Highly selective separation of individual platinum group metals (Pd, Pt, Rh) from acidic chloride media using phosphonium-based ionic liquid in aromatic diluent

Abstract

An effective solvent extraction-based method has been developed to recover individual platinum group metals (PGMs) (i.e. Pd, Pt and Rh) with high purity from acidic chloride media using phosphonium-based ionic liquid Cyphos IL 101 ([P₆₆₆₁₄]⁺Cl⁻) diluted in xylene used as a model of industrial diluents such as SOLVESSO 150. The system showed the selective co-extraction of Pd(II) and Pt(IV) from a feed solution containing 100 mg L⁻¹ Pt(IV), 55 mg L⁻¹ Pd(II), and 25 mg L⁻¹ Rh(III) in HCl via an anion exchange mechanism, leaving Rh(III) in the raffinate. A McCabe–Thiele diagram demonstrated the quantitative extraction of Pd(II) and Pt(IV) with 0.6 g L⁻¹ [P₆₆₆₁₄]⁺Cl⁻ with two counter-current stages at an organic/aqueous (O/A) phase volume ratio of 3/2 at 298 K. The stoichiometry of complexes formed in the organic phases was determined using Job's method and further characterized by ¹H, ¹³C, ³¹P NMR and UV-vis spectroscopy. More importantly, a two-step stripping strategy was successfully adapted to treat the loaded organic phases with low aggressive media NaSCN and thiourea/HCl for the selective recovery of Pt(IV) and Pd(II), respectively, with high purity >99.9% (w). Three counter-current stages are required for the quantitative stripping of Pt(IV) with 0.1 mol L⁻¹ NaSCN at an O/A ratio of 3, whereas the total stripping of Pd(II) with 0.01 mol L⁻¹ acidic thiourea requires only one stage at an O/A ratio of 5. A series of extraction-stripping processes of up to 5 cycles indicated a possible recirculation of regenerated solvent without loss of performance. Simulated experiments in continuous modes suggest that phosphonium-based ionic liquid such as Cyphos IL 101 can be advantageously used as extractant to recover individually Pd, Pt and Rh from acidic chloride feed solutions as those encountered in the PGM's recycling industry.

---

1 Introduction

The platinum group metals (PGMs) palladium, platinum, and rhodium represent the key materials for various applications, including automotive exhaust gas treatment, catalysis, etc. Because there are currently no adequate alternatives, the importance is steadily increasing, whereas the supply of these metals is experiencing a shortage. For instance, it was reported by Johnson Matthey Catalysts that the gross demands for Pt, Pd and Rh in the field of autocatalysts in 2012 were 97.2 t (over 44% of the gross Pt demand), 216.8 t (72% of the gross Pd demand), and 24.9 t (78% of the gross Rh demand), while the recycled Pt, Pd and Rh from auto-exhaust catalysts accounted for 39.7 t, 57.9 t and 8.7 t, respectively.¹ Thus, from an economic and an environmental perspective, the recycling process is essential to satisfy the demand for PGMs and for reducing the effect of the balance problem and environmental issues related to mining.²

The separation and purification processes of PGMs face a continual challenge due to their refractory chemical properties and the formation of many chemical species in chloride media, which is typically chosen for their hydrometallurgical recycling. Traditional refining of PGMs involved a series of precipitation-dissolution steps that are no longer considered efficient in terms of the degree of separation, the yields and the complexity of the operations. Thus, a considerable amount of research and development has been conducted over the last decades to replace traditional refining practices with solvent extraction processing technology. In general, solvent extraction-based refining processes for PGMs provide many advantages over the traditional precipitation methods, primarily including lower inventory of metals with the high value due to reduced overall processing time, improved primary yields, reduced process recycles, flexibility and versatility, and the capability for continuous operation and process control. Thus, their usage in industry is expected to become a standard practice.³

The liquid–liquid extraction of PGMs from acidic chloride media has been extensively reported using molecular solvents, such as alkyl derivatives of 8-hydroxyquinoline,^4–7 hydroxyoximes,⁸ organo-phosphorus extractants,⁹ hydrophobic amines,^10,11 and quaternary ammonium salts.^12,13 More recently, room-temperature ionic liquids (referred to herein as ionic liquids) have been widely considered in hydrometallurgy towards innovative and sustainable metal extraction processes.^14–21 Ionic liquids are defined as organic salts that consist entirely of ions and have melting point below 100 °C.^22–24 Due their nature, i.e., typically [R]⁺[X]⁻ and the huge diversity of possible choices regarding their cationic and anionic moieties, the ionic liquids can exhibit amazing properties. For instance, tri(hexyl)tetradecylphosphonium chloride (Cyphos IL 101 or [P₆₆₆₁₄]⁺Cl⁻), as an extractant, behaves as a liquid anion exchanger, as its chloride counter-ion can be easily exchanged with aqueous anionic species such as [PdCl₄]²⁻, [PtCl₆]²⁻, etc.^25–27 On the other hand, imidazolium ionic liquids such as [C_nmim]⁺[NTf₂]⁻ can extract aqueous cationic species as [Na·DCH18C6]⁺ (DCH18C6 = dicyclohexano-18-crown-6) by exchange with its [C_nmim]⁺ which is released in aqueous phase.^28–31

In practice, ionic liquids with hydrophobic cationic moieties associated with simple anions such as Cl⁻, NO₃⁻, etc., as for instance tri(hexyl)tetradecylphosphonium chloride (Cyphos IL 101 or [P₆₆₆₁₄]⁺Cl⁻), are preferred in liquid–liquid extraction. However, the use of ionic liquids as green extractant in the pure state for metal extraction purposes remain limited due to their high viscosity, which has a negative effect on the mass transfer and/or the extraction kinetics and makes their implementation in a continuous extraction process challenging.^32–34 For example, the separation of transition metals from rare earth elements with undiluted ionic liquid [P₆₆₆₁₄]⁺Cl⁻ has been reported.³³ Although the metals could be separated very efficiently, this separation process has the disadvantages mentioned above. Another disadvantage of this ionic liquid extraction system is the difficult stripping of iron from the loaded organic phase.^33,35,36 More recently, [P₆₆₆₁₄]⁺Cl⁻ has been also used for preparing new extractant impregnated resins (EIRs), which are alternative systems that combine the advantages of both ionic liquids and resins. The immobilization of ionic liquids on resins used for PGMs extraction has been designed using impregnation procedures, or encapsulation processes.^37–39 Unfortunately, the main drawbacks of EIRs are the leakage of the ionic liquid from the resins/capsules as well as the partial hydrolysis of the capsules under highly acidic condition (>5.0 mol L⁻¹ HCl), which lead to a steady loss of sorptive capacity towards PGMs after several cycles of application. Furthermore, the presence of thiourea in the porous after PGMs desorption may limit the efficiency of the sorbent for next PGMs sorption cycles.

Therefore, most authors have exploited phosphonium-based ionic liquids by considering them as new extractants and/or ion-exchangers to be used in the classical approach, i.e., diluted in molecular solvents to reduce the viscosity of the organic phase.^{25–27,40,41} Cieszynska et al. studied the extractive recovery of Pd(II) from single-element hydrochloric solution with Cyphos IL 101 diluted in toluene. The extraction stoichiometry of Pd(II) at different acidity was determined with slope analysis method. It was assumed that Pd(II) extraction with [P₆₆₆₁₄]⁺Cl⁻ proceeds according to an anion-exchange mechanism.^25,26 The extraction of Pd(II) over metal impurities (Ni(II), Cu(II), Pb(II), Fe(III), Rh(III) and Ru(III)) in HCl with phosphonium extractants diluted in toluene has been also reported by the same authors.²⁷ Spot tests revealed that Pt(IV) and Pd(II) were mostly co-extracted by Cyphos IL 101 and that Pd(II) was poorly separated from Pt(IV) at the stripping stage with 0.5 mol L⁻¹ NH₄OH as a stripping agent. In fact, the use of NH₄OH for stripping purpose appears fairly limited in selective recovery of PGMs due to the drawbacks of third phase formation, solvent decomposition, metal impurities (Pb(II), Fe(II)) co-precipitation at high pH, and stable complexation with other metal impurities (Cu(II), Ni(II)).

The present study addresses an efficient and highly pure PGMs recovery (i.e., Pd, Pt and Rh) (>99.9%) in acidic chloride media using Cyphos IL 101 as extractant and xylene as a model of industrial diluents such as SOLVESSO 150. The strategy of two-step selective stripping with the use of NaSCN and thiourea/HCl as stripping agents for Pt(IV) and Pd(II), respectively, has been successfully adapted to the case of co-extraction of Pt(IV) and Pd(II) with Cyphos IL 101. Furthermore, the added values of the paper include: (i) optimization for both extraction and selective stripping of PGMs, (ii) characterization of the extracted species with Job's continuous variation method and spectroscopies (¹H, ¹³C, and ³¹P NMR, and UV-vis), (iii) regeneration of the organic phases up to 5 extraction-stripping cycles without loss in performance, (iv) simulation in continuous mode suggesting that Cyphos IL 101 can be advantageously used as extractant to recover individually Pd, Pt and Rh from acidic chloride feed solutions as those encountered in the PGM's recycling industry.

2 Experimental

2.1 Chemicals and reagents

Phosphonium-based ionic liquid Cyphos IL 101, [P₆₆₆₁₄]⁺Cl⁻ (>97.7%, m.p. 223 K), from Cytec Industries Inc. (Canada) was used without further purification. Due to the limited solubility of [P₆₆₆₁₄⁺]Cl⁻ in kerosene, extra pure xylene (>99%, b.p.p 412 K, Junsei Chemical Co., Japan) was chosen as a model of industrial diluents such as SOLVESSO 150.

Considering the most frequently associated metals in PGM's recycling solutions, as those issued from the leachate of spent automotive catalysts, an aqueous solution containing 100 mg L⁻¹ (i.e., 0.51 mmol L⁻¹) Pt(IV), 55 mg L⁻¹ (i.e., 0.51 mmol L⁻¹) Pd(II), and 25 mg L⁻¹ (i.e., 0.24 mmol L⁻¹) Rh(III) was prepared by dissolving the required amount of H₂PtCl₆·6H₂O, PdCl₂, and RhCl₃·xH₂O (Sigma-Aldrich) in HCl solutions, respectively. Aluminum and magnesium, which are commonly present as cationic species in the acidic leaching solution of spent autocatalysts, were not taken into account in this study. NaCl (>99.9%, Junsei Chemical Co., Japan) was used to adjust the chloride concentration in feed solutions. All of the other chemicals used were of analytical grade (Junsei Chemical Co., Ltd).

2.2 Apparatus and characterization

The concentration of metal ions in the aqueous phase was measured with an inductively coupled plasma atomic emission spectrometer (ICP-AES, iCAP 6000 series, Thermo Scientific). The concentration of chloride ions in the aqueous solutions was determined using the Mohr method (direct precipitation titration with AgNO₃ using potassium dichromate as an indicator). The water content of ionic liquids phase was analysed by Karl-Fischer titration technique (model Schott Instruments Titro-Line KF) and it was found that the water content to be less than 300 mg L⁻¹ for all samples, which is close to the solubility of water in xylene. Ultraviolet-visible (UV-vis) spectroscopy was measured on a Shimadzu UV-1601PC spectrophotometer in the range between 400 and 600 nm at room temperature in a 1 cm quartz cuvette. The ¹H, ¹³C and ³¹P NMR spectra of neat compounds were recorded on a Bruker ARX-500 spectrometer. The ionic liquid phases fully loaded with PGMs were concentrated to dryness under reduced pressure in a rotary evaporator at 343 K. The residues were diluted in CDCl₃ before NMR analysis. In the supplement section, the chemical shifts are reported in ppm downfield from the external references (CH₃)₄Si (¹H and ¹³C NMR) and H₃PO₄ (³¹P NMR). Fig. 1 shows the schematic structures of ionic liquid phases before and after being loaded with PGMs.

Fig. 1 Schematic structures of fresh ionic liquid [P₆₆₆₁₄]⁺Cl⁻ (1), Pd(II) loaded ionic liquid [P₆₆₆₁₄]⁺[PdCl_n]⁽ⁿ⁻²⁾⁻ (2), and Pt(IV) loaded ionic liquid [P₆₆₆₁₄]⁺[PtCl_m]^(m−4)− (3).

2.3 Extraction procedures

Batch tests were conducted using 10 mL vials. Unless otherwise stated, equal volumes of aqueous and organic phases were shaken on a Recipro Shaker (RS-1, Jeio Tech, Korea) at room temperature (298 ± 1 K) for 10 min to ensure that equilibrium had been attained. After phase disengagement, the aqueous phase was properly diluted, and the metal content was determined by ICP-AES. The metal concentration in the organic phase was calculated from the difference between the metal concentration in the aqueous phase before and after extraction.

The extraction efficiency (E%) was calculated using eqn (1):

where V_aq and V_org refer to the volumes of aqueous and organic phases, respectively. The distribution ratio, D, was defined as the ratio of metal present in the organic phase after extraction [M]_org and the raffinate [M]_aq:

The PGMs extraction distribution isotherms have been obtained with fresh or recycled organic [P₆₆₆₁₄]⁺Cl⁻ by equilibrating a feed solution containing 100 mg L⁻¹ Pt(IV), 55 mg L⁻¹ Pd(II), and 25 mg L⁻¹ Rh(III) in 0.1 mol L⁻¹ HCl with the organic phase 0.6 g L⁻¹ [P₆₆₆₁₄]⁺Cl⁻ at different phase ratios from 1/5 to 5/1 while maintaining the total volume of the phases constant.

2.4 Stripping procedures

For stripping studies, loaded organic phases containing Pd(II) and Pt(IV) were individually mixed with various aqueous strip solutions (NH₄OH, Na₂S₂O₃, Na₂S₂O₄, NaSCN, (NH₄)₂CS) at a unit phase volume ratio, unless otherwise stated. The stripping experiments of PGMs were performed at room temperature (298 ± 1 K) for 10 min, which was sufficient to attain equilibrium. Strip liquors were subsequently separated from the organic phase, and metal concentrations were analysed by ICP-AES. The percentage stripping (% S) was calculated as below:where V_st, V_lorg, [M]_st, and [M]_lorg are the volumes and metal concentrations in the strip liquor and the loaded organic phase after stripping, respectively. The stripped organic phases were reused in further cycles of the extraction-stripping process. All of the experiments were performed in triplicate and experimental error did not exceed 5%.

3 Results and discussion

3.1 Influence of various parameters on the PGMs extraction

3.1.1 Effect of ionic liquid concentration. Bach tests were conducted to determine the influence of the [P₆₆₆₁₄]⁺Cl⁻ concentration on the extraction behavior of PGMs. The initial ionic liquid concentrations were varied in the range of 0.1 to 4.0 g L⁻¹ (i.e., 0.19 to 7.7 mmol L⁻¹), while other parameters were kept constant at an O/A phase volume ratio of 1 and 0.1 mol L⁻¹ HCl. As seen in Fig. 2, Pd(II) and Pt(IV) are preferentially extracted into the organic phase, while Rh(III) remains in the aqueous phase. The extraction efficiencies of Pd(II) and Pt(IV) sharply increase as the [P₆₆₆₁₄]⁺Cl⁻ concentration increases up to 2.0 g L⁻¹ at which the quantitative extraction of both metals is achieved. In detail, the extraction efficiency rises from 21.9% to 99.9% for Pd(II) and from 8.51% to 98.1% for Pt(IV) as the [P₆₆₆₁₄]⁺Cl⁻ concentration varies from 0.2 to 2.0 g L⁻¹.

Fig. 2 Effect of [P₆₆₆₁₄]⁺Cl⁻ concentration on the PGMs extraction. Organic phase: 0.1–4.0 g L⁻¹ [P₆₆₆₁₄]⁺Cl⁻; aqueous phase: 100 mg L⁻¹ Pt(IV), 55 mg L⁻¹ Pd(II), 25 mg L⁻¹ Rh(III), 0.1 mol L⁻¹ HCl; O/A = 1; t = 10 min; T = 298 K.

As reported in the literature, the extraction of metal species by [P₆₆₆₁₄]⁺Cl⁻ proceeds through an anion-exchange mechanism, as schematically exemplified below:

where the overbars denote the species in the organic phase, and [A]ⁿ⁻ refers to the anionic metal species to be extracted. The extraction efficiency of Rh(III) is less than 2%, in agreement with the fact that only the species [RhCl₂(H₂O)₄]⁺ and [RhCl₃(H₂O)₃], which cannot be involved in an anion exchange mechanism such as in eqn (4), are observed in HCl solutions of concentrations ranging from 0 to 1.0 mol L⁻¹.^42–44 Note that neither a stable emulsion nor a third phase formation was observed in this series of experiments. For further studies, an organic phase containing 0.6 g L⁻¹ [P₆₆₆₁₄]⁺Cl⁻ (i.e., 1.16 mmol L⁻¹) was selected.

3.1.2 Effect of hydrogen ion concentration. The effect of hydrogen ion concentration on the extraction of PGMs was investigated in the range of 0.1 to 4.0 mol L⁻¹ H⁺ while maintaining the ionic strength at 4.0 mol L⁻¹ (H, Na)Cl in the aqueous phase. The initial concentrations of PGMs in the aqueous solutions and [P₆₆₆₁₄]⁺Cl⁻ in the organic solutions were kept constant in the experiments. Fig. 3a exhibits that the increase in hydrogen ion concentration in the investigated range at a fixed chloride concentration had virtually no influence on the PGMs extraction under the present experimental conditions. The extraction efficiencies of Pd(II) and Pt(IV) remain constant at 36.1% and 53.5%, respectively.

Fig. 3 Effect of H⁺ and total chloride concentrations on the PGMs extraction. Organic phase: 0.6 g L⁻¹ (i.e., 1.16 mmol L⁻¹) [P₆₆₆₁₄]⁺Cl⁻; aqueous phase: 100 mg L⁻¹ Pt(IV), 55 mg L⁻¹ Pd(II), 25 mg L⁻¹ Rh(III), (a) 0.1–4.0 mol L⁻¹ H⁺, 4.0 mol L⁻¹ Cl⁻, 0–3.9 mol L⁻¹ Na⁺; (b) 0.1–4.0 mol L⁻¹ HCl; (c) 0.1 mol L⁻¹ H⁺, 0.1–4.0 mol L⁻¹ Cl⁻, 0–3.9 mol L⁻¹ Na⁺; O/A = 1; t = 10 min; T = 298 K.

3.1.3 Effect of total chloride ion concentration. Two sets of experiments were performed to study the extraction behaviors of PGMs with [P₆₆₆₁₄]⁺Cl⁻ at various chloride concentrations. First, aqueous solutions containing 0.1 to 4.0 mol L⁻¹ HCl were mixed with equal volumes of the same organic phase of 0.6 g L⁻¹ [P₆₆₆₁₄]⁺Cl⁻. Fig. 3b shows a significant decreasing trend in Pd(II) extraction with the increase in the HCl concentration. For example, the Pd(II) extraction efficiency decreases from 77.5% to 35.8%, corresponding to a significant decrease in the distribution ratio (D) from 2.68 to 0.56 as the HCl concentration increases from 0.1 mol L⁻¹ to 4.0 mol L⁻¹. Meanwhile, the Pt(IV) extraction efficiency from HCl solutions remains constant at approximately 62.0% when the HCl concentration increases from 0.1 to 2.0 mol L⁻¹. These observations are consistent with these described by Cieszynska and Wisniewski.^25–27 A further increase in the HCl concentration in the range of 2.0 to 4.0 mol L⁻¹ results in a slight decrease in the Pt(IV) extraction efficiency from 61.8 to 50.7%, respectively. In addition to the speciation of PGM's, which may also play a role, the decrease in the Pd(II) and Pt(IV) extraction reported above when the concentration of HCl is increased may be a result of the negative mass effect of chloride ions, as illustrated in eqn (4) (at constant speciation of [A]ⁿ⁻) and to a lesser degree, to the co-extraction of HCl according to eqn (5) at a high HCl concentration, as reported in the case of tri-n-hexyl amine hydrochloride.⁴⁵

It appears that rhodium(III) is not extracted by [P₆₆₆₁₄]⁺Cl⁻ between 0.1 and 4.0 mol L⁻¹ HCl, although anionic species such as [RhCl₄(H₂O)₂]⁻, [RhCl₅(H₂O)]²⁻, and [Rh₂Cl₉]³⁻ exist in aqueous phase above 2.0 mol L⁻¹ HCl.⁴² In fact, Rh(III) forms octahedral complexes with anions such as halides and with oxygen-containing ligands. With respect to solvent extraction, the highly charged octahedral complexes, such as [RhCl₆]³⁻, [RhCl₄(H₂O)₂]⁻, and [RhCl₅(H₂O)]²⁻, are particularly difficult to extract due to steric effects.⁴⁴

Second, the effect of total chloride ion concentration on the PGMs extraction with 0.6 g L⁻¹ [P₆₆₆₁₄]⁺Cl⁻ was determined in the range of 0.1 to 4.0 mol L⁻¹ by adding the required amounts of NaCl at a given proton concentration of 0.1 mol L⁻¹. As shown in Fig. 3c, the increasing chloride concentration is responsible for a significant decrease in the Pd(II) extraction efficiency. The percentage extraction of Pd(II) markedly decreased from 72.5 to 34.4%, or in terms of distribution ratio (D), from 2.64 to 0.52, as the concentration of Cl⁻ increased from 0.1 to 4.0 mol L⁻¹, respectively. The extraction curve reported in Fig. 3c is qualitatively similar to the extraction curve shown in Fig. 3b. These observations indicate that the concentration of chloride ion has a strong influence on the yield of extracted Pd(II) with [P₆₆₆₁₄]⁺Cl⁻. Thus, the effect of HCl concentration reported in Fig. 3b is mainly due to the concomitant variation in the chloride ion concentration. The extraction behavior of Pt(IV) remains constant at approximately 60% when the Cl⁻ concentration increases from 0.1 to 2.0 mol L⁻¹. A further increase in the Cl⁻ concentration in the range of 2.0 to 4.0 mol L⁻¹ results in a slight decrease in the Pt(IV) extraction efficiency from 59.7 to 50.1%, respectively.

3.1.4 Speciation of the extracted species. Job's method of continuous variation was used to investigate the main species formed during the extraction of Pd(II) with [P₆₆₆₁₄]⁺Cl⁻.⁴⁶ For this treatment, the sum of the initial concentration of Pd(II) in the aqueous phase and the concentration of [P₆₆₆₁₄]⁺Cl⁻ in the organic phase, for a phase volume ratio of 1 was maintained constant at 1.5 mmol L⁻¹, while the molar fraction of Pd(II) was varied from 0.10 to 0.90 in each mixture. The concentration of HCl in the aqueous solutions was maintained at two different values, i.e., 0.1 mol L⁻¹ and 4.0 mol L⁻¹.

Examination of Fig. 4a shows that the plot of the concentration of Pd(II) extracted into the organic phase versus the molar fraction x = [Pd(II)]/[[Pd(II)] + [P₆₆₆₁₄]⁺Cl⁻] at constant [[Pd(II)] + [P₆₆₆₁₄]⁺Cl⁻]] and at a phase volume ratio of 1 exhibits a maximum at x = 0.50 and 0.25 when the extraction is performed a 0.1 mol L⁻¹ and 4.0 mol L⁻¹ HCl, respectively. This indicates that the extracted species exhibit 1 : 1 and 1 : 3 metal : ligand stoichiometries at 0.1 mol L⁻¹ and 4.0 mol L⁻¹ HCl, respectively.⁴⁶ Similarly, the nature of the Pt(IV) complex extracted with [P₆₆₆₁₄]⁺Cl⁻ at 0.1 mol L⁻¹ HCl was also determined using Job's method. Examination of Fig. 4b shows that the plot of the concentration of Pt(IV) extracted into the organic phase versus the molar fraction x = [Pt(IV)]/[[Pt(IV)] + [[P₆₆₆₁₄]⁺Cl⁻]] at constant [[Pt(IV)] + [[P₆₆₆₁₄]⁺Cl⁻]] and at a phase volume ratio of 1 exhibits a maximum at x = 0.34, which indicates that the extracted species exhibits a 1 : 2 metal : ligand stoichiometry.

Fig. 4 Job's plots for the PGMs extraction. (a) ; organic phase: 0.08–0.7 g L⁻¹ [P₆₆₆₁₄]⁺Cl⁻; aqueous phase: 16–144 mg L⁻¹ Pd(II), 0.1 mol L⁻¹ and 4.0 mol L⁻¹ HCl; O/A = 1; t = 10 min; T = 298 K. (b) ; organic phase: 0.08–0.7 g L⁻¹ [P₆₆₆₁₄]⁺Cl⁻; aqueous phase: 29–264 mg L⁻¹ Pt(IV), 0.1 mol L⁻¹ HCl; O/A = 1; t = 10 min; T = 298 K.

In HCl media, [PtCl₆]²⁻ is the predominant species, even at 0.1 mol L⁻¹ HCl.⁴⁷ Thus, the value x = 0.34 found in Fig. 4b indicates that the extraction of Pt(IV) with [P₆₆₆₁₄]⁺Cl⁻ occurs according to the following equation:

This finding is in agreement with slope value obtained in Fig. S1† as well as the literature concerning the extraction of Pt(IV) from HCl media by liquid anion exchangers.^47,48 Furthermore, eqn (6) is in agreement with the absence of effect of H⁺ concentration, as observed in Fig. 3a at constant ionic strength. On the other hand, from eqn (6), a significant decrease of the yield of extraction of Pt(IV) is expected when the concentration of chloride ion is increased due the negative mass effect of these ions; however, this is not the case. The Pt(IV) extraction yield remains nearly constant between 0.1 and 2.0 mol L⁻¹ Cl⁻ and only slightly decreases above 2.0 mol L⁻¹ Cl⁻ (Fig. 3b and c). The distribution data reported in Fig. 3b and c do not reflect the pure effect of the chloride ion concentration as expected, but are the results of a combination of several phenomena, including: (i) the negative mass effect of chloride ions as appearing in eqn (6), (ii) a competitive extraction of HCl as exemplified by eqn (5), (iii) an effect resulting from the change of the activity coefficients (salting-out effects) due the increase of ionic strength I from 0.1 mol L⁻¹ HCl (I = 0.1 mol L⁻¹) to 4.0 mol L⁻¹ HCl (Fig. 3b) or to 0.1 mol L⁻¹ HCl + 3.9 mol L⁻¹ NaCl (Fig. 3c) (I = 4.0 mol L⁻¹),^49–52 and possibly, (iv) a variation of the [P₆₆₆₁₄]⁺Cl⁻ concentration along the curve due a decrease in the Pd(II) extraction. In the case of the Pt(IV) extraction with [P₆₆₆₁₄]⁺Cl⁻ from chloride media under the conditions in Fig. 3b and c, all of these effects appear to roughly counterbalance themselves, except at concentration above 2.0 mol L⁻¹ HCl, which is not the case of Pd(II).

In hydrochloric acid/chloride media, Pd(II) can form four chlorocomplexes, i.e., [PdCl]⁺, PdCl₂, [PdCl₃]⁻ and [PdCl₄]²⁻. In reality, [PdCl₄]²⁻ is the predominant species, but the latter is in rapid equilibrium with the species [PdCl₃(H₂O)]⁻, which is more labile and whose relative concentration is significant at a low HCl concentration.^46,53 The values x = 0.50 and x = 0.25 found in Fig. 4a for 0.1 and 4.0 mol L⁻¹ HCl, respectively, suggest that the extraction of Pd(II) with [P₆₆₆₁₄]⁺Cl⁻ occurs according to the following equations:

- At 0.1 mol L⁻¹ HCl/Cl⁻:

or, as [PdCl₃]⁻ and [PdCl₄]²⁻ are in rapid equilibrium:

- At 4.0 mol L⁻¹ HCl/Cl⁻:

where [P₆₆₆₁₄]⁺Cl⁻ could play the role of anion exchanger and of solvating agent.

The formation of [P₆₆₆₁₄]⁺ [PdCl₃]⁻ at a low HCl concentration is in agreement with the slope value obtained in Fig. S1† as well as the findings of Cieszynska and Wisniewski.^26,27 On the other hand, the 1 : 3 metal : ligand stoichiometry observed at 4.0 mol L⁻¹ HCl is unexpected.

3.1.5 Characterization of the extracted species by ¹H and ³¹P NMR spectroscopies. Examination of the ¹H NMR spectra (see supplement section) shows the chemical shifts of the methylene protons (H_a and H_b – Fig. 1) adjacent to the phosphorus atom come into resonance at a lower frequency with the loading of PGMs into the organic phase. Indeed, the chemical shifts of protons H_a decrease from 2.6823–2.5965 to 2.5253–2.4676 ppm for Pd(II) extraction and to 2.4171–2.3623 ppm for Pt(IV) extraction, which is statistically significant as the standard deviation in chemical shifts of the solvent CDCl₃ (δ 7.2876, 7.2873, and 7.2873 ppm) used for ¹H NMR is ±0.0002 ppm. This is an evidence of the existence an interaction between the chloro-metal complexes ([PdCl_n]⁽ⁿ⁻²⁾⁻ and [PtCl₆]²⁻) loaded into the organic phase and the cationic [P₆₆₆₁₄]⁺ moiety of Cyphos IL 101. Further evidence is given by ³¹P NMR spectroscopy. In particular, the singlet peak of [–(CH₂)₄P]⁺ at 32.69 ppm is shifted to 32.39 (Δδ = 0.3 ppm) and 32.48 ppm (Δδ = 0.2 ppm) with Pd(II) and Pt(IV) extraction, respectively.⁵⁴

3.1.6 Thermodynamic parameters for the PGMs extraction. The effect of temperature on Pd(II) and Pt(IV) extraction by [P₆₆₆₁₄]⁺Cl⁻ was examined in a water-jacketed reactor. Temperature was varied over the range 288−338 K using a circulating water bath (±0.2 °C). Other parameters, meanwhile, were kept constant at 100 mg L⁻¹ Pt(IV), 55 mg L⁻¹ Pd(II), 25 mg L⁻¹ Rh(III) in 0.1 mol L⁻¹ HCl, 0.6 g L⁻¹ [P₆₆₆₁₄]⁺Cl⁻, and O/A = 1. Fig. 5a shows that increasing temperature from 288 to 338 K leads to a decrease in the percentage extraction from 72.3% to 54.7% for Pd(II), and from 60.9% to 53% for Pt(IV), respectively. This demonstrates that the reaction is exothermic.

Fig. 5 (a) Effect of temperature on the PGMs extraction; (b) plot of log K vs. 1000/T. Organic phase: 0.6 g L⁻¹ [P₆₆₆₁₄]⁺Cl⁻; aqueous phase: 100 mg L⁻¹ Pt(IV), 55 mg L⁻¹ Pd(II), 25 mg L⁻¹ Rh(III), 0.1 mol L⁻¹ HCl; O/A = 1; t = 10 min; T = 288–338 K.

The extraction equilibrium constants, K_ex-Pd(II) and K_ex-Pt(IV), associated to eqn (7) and (6) can be written as follows if it is assumed that the activity coefficients are approximately constant (γ = 1) at 0.1 mol L⁻¹ HCl:

Taking logarithm and re-arranging eqn (10) and (11), we have:

log K_ex-Pd(II) = log D_Pd(II) − log[[P₆₆₆₁₄]⁺Cl⁻] + 2 log[Cl⁻] (12)

log K_ex-Pt(IV) = log D_Pt(IV) − 2 log[[P₆₆₆₁₄]⁺Cl⁻] + 2 log[Cl⁻] (13)

The log K_ex-Pd(II) and log K_ex-Pt(IV) values for Pd(II) and Pt(IV) extraction are estimated according to eqn (12) and (13), as represented in Table 1. The standard enthalpy change, ΔH°, for the Pd(II) and Pt(IV) extraction with [P₆₆₆₁₄]⁺Cl⁻ can be evaluated using Van't Hoff equation provided that the temperature change is relative small:

where R is the universal gas constant and C is integration constant which was assumed to be constant at a particular temperature under the experimental conditions. The plot of log K_ex versus 1/T yields straight lines with slopes proportional to the enthalpy. Examination of Fig. 5b shows two straight lines with slopes of 1.678 and 2.310 for Pd(II) and Pt(IV) extraction, respectively. The negative values of the enthalpy changes, ΔH° −32.1 kJ mol⁻¹ and −44.2 kJ mol⁻¹ for the Pd(II) and Pt(IV) extraction calculated from the slopes of the linear relations indicate the exothermic nature of the PGMs extraction processes. Such negative ΔH° values are agreement with those reported by Cieszynska et al. regarding Pd(II) extraction with Cyphos IL 104.²⁵ The formation of ionic bonding, which is involved to electrostatic interaction between phosphonium cation and chloro-metal complexes ([PdCl₄]²⁻, [PtCl₆]²⁻), can be attributed to the negative enthalpy contribution (ΔH° < 0). Thus, the PGMs extractions are preferably carried out at ambient temperatures.

Table 1 Equilibrium constants and thermodynamic parameters for the PMGs extraction at different temperatures. Organic phase: 0.6 g L⁻¹ [P₆₆₆₁₄]⁺Cl⁻; aqueous phase: 100 mg L⁻¹ Pt(IV), 55 mg L⁻¹ Pd(II), 25 mg L⁻¹ Rh(III), 0.1 mol L⁻¹ HCl; O/A = 1; t = 10 min; T = 288–338 K

Temp (K)	log Kex-Pd(II)			log Kex-Pt(IV)
288	2.54	−14.0	−18.1	6.43	−35.4	−8.8
298	2.27	−12.9	−19.2	6.01	−34.3	−9.9
308	2.11	−12.5	−19.6	5.78	−34.1	−10.1
318	1.92	−11.7	−20.4	5.54	−33.7	−10.5
328	1.79	−11.2	−20.9	5.36	−33.7	−10.5
338	1.67	−10.8	−21.3	5.21	−33.7	−10.5

---

The overall standard Gibbs free energy change, ΔG°, is related to the equilibrium constant by the relationship:

ΔG° = −2.303RT log K_ex (16)

Once the overall free energy and enthalpic change have been obtained, the standard entropic contributions TΔS° can be obtained from the relationship:

TΔS° = ΔH° − ΔG° (17)

Table 1 lists the values for log K_ex, ΔG°, and TΔS° for Pd(II) and Pt(IV) extraction with [P₆₆₆₁₄]⁺Cl⁻ at 0.1 mol L⁻¹ HCl. The evaluated thermodynamic parameters provide further information into both the aqueous phase complexation and the organic phase reactions. As TΔS° > ΔH°, the standard Gibbs free energy changes ΔG° are negative, which refer to the spontaneous nature of the Pd(II) and Pt(IV) extraction reactions with [P₆₆₆₁₄]⁺Cl⁻ at 0.1 mol L⁻¹ HCl. The negative values ΔS°, −64.3 J K⁻¹ mol⁻¹ for Pd(II) and −33.3 J K⁻¹ mol⁻¹ for Pt(IV) extraction, imply the degree of order has increased during the extraction process. In fact, the ionic bonding, which involves the electrostatic attraction between oppositely charged ions, is responsible for a decrease in the randomness of the system and gives a negative entropy contribution. Furthermore, the steric hindrance (steric crowding) of the extracted species formed in the organic phase may also result in negative entropy change.

3.2 Stripping of PGMs from the loaded organic phase

3.2.1 Selection of stripping reagent. Stripping of PGMs from the loaded organic phase is a critical property to selectively recover the metals and to recycle the organic phase used for next extraction-stripping cycles. The loaded organic phase containing 100 mg L⁻¹ Pd(II) and 55 mg L⁻¹ Pt(IV) was prepared in two co-current stages (two successive contacts of 0.6 g L⁻¹ [P₆₆₆₁₄]⁺Cl⁻ with fresh aqueous phase at the second contact at 0.1 mol L⁻¹ HCl) (see below). Stripping of PGMs from the loaded organic with deionized water was inefficient because of strong ionic bonds between [P₆₆₆₁₄]⁺ and the negative charge of chloro-metal species ([PdCl_n]⁽ⁿ⁻²⁾⁻ and [PtCl₆]²⁻) in the organic phase. The inefficiency of water is also because water has no anion species to exchange and the barrier of the interface may prevent the hydrolysis of the chloro-complexes of Pd(II) and Pt(IV). Therefore, aqueous solutions that may form stable complexes with PGMs, such as NH₄OH, Na₂S₂O₃, Na₂S₂O₄, NaSCN, and (NH₄)₂CS/HCl, were individually used as stripping reagents. As shown in Table 2, Na₂S₂O₃ and Na₂S₂O₄ solutions yielded the lowest stripping efficiency for PGMs, whereas the quantitative stripping of Pd(II) can be achieved with NH₄OH and (NH₄)₂CS/HCl solutions, which results in the co-stripping of 81.9% and 10.5% of Pt(IV), respectively. Among these strip solutions, NaSCN exhibits a selective stripping of Pt(IV) over Pd(II). Optimized conditions for stripping of Pd(II) and Pt(IV) were investigated in further experiments.

Table 2 Stripping behavior of PGMs from the organic phases using various strip solutions. Organic phase (previously loaded at 0.1 mol L⁻¹ HCl): 100 mg L⁻¹ Pt(IV), 55 mg L⁻¹ Pd(II); strip solution: 0.01 mol L⁻¹; O/A = 1; t = 10 min; T = 298 K

Strip solution	Stripping efficiency, %
	Pd	Pt
NH4OH	94.6	81.9
Na2S2O3	<0.1	1.45
Na2S2O4	<0.1	7.33
NaSCN	<0.1	67.4
(NH4)2CS in 5% v/v HCl	100	10.5

---

3.2.2 Effect of strip solution concentration. The stripping was considered using an acidic thiourea solution or a sodium thiocyanide one as the strip phase at various concentrations and an O/A ratio of 1. As shown in Fig. 6a, the percentage stripping of Pd(II) increased significantly from the 4.4% to 99.9% as thiourea concentration varied from 10⁻⁴ to 10⁻² mol L⁻¹ in 5% v/v of HCl. Along with Pd(II) stripping, Pt(IV) was also co-stripped from the loaded organic phase by up to 10.5% using 10⁻² mol L⁻¹ thiourea.

Fig. 6 Effect of thiourea and NaSCN concentrations on the PGMs stripping. Organic phase (previously loaded at 0.1 mol L⁻¹ HCl): 100 mg L⁻¹ Pt(IV), 55 mg L⁻¹ Pd(II); strip solution: (a) 10⁻⁴ to 10⁻² mol L⁻¹ (NH₄)₂CS in 5% v/v HCl; (b) 10⁻³ to 10⁻¹ mol L⁻¹ NaSCN; O/A = 1; t = 10 min; T = 298 K.

The percentage stripping of Pt(IV) increased from 7.9% to 92.8% as the NaSCN increased from 10⁻³ to 10⁻¹ mol L⁻¹ (Fig. 6b). In contrast to the acidic thiourea solution, NaSCN exhibits a selective stripping of Pt(IV) over Pd(II) in the investigated range. Therefore, the sequence of strip solutions used plays an important role in the selective recovery of PGMs. Thus, the NaSCN solution is preferably used in the first stripping of Pt(IV) followed by a second stripping of Pd(II) with acidic thiourea solution.

The presence of stable complexes, [Pt(SCN)₆]²⁻ and [Pd(NH₂CSNH₂)₄]²⁺, in the aqueous phase has been reported in the literature.⁵⁵ Therefore, the mechanism of Pt(IV) and Pd(II) stripping can be written as follows:

Note that the kinetics of PGMs stripping in both of the cases was very fast, and the equilibrium state was achieved within 1 min. Neither emulsion nor precipitation was observed, which suggests the possibility of operating the stripping in continuous modes.

3.3 Extraction-stripping cycles

To verify the recyclability of the organic phase, a series of extraction-stripping experiments was performed as follows. A fresh and/or regenerated organic phase of 0.6 g L⁻¹ [P₆₆₆₁₄]⁺Cl⁻ was equilibrated with an equal volume of feed solution in two co-current stages at 0.1 mol L⁻¹ HCl and O/A = 1. After phase separation, the loaded organic phase containing 100 mg L⁻¹ Pt(IV) and 55 mg L⁻¹ Pd(II) was stripped with a 0.1 mol L⁻¹ NaSCN solution (three co-current stages), followed by Pd(II) stripping with 0.01 mol L⁻¹ CS(NH₂)₂ in 5% HCl (two co-current stages) at an O/A ratio of 1. The stripped organic phase was then reused for a new extraction-stripping cycle under similar experimental conditions. As seen in Fig. 7, the recycled organic phase exhibits the same extractive behaviors (at first stage of the two co-current stages) as the fresh solution of [P₆₆₆₁₄]⁺Cl⁻ for at least 5 cycles. The average extraction efficiencies of Pd(II), Pt(IV), and Rh(III) were 78%, 63%, and less than 2%, respectively. The result indicates that the organic phase can be easily recycled without loss of performance. From an economical perspective, this is beneficial for industrial processes.

Fig. 7 Extraction of PGMs (at first stage of the two co-current stages) with fresh and regenerated [P₆₆₆₁₄]⁺Cl⁻. Extraction (two co-current stages): fresh and/or regenerated organic phase: 0.6 g L⁻¹ [P₆₆₆₁₄]⁺Cl⁻; aqueous phase: 100 mg L⁻¹ Pt(IV), 55 mg L⁻¹ Pd(II), 25 mg L⁻¹ Rh(III), 0.1 mol L⁻¹ HCl; O/A = 1; t = 10 min; T = 298 K. Stripping: organic phase (previously loaded at 0.1 mol L⁻¹ HCl): 100 mg L⁻¹ Pt(IV), 55 mg L⁻¹ Pd(II); strip solution: 0.1 mol L⁻¹ NaSCN (three co-current stages) and 0.01 mol L⁻¹ (NH₄)₂CS in 5% v/v HCl (two co-current stages); O/A = 1; t = 10 min; T = 298 K.

3.4 McCabe–Thiele diagrams

3.4.1 Extraction isotherms and counter-current extraction simulated with a cascade of batch extractors. Examination of Fig. 8 shows that two theoretical stages would be sufficient to quantitatively extract Pd(II) and Pt(IV) using 0.6 g L⁻¹ [P₆₆₆₁₄]⁺Cl⁻ at an O/A ratio of 3/2, leaving Rh(III) in the raffinate solution. Subsequently, the extraction of PGMs was performed on a batch-wise laboratory scale to simulate a 2-stage continuous counter-current extraction using 0.6 g L⁻¹ [P₆₆₆₁₄]⁺Cl⁻ at an O/A ratio of 3/2 to confirm the McCabe–Thiele diagram prediction (Fig. 9). The procedure consists of repeated introductions of feed solution and fresh organic phase into a series of batch extractions in separatory funnels, along with the withdrawal of the extract and raffinate phases. After a number of extraction cycles, the system should approach steady state, and the liquids in the funnels resemble the streams that would exist in an actual continuous countercurrent extraction. Fig. 9 shows the profile of the PGMs concentration in the organic and aqueous phases at a steady state attained within 5 complete cycles of operation. The final raffinate containing less than 1 mg L⁻¹ Pd(II) and Pt(IV) corresponds to over 99% extraction of the metals into the organic phase. An analysis of the loaded organic phase containing 37 mg L⁻¹ Pd(II) and 65 mg L⁻¹ Pt(IV) confirms the mass balance during the extraction stage. The Pd(II) and Pt(IV) extraction in the first stage was 85.3% and 79.7%, respectively, and in second stage, the values were found to be 99.9% and 99.6%, respectively. The results obtained are in agreement with the data predicted in the McCabe–Thiele plot.

Fig. 8 McCabe–Thiele diagram for the PGMs extraction. Organic phase: 0.6 g L⁻¹ [P₆₆₆₁₄]⁺Cl⁻; feed aqueous phase: 100 mg L⁻¹ Pt(IV), 55 mg L⁻¹ Pd(II), 25 mg L⁻¹ Rh(III), 0.1 mol L⁻¹ HCl; t = 10 min; T = 298 K.

Fig. 9 Counter-current mode simulated with a cascade of batch extractors and mass balance for the extraction and stripping of PGMs. Extraction: organic phase: 0.6 g L⁻¹ [P₆₆₆₁₄]⁺Cl⁻, aqueous phase: 100 mg L⁻¹ Pt(IV), 55 mg L⁻¹ Pd(II), 25 mg L⁻¹ Rh(III), 0.1 mol L⁻¹ HCl; O/A = 3/2; 2 counter-current stages, t = 10 min; T = 298 K. First stripping: loaded organic phase: 65 mg L⁻¹ Pt(IV), 37 mg L⁻¹ Pd(II); strip solution: 0.1 mol L⁻¹ NaSCN; O/A = 3/1; 3 counter-current stages, t = 10 min; T = 298 K. Second stripping: loaded organic phase: 37 mg L⁻¹ Pd(II); strip solution: 0.01 mol L⁻¹ thiourea in 5% v/v HCl; O/A = 5/1; single counter-current stage, t = 10 min; T = 298 K.

3.4.2 Stripping isotherms and counter-current stripping simulated with a cascade of batch extractors. As mentioned above, a 0.1 mol L⁻¹ NaSCN solution was chosen as the selective stripping phase for Pt(IV). To investigate the number of stages required for the stripping of Pt(IV) in the continuous mode, the McCabe–Thiele diagram was plotted, as shown in Fig. 10. The loaded organic phases containing 100 mg L⁻¹ Pt(IV) and 55 mg L⁻¹ Pd(II) were mixed with 0.1 mol L⁻¹ NaSCN at various O/A ratios from 1/5 to 5/1 while maintaining the total volume of the phases constant. Fig. 10 reveals that two theoretical stages are required to quantitatively strip Pt(IV) from the loaded organic phase at an O/A ratio of 1, while three theoretical counter-current stages are required at an O/A = 3/1. Increasing the O/A ratio leads to an increase in the number of theoretical stages required for quantitative stripping of Pt(IV). Considering the enrichment of Pt(IV) in the strip liquor and the number of theoretical stages, the O/A ratio of 3/1 is preferable to that of 1/1 for stripping of Pt(IV).

Fig. 10 McCabe–Thiele diagram for the Pt(IV) stripping. Organic phase (previously loaded at 0.1 mol L⁻¹ HCl): 100 mg L⁻¹ Pt(IV), 55 mg L⁻¹ Pd(II); strip solution: 0.1 mol L⁻¹ NaSCN; O/A = 1/5 to 5/1; t = 10 min; T = 298 K.

Afterward, a 3-stage batch counter-current stripping study was performed using a 0.1 mol L⁻¹ NaSCN solution at an O/A ratio of 3/1. The organic phase obtained from the previous experiment of the two-stage counter-current for PGMs extraction contains 65 mg L⁻¹ Pt(IV) and 37 mg L⁻¹ Pd(II). Fig. 9 represents the PGMs concentrations in the strip liquor and in the stripped organic phase at steady state. The results correspond well with the prediction obtained from the McCabe–Thiele diagram. Furthermore, the mass balance during the stripping was also considered. The ultimate strip liquor contained over 197 mg L⁻¹ Pt(IV), which indicated that the Pt(IV) stripping efficiency was more than 99.9%. Thus, the concentration of platinum in the strip liquor was enriched 3-fold after three counter-current stripping stages. Note that the co-stripping of Pd(II) from the loaded organic was less than 0.1% using 0.1 mol L⁻¹ NaSCN, even in the simulated counter-current mode.

Subsequently, the Pd(II)-loaded organic phase was used for a stripping study in a further investigation using an acidic thiourea solution. As shown in Fig. 9, the quantitative stripping of Pd(II) with 0.1 mol L⁻¹ thiourea in 5% HCl was achieved within one stage at an O/A ratio of 5/1. The stripping of Pd(II) with the (NH₄)₂CS/HCl system was rapid and efficient. The strip liquor contains up to 183 mg L⁻¹ Pd(II) corresponding to 3.3 times of enrichment of Pd(II) in the feed solution. Neither precipitation nor third phase formation was observed during the stripping process.

It is of interest that Cieszynska et Wisniewski have previously investigated the extraction of PGMs from chloride media with Cyphos IL 101 and Cyphos IL 104.^25–27 These authors have shown that Pt(IV) and Pd(II) were mostly co-extracted by Cyphos IL 101 and that Pd(II) was poorly separated from Pt(IV) at the stripping stage by using 0.5 mol L⁻¹ NH₄OH as a stripping agent. Conversely, the flowsheet given in Fig. 9 based on the use of Cyphos IL 101 as an extractant and on the use of NaSCN and thiourea/HCl as stripping agents for Pt(IV) and Pd(II), respectively, is highly effective for the recovery of these two metals from HCl solutions and leads to Pt(IV) with a purity higher than 99.9% (w) vs. Pd(II) + Rh(III) and Pd(II) with a purity higher than 99.9% (w) vs. Pt(IV) + Rh(III). The strategy of using NaSCN to strip selectively Pt(IV) in the first place and to use thiourea/HCl to strip remaining Pd(II) in a second place, after their previous co-extraction with Alamine 300, was proposed by Swain et al.¹⁰ As shown in Fig. 9, this strategy of two step selective stripping has been successfully adapted here to the case of co-extraction of Pt(IV) and Pd(II) with Cyphos IL 101.

4 Conclusions

Phosphonium-based ionic liquid [P₆₆₆₁₄]⁺Cl⁻ can be successfully used for the selective co-extraction of Pd(II) and Pt(IV) from acidic chloride media, leaving Rh(III) in the raffinate. The extraction efficiencies of Pd(II) and Pt(IV) were contingent upon the concentrations of [P₆₆₆₁₄]⁺Cl⁻ and chloride ions in the feed solution. Job's continuous variation method revealed the Pd(II)/[P₆₆₆₁₄]⁺Cl⁻ stoichiometry was 1/1 and 1/3 at 0.1 and 4.0 mol L⁻¹ HCl, respectively, whereas the Pt(IV)/[P₆₆₆₁₄]⁺Cl⁻ stoichiometry remained constant at 1/2, regardless of the acidity of the aqueous phase. The complexes formed in the ionic liquid phases via anionic exchange mechanism were further characterized by ¹H and ³¹P NMR spectroscopies.

The selective recovery of Pt(IV) from the loaded organic phase was first achieved by selective and complete stripping with 0.1 mol L⁻¹ NaSCN, then Pd(II) was stripped with a 0.01 mol L⁻¹ acidic thiourea solution. A series of extraction-stripping cycles showed that the organic phase could be recycled without loss of performance compared with the fresh ionic liquid for the extraction of PGMs from HCl/Cl⁻ media. Experiments performed with a cascade of batch extractors simulating counter-current modes provided results that are in good agreement with the data predicted in the McCabe–Thiele plots for PGMs extraction and stripping. Particularly, Pd(II) and Pt(IV) were quantitatively extracted with 0.6 g L⁻¹ [P₆₆₆₁₄]⁺Cl⁻ from 0.1 mol L⁻¹ HCl initially containing 100 mg L⁻¹ Pt(IV), 55 mg L⁻¹ Pd(II), and 25 mg L⁻¹ Rh(III) within two counter-current stages at O/A = 3/2, leaving Rh(III) in the raffinate. The complete stripping of Pt(IV) with a 0.1 mol L⁻¹ NaSCN solution required three counter-current stages at O/A = 3/1, whereas the quantitative stripping of Pd(II) with 0.01 mol L⁻¹ thiourea in 5% HCl was achieved within a single counter-current stage at O/A = 5/1.

An integrated process based on the use of Cyphos IL 101 as an extractant and on the use of NaSCN and thiourea/HCl as stripping agents for Pt(IV) and Pd(II), respectively, is highly effective for the recovery of these two metals from HCl solutions and leads to Pt(IV) with a purity higher than 99.9% (w) vs. Pd(II) + Rh(III) and Pd(II) with a purity higher than 99.9% (w) vs. Pt(IV) + Rh(III). The strategy of using NaSCN to strip selectively Pt(IV) in the first place and to use thiourea/HCl to strip remaining Pd(II) in a second place, has been successfully adapted to the case of co-extraction of Pt(IV) and Pd(II) with Cyphos IL 101. Moreover, a counter-current extraction-stripping simulated with a cascade of batch extractors suggests that phosphonium-based ionic liquid such as Cyphos IL 101 can be advantageously used as extractant to recover individually Pd, Pt and Rh from acidic chloride feed solutions as those encountered in the PGM's recycling industry.

Acknowledgements

This work was supported by the basic research program of the Korea Institute of Geoscience and Mineral Resources (KIGAM). Support by Cytec Industries Inc. (Canada) is gratefully acknowledged for donating Cyphos IL 101. Thanks are also due to Dr Vinay Kumar, a visiting scientist under the Brain Pool program of KIGAM, for his valuable comments and suggestions which have led to significant improvement on the quality of the paper.

References

1. Johnson Matthey Catalysts, Platinum 2013: Interim Review, http://www.platinum.matthey.com/documents/market-review/2013-interim/full-review/english.pdf, accessed 06/20/2015, 2015.
2. R. M. Izatt, S. R. Izatt, N. E. Izatt, K. E. Krakowiak, R. L. Bruening and L. Navarro, Green Chem., 2015, 17, 2236–2245.
3. T. C. Lo, in Handbook of Separations Techniques for Chemical Engineers, ed. P. A. Schwietzer, McGraw-Hill, New York, 1996, p. 1-450.
4. B. Cote and G. P. Demopoulos, Solvent Extr. Ion Exch., 1994, 12, 393–421.
5. B. Cote and G. P. Demopoulos, Solvent Extr. Ion Exch., 1994, 12, 517–540.
6. E. Benguerel, G. P. Demopoulos, G. Cote and D. Bauer, Solvent Extr. Ion Exch., 1994, 12, 497–516.
7. E. Benguerel, G. Cote, A. Lautié, G. Demopoulos and D. Bauer, J. Chem. Technol. Biotechnol., 1995, 62, 380–384.
8. M. V. Rane and V. Venugopal, Hydrometallurgy, 2006, 84, 54–59.
9. V. I. Men'shikov, I. Y. Voronova, O. A. Proidakova, S. F. Malysheva, N. I. Ivanova, N. A. Belogorlova, N. K. Gusarova and B. A. Trofimov, Russ. J. Appl. Chem., 2009, 82, 183–189.
10. B. Swain, J. Jeong, S.-k. Kim and J.-c. Lee, Hydrometallurgy, 2010, 104, 1–7.
11. P. P. Sun and M. S. Lee, Hydrometallurgy, 2011, 109, 181–184.
12. V. V. Belova, A. I. Khol'kin and T. I. Zhidkova, Theor. Found. Chem. Eng., 2007, 41, 743–751.
13. P. Giridhar, K. A. Venkatesan, T. G. Srinivasan and P. R. V. Rao, Hydrometallurgy, 2006, 81, 30–39.
14. U. Domańska, A. Pobudkowska and F. Eckert, Green Chem., 2006, 8, 268–276.
15. Z. Kolarik, Solvent Extr. Ion Exch., 2013, 31, 24–60.
16. P. R. Vasudeva Rao, K. A. Venkatesan, A. Rout, T. G. Srinivasan and K. Nagarajan, Sep. Sci. Technol., 2011, 47, 204–222.
17. J. G. Huddleston, H. D. Willauer, R. P. Swatloski, A. E. Visser and R. D. Rogers, Chem. Commun., 1998, 1765–1766.
18. A. P. de los Rios, F. J. Hernandez-Fernandez, L. J. Lozano, S. Sanchez, J. I. Moreno and C. Godinez, J. Chem. Eng. Data, 2010, 55, 605–608.
19. M. Mincher, D. Quach, Y. Liao, B. Mincher and C. Wai, Solvent Extr. Ion Exch., 2012, 30, 735–747.
20. N. Papaiconomou, S. Génand-Pinaz, J.-M. Leveque and S. Guittonneau, Dalton Trans., 2013, 42, 1979–1982.
21. Y. Zuo, Y. Liu, J. Chen and D. Li, Ind. Eng. Chem. Res., 2008, 47, 2349–2355.
22. X. Sun, H. Luo and S. Dai, Chem. Rev., 2012, 112, 2100–2128.
23. J. P. Hallett and T. Welton, Chem. Rev., 2011, 111, 3508–3576.
24. H. Niedermeyer, J. P. Hallett, I. J. Villar-Garcia, P. A. Hunt and T. Welton, Chem. Soc. Rev., 2012, 41, 7780–7802.
25. A. Cieszynska and M. Wiśniewski, Hydrometallurgy, 2012, 113–114, 79–85.
26. A. Cieszynska and M. Wisniewski, Sep. Purif. Technol., 2010, 73, 202–207.
27. A. Cieszynska and M. Wisniewski, Sep. Purif. Technol., 2011, 80, 385–389.
28. M. L. Dietz and J. A. Dzielawa, Chem. Commun., 2001, 20, 2124–2125.
29. M. L. Dietz, J. A. Dzielawa, I. Laszak, B. A. Young and M. P. Jensen, Green Chem., 2003, 5, 682–685.
30. M. L. Dietz and D. C. Stepinski, Green Chem., 2005, 7, 747–750.
31. M. P. Jensen, J. A. Dzielawa, P. Rickert and M. L. Dietz, J. Am. Chem. Soc., 2002, 124, 10664–10665.
32. M. Zhu, J. Zhao, Y. Li, N. Mehio, Y. Qi, H. Liu and S. Dai, Green Chem., 2015, 17, 2981–2993.
33. T. V. Hoogerstraete, S. Wellens, K. Verachtert and K. Binnemans, Green Chem., 2013, 15, 919–927.
34. D. Parmentier, S. J. Metz and M. C. Kroon, Green Chem., 2013, 15, 205–209.
35. S. Wellens, R. Goovaerts, C. Möller, J. Luyten, B. Thijs and K. Binnemans, Green Chem., 2013, 15, 3160–3164.
36. A. Rout, S. Wellens and K. Binnemans, RSC Adv., 2014, 4, 5753–5758.
37. R. Navarro, I. Saucedo, C. Gonzalez and E. Guibal, Chem. Eng. J., 2012, 185–186, 226–235.
38. T. Vincent, A. Parodi and E. Guibal, Sep. Purif. Technol., 2008, 62, 470–479.
39. T. Vincent, A. Parodi and E. Guibal, React. Funct. Polym., 2008, 68, 1159–1169.
40. V. T. Nguyen, J.-c. Lee, J. Jeong, B.-s. Kim, G. Cote and A. Chagnes, Ind. Eng. Chem. Res., 2015, 54, 1350–1358.
41. M. Regel-Rosocka, Sep. Purif. Technol., 2009, 66, 19–24.
42. A. C. Samuels, C. A. Boele, K. T. Bennett, S. B. Clark, N. A. Wall and A. E. Clark, Inorg. Chem., 2014, 53, 12315–12322.
43. W. Robb and G. M. Harris, J. Am. Chem. Soc., 1965, 87, 4472–4476.
44. E. Benguerel, G. P. Demopoulos and G. B. Harris, Hydrometallurgy, 1996, 40, 135–152.
45. M. L. Good, S. E. Bryan, F. F. Holland Jr. and G. J. Maus, J. Inorg. Nucl. Chem., 1963, 25, 1167–1173.
46. S. Damach, G. Cote and D. Bauer, C. R. Acad. Sci., Ser. II: Mec., Phys., Chim., Sci. Terre Univers, 1987, 304, 889–892.
47. M. S. Lee, J. Y. Lee, J. R. Kumar, J. S. Kim and J. S. Sohn, Mater. Trans., 2008, 49, 2823–2828.
48. F. J. Alguacil, A. Cobo, A. G. Coedo, M. T. Dorado and A. Sastre, Hydrometallurgy, 1997, 44, 203–212.
49. M. Tanaka, T. Tamagawa and Y. Hamada, Mater. Trans., 1992, 33, 380–390.
50. J. J. C. Jansz, Hydrometallurgy, 1983, 11, 13–31.
51. G. Cote, A. Jakubiak, D. Bauer, J. Szymanowski, B. Mokili and C. Poitrenaud, Solvent Extr. Ion Exch., 1994, 12, 99–120.
52. G. Cote and A. Jakubiak, Hydrometallurgy, 1996, 43, 277–286.
53. D. J. A. De Waal and W. Robb, Int. J. Chem. Kinet., 1974, 6, 309–321.
54. D. Cholico-Gonzalez, M. Avila-Rodriguez, G. Cote and A. Chagnes, J. Mol. Liq., 2013, 187, 165–170.
55. S. I. Ginzburg, N. A. Ezerskaya, I. V. Prokof’eva, N. V. Fedorenko, V. I. Shlenskaya and N. K. Bel'skii, Analytical chemistry of Platinum metals, John Wiley & Son, New York, 1975, pp. 116–143.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra09328k

---