueous solution. Aqueous phase = 0.5 M H2SO4 and 1 g/L of Co(II). Equilibration time = 20 min, temperature = 25 °C, A/O ratio = 2:1.

Effect of Sulfuric Acid Concentration

The extraction of Co(II) from aqueous solutions of 1 M HCl, 3 M HCl, and 3 M NaCl at different concentrations of H2SO4 and 1 g/L of Co(II) was studied. The extraction of Co(II) was done with the [P88814][Cl] IL and compared with A336.

In Figure , it is observed that there is a linear dependence between the Co(II) extraction and the concentration of H2SO4. For the extraction with [P88814][Cl], the tendency to increase the Co(II) extraction with the increase in the concentration of H2SO4 is similar at 1 M HCl, 3 M HCl, and 3 M NaCl media. A linear dependence between the concentration of H2SO4 and Co(II) extraction from a 3 M HCl medium with the A336 IL was also found. The dependence of H2SO4 is much more significant to the Co(II) extraction with the A336 IL compared to that with [P88814][Cl]. Therefore, a lower concentration of H2SO4 is required for ionic phosphonium liquids.

Figure 4

Effect of the initial concentration of H2SO4 on the Co(II) extraction with [P88814][Cl] (3 M NaCl, 3 M HCl, and 1 M HCl) and A336 (3 M HCl). Aqueous phase = 1 g/L of Co(II). Equilibration time = 20 min, temperature = 25 °C, A/O ratio = 2:1.

Extraction Mechanism

The UV/vis absorption spectra shown in Figure for the [P88814][Cl] IL after the Co(II) extraction from NaCl, HCl, and H2SO4 media exhibit an absorption band between 600 and 750 nm, which is characteristic of the CoCl42– complex.[61] This suggests that the ILs solubilize the CoCl42– complex and release two chloride ions to the aqueous phase in order to maintain the neutrality. This seems to be true regardless of the matrix of the aqueous solution. One similar spectrum was obtained for the A336 IL after the Co(II) extraction from HCl media. Therefore, the Co(II) extraction begins in the aqueous phase with the formation of the CoCl42– complex, which predominates as the chloride concentration increases.[62]

Figure 5

UV/vis absorption spectra for the loaded [P88814][Cl] from aqueous solutions of different matrices (NaCl, HCl, H2SO4) and the loaded A336 from HCl.

The reaction of the Co(II) extraction represented by an anion exchange according to eq is proposed. For each mole of the CoCl42– complex solubilized in the IL phase, 2 mol of chloride ions is exchanged. The formation of the CoCl42– complex is favored by the presence of ions H+ and HSO4–, which allow dehydration of the Co(II) ions.[63]

The stoichiometry in 5 was determined using the slope analysis method. Neglecting the effects of nonlinearity, we have an approximation to the equilibrium constant given by eq .

Taking into account that the distribution coefficient is given by , replacing in eq and taking logarithm on both sides, eq is obtained.

The [P8888][Cl] IL was diluted in toluene, and different concentrations were obtained. Co(II) extraction experiments were carried out in 3 M NaCl and 4 M NaCl solution. The initial concentration of Co(II) was 0.5 g/L. The slope analysis from Figure shows straight lines, as predicted by eq . Slope values of 1.75 and 1.79 were obtained for chloride concentrations of 3 and 4 M, respectively. These values are close to 2, which confirm that the loaded IL phase contains about 2 mol of IL per mole of extracted cobalt(II) ion according to eq and also takes into account the UV/vis results. For the plot of log D versus log[A336], a slope of 1.81 was obtained in 4 M NaCl media (Figure S1). This suggests that the extraction mechanism is similar to that found for phosphonium ILs. In addition, the UV/vis spectra (Figure ) for the charged A336 IL show the presence of the CoCl42– complex.

Figure 6

Graph of log D vs log([P8888][Cl]) for slope analysis in the Co(II) extraction from 3 M NaCl and 4 M NaCl aqueous solutions. Aqueous phase = 0.5 g/L Co(II). Equilibration time = 20 min, temperature = 25 °C, A/O ratio = 2:1.

These experiments were repeated in 4 M HCl media. A slope of 0.98 was obtained, which is very far from 2 (Figure S2). In this case, the presence of the H+ ions released by the HCl modifies the extraction mechanism. A coextraction of HCl is proposed according to the addition reaction given by eq . A slope of 0.79 was obtained for the plot of log D versus log(A336) when the Co(II) extraction was made from a 4 M solution of HCl (Figure S1).

The Co(II) extraction with the [P88814][Cl] IL from 5 M NaCl and 5 M HCl aqueous solutions, both with 1 g/L of Co(II) and A/O = 1, was done. For the NaCl media, a Co(II) extraction of 99.5% was obtained, while for the HCl media, the Co(II) extraction was 98.1%. The [P88814][Cl] IL loaded came in contact with deionized water (pH ≈ 6) at a 1/1 ratio. For the [P88814][Cl] IL loaded from the NaCl media, the discharge percentage was 98.3% with a final pH of the discharge solution of 4.3. The decrease in pH of the water is due to the presence of chloride ions which alter the activity of the water. For the [P88814][Cl] loaded from the HCl media, the discharge percentage was 61.9% with a final pH of the discharge solution of 1.4. The decrease in pH is due to the discharge of coextracted H+ ions. The discharge of H+ ions affects the Co(II) stripping, being only 61.9% compared to 98.3% for the case that H+ is not present.

Effect of System Temperature

The Co(II) concentration in the aqueous phase was 1 g/L (0.017 M); therefore, it can be said that chloride ions exist in large excess compared to Co(II) ions, so it is considered that the total concentration of chloride ions is approximately constant. At equilibrium, the change in Gibbs free energy associated with the extraction reaction at a given temperature can be determined according to eq .where R is the universal gas constant and T is the absolute temperature at which the extraction process is carried out. The change of free energy is related to the changes of enthalpy and entropy by eq .

Equation is derived from eqs and 10.

Assuming ΔH° and ΔS° being very weak with the change in temperature, their values were calculated from the slope and intercepts of the lines of the van’t Hoff plot presented in Figure . The Co(II) extraction was carried out in an aqueous solution of 3 M NaCl, 0.5 M H2SO4, and 1 g/L of Co(II). The results are summarized in Table . An increase in the extraction of Co(II) was found with the increase in the temperature of the extraction process. Positive values for ΔH° indicate an endothermic process. The endothermic process is also due to the need to break the Co(II)–H2O bond of the cobalt hydration in the aqueous solution. This implies that the formation of the complex is thermodynamically unfavorable, which indicates that the formation of the CoCl42– complex occurs mainly at the interphase, accompanied by the dehydration of the metal center. The increase in entropy after the dehydration of the metal center drives the extraction. The high value of the extraction entropy could be attributed to the loss of an organized aggregation structure within the IL phase after the transfer of the CoCl42– complex between both phases (Quinn et al., 2013). The negative value to the Gibbs free energy change indicates that all the ILs studied are favorable for the Co(II) extraction process.

Figure 7

Graph of log D vs 1000/T for the Co(II) extraction with the [P888][Cl] ILs (n = 8, 14, and 16), [P66614][Cl], and A336. Aqueous phase = 0.5 M H2SO4, 3 M NaCl, and 1 g/L of Co(II). Equilibration time = 20 min, A/O ratio = 2:1.

Table 3

Thermodynamic Parameters for the Co(II) Extraction

IL	A336	[P8888][Cl]	[P88814][Cl]	[P88816][Cl]	[P66614][Cl]
ΔH° (kJ mol–1 K–1)	47.17	27.64	16.91	32.72	23.90
ΔS° (J mol–1 K–1)	163.96	120.69	74.11	128.73	107.38
ΔGex (kJ mol–1 K–1)	–1.72	–8.34	–5.19	–5.66	–8.12

The Co(II) extraction has a linear dependence with the process temperature (Figure S3). The increase is similar for the [P8888][Cl] and [P88814][Cl] ILs where the slope of the trend line is similar. There is an increase of 3.23% points in the Co(II) extraction when passing from 298.15 to 318.15 K for the [P8888][Cl] IL. There is a more significant temperature dependence for the [P88816][Cl] IL. In this case, the Co(II) extraction increased in 8.56% points when the system temperature increased from 298.15 to 318.15 K. The extraction with the [P66614][Cl] IL was compared. The behavior in the cobalt extraction with the temperature change is practically the same compared to the [P8888][Cl] IL. However, there is a great difference between their viscosities; the viscosity of the [P66614][Cl] IL is 7 times higher than the viscosity of the [P8888][Cl] IL at 25 °C (Table ).

Table 4

Viscosity Values at 25 °C

IL	[P8888][Cl]	[P88814][Cl]	[P88816][Cl]	[P66614][Cl]
viscosity (mm2/s)	284.35	49.79	587.97	2188.66

Conclusions

The use of phosphonium ILs and that based on ammonium was studied as an alternative to traditional organic solvents for the separation of cobalt and nickel. The ILs studied showed selectivity for the cobalt extraction from nickel. For phosphonium ILs, the equilibrium in the extraction is reached after 10 min under the experimental conditions studied. Cobalt extractions near 98% with the trioctyl(alkyl)phosphonium chloride ILs were obtained. The percentage of cobalt extraction was practically the same as that obtained with Cyphos IL 101 but much higher compared to A336, both commercially available. The symmetry of the cation could affect cobalt extraction, but there is no enough evidence about it. Additional studies are necessary to establish this relationship.

The Co(II) extraction is strongly dependent on the initial chloride concentration in the aqueous solution and slightly dependent on the sulfuric acid concentration. The extraction of cobalt occurs through an anion exchange mechanism favored by the presence of hydrogen ions and sulfuric acid, which allow the break of the cobalt hydration bond. The HCl coextraction is carried out in an acidic medium and has a negative effect on stripping. The cobalt extraction with the phosphonium ILs depends slightly on the temperature, resulting in an endothermic process. For the A336 IL, there is a strong dependence on the temperature of the system, which may be due to its high viscosity.

Cobalt can be extracted from an aqueous solution in sulfate media adding sodium chloride to the solution. Cobalt extraction with phosphonium ILs has additional advantages as the process can be carried out with sodium chloride instead of hydrochloric acid, which translates into easy handling, lower costs in reagents and equipment, and risk reduction. The process is facilitated by the cobalt stripping into water without adding acids, which also reduced the cost.