Extraction kinetics of Fe(III) by di-(2-ethylhexyl) phosphoric acid using a Y–Y shaped microfluidic device

Highlights

• A microfluidic device was used to characterise Fe^III extraction kinetics.

• The forward rate constant was determined using finite-volume numerical simulations.

• Mass transport occurred under a mixed reaction–diffusion transfer resistance regime.

• Insight is provided for the suppression of Fe^III extraction rates.

Abstract

The extraction kinetics of Fe^III by di-(2-ethylhexyl) phosphoric acid (D2EHPA) were investigated using a Y–Y shaped microfluidic device. Finite-volume simulations were used to examine the accuracy of a single-step dimeric reaction mechanism in fitting the experimental data. Results demonstrate the validity of the proposed mechanism and show that Fe^III extraction occurred at a slow rate (second-order forward rate constant of k₁ = (3.0 ± 0.1) × 10⁻⁶ m⁴/mol s) under a mixed reaction–diffusion resistance regime. The present study provides insight for the control of Fe^III extraction rates in hydrometallurgical processes.

Introduction

Solvent extraction (SX), a widely used hydrometallurgical separations technique, exploits an aqueous/organic interface for the purposes of purification and concentration (Lo et al., 1983). SX-based processes play an increasingly important role in mineral processing (Cote, 2000, Flett, 1977, Flett, 2004, Flett, 2005, Habashi, 1993, Lo et al., 1983, Ritcey, 2006, Sole et al., 2005), however SX plants pose an environmental hazard due to the large quantities of organic solvents required (Anastas and Warner, 2000). The next generation of processing equipment will therefore need to incorporate reductions in solvent use (Ji et al., 2006). New designs have been investigated – including hollow-fibre contactors (Gabelman and Hwang, 1999) – and installed, for example column-based contactors in the uranium industry (Bart, 2005, Eccles, 2000).

Equilibrium concepts have traditionally been used for the design of conventional mixer–settler SX units with large residence times. New equipment, however, will utilise shorter timescales of liquid–liquid contact, meaning that the contribution of kinetics to mass transfer becomes important (Ji et al., 2006). Hence, reliable kinetic data is required specific to the extraction reactions envisaged for new contactor technologies. Kinetic determinations, however, can be difficult in SX systems. Multiple mass transfer mechanisms exist, including bulk species diffusion and reaction at the two-phase interface. Transfer resistance due to diffusion can dominate over that due to the reaction, making it difficult to access the kinetic regime experimentally (Ciceri et al., 2013). The use of microfluidic devices can potentially address this problem. As the flow regime within a microfluidic device is laminar, numerical analysis can be used to determine extraction kinetics from experimental data. This advantage – combined with reduced sample quantities, large specific interfacial areas and short contact times between the liquid phases – makes microfluidic solvent extraction (μSX) well suited for the study of SX kinetics. An overview of μSX studies has been given previously (Ciceri et al., 2013).

Iron is present as a contaminant in many hydrometallurgical processes (Biswas et al., 2007, Demopoulos and Gefvert, 1984, Flett, 2005, Sahu and Das, 2000, Sole et al., 2005, Yu et al., 1989). One approach to minimising product impurity levels has been to choose extractant–solvent combinations that do not react with iron or react only selectively (Demopoulos and Gefvert, 1984, Haggag et al., 1977, Sahu and Das, 1997, Sahu and Das, 2000). An alternative strategy is to use kinetics to suppress Fe extraction rates. Kinetic information, however, is scarce for typical reaction scenarios involving iron. The kinetics of Fe^III by di-(2-ethylhexyl) phosphoric acid (D2EHPA), for example, are not well understood – despite this being a relatively slow reaction with long equilibration times (Roddy et al., 1971). An overview of available equilibrium and kinetic information for the reaction is given in Table 1, Table 2, respectively.

Above a pH of 1, Fe^III is known to undergo self-hydrolysis: the overall set of hydrolytic equilibria is complex and involves several chemical species (Arden, 1951, Blesa and Matijevic, 1989, Cotton and Wilkinson, 1988, Dousma et al., 1979, Kobylin et al., 2007, Majzlan and Myneni, 2005, Milburn and Vosburgh, 1955, Rose and Waite, 2007, Sapieszko et al., 1977). For cases in which speciation is neglected, a simplified extraction mechanism is given by

$$\text{F}{\text{e}}_{(aq)}^{\text{III}}+x{(\text{HL})}_{2(org)}\rightleftarrows {{[{\text{FeL}}_x{(\text{HL})}_x]}_{(org)}}^{(3-x)+}+x{{\text{H}}_{(aq)}}^{+}\text{,}$$

where HL denotes the monomeric form of the D2EHPA extractant, (HL)₂ the dimeric form and (HL)_x a polymeric form with degree of polymerisation x. As shown in Table 1, several values of x have been reported for Eq. (1), indicating that speciation of Fe^III under the various reported experimental conditions is important (Yu et al., 1989). The presence of ligands can change the aqueous-phase speciation: SO₄²⁻ ions, for example, strongly coordinate the Fe^III metallic centre (Dousma et al., 1979, Flynn, 1984) to give FeSO₄⁺. Consequently, [FeL_x(HL)_x]^(3−x)+ is not necessarily the extraction product as implied by Eq. (1).

Ultraviolet–visible spectroscopy has been used to determine the nature of the extraction product in the organic phase (Golovanov, 1998, Van de Voorde, 2008, Van de Voorde et al., 2007, Whitcomb, 1987). However, little information can be gained from this technique. Similarly, the infrared analysis of extracted complexes has been attempted by Smythe et al. (1968), however specific structural conclusions could not be made. Whitcomb et al. (1992) reported the existence of an FeOPFe bimetallic complex that was shown to polymerise and coordinate water molecules or nitrate ions; the possibility of sulfate ion coordination was not excluded. Sulfate-coordinated compounds such as [FeL(HL)SO₄] could also be expected in the organic phase. There is little experimental evidence to support this hypothesis, however, as the structural conformation of the extracted compounds of Fe^III are complicated (Good et al., 1963) and are yet to be clarified.

Both aqueous-bulk and interfacial reaction mechanisms have been previously proposed for the present system. Karpacheva and Ilozheva (1969) proposed an aqueous-bulk mechanism for the extraction of Fe^III by D2EHPA in a water/synthine system using a highly stirred apparatus and a narrow concentration range. Sato et al. (1985) considered an aqueous-bulk mechanism modified to account for the formation of hydrolysed FeOH²⁺ species. Roddy et al. (1971) concluded that interfacial saturation by a partially complexed Fe^III intermediate was important for quiescent interfaces using a Lewis cell (Lewis, 1954a, Lewis, 1954b), but not for dispersive mixing. Matsuyama et al. (1990) determined the rate-limiting step to be an Eigen mechanism (Cotton and Wilkinson, 1988) involving FeOH⁺ and either HL or L^– using a water/ethanol system, and determined this step to be exclusively interfacial using a water/benzene system.

Speciation effects have been considered in the formulation of kinetic models. Islam and Biswas (1981) reported a change in the extraction stoichiometry due to a change in concentration of hydrolysed Fe^III in the aqueous phase. Using a single-drop technique, Biswas and Begum (1999) accounted for the speciation of Fe^III in chloride media. Other studies have emphasised the role of FeOH²⁺ species in the extraction mechanism (El-Nadi and El-Hefny, 2010, Flynn, 1984, Islam and Biswas, 1981, Matsuyama et al., 1990, Roddy et al., 1971, Sato et al., 1985). The effect of specific anions such as Cl⁻ on the hydrolysis has also been discussed. Sato et al. (1985) discussed both ion-exchange and associative mechanisms as a function of the HCl concentration in the aqueous phase. Reaction mechanism uncertainty has previously been discussed for Fe^III extraction from sulfuric acid media (El-Nadi and El-Hefny, 2010, Haggag et al., 1977, Islam et al., 1979, Suarez et al., 2002).

In this study, the extraction kinetics of Fe^III in a sulfate medium by D2EHPA are investigated using a Y–Y shaped microfluidic cell. Aqueous bulk mechanisms have been excluded due to the low solubility of D2EHPA in the aqueous phase (Ciceri et al., 2013). The reaction of hydrolytic species has been neglected due to the strong coordination of SO₄^2– to the Fe^III metallic centre. A single-step dimeric mechanism is examined using finite-volume numerical simulations in combination with data obtained from μSX experiments.