Competition between lanthanides in extraction with tri-n-butyl phosphate in supercritical CO₂ from solid nitrates

Abstract

Competition extraction of lanthanide nitrates with tri-n-butyl phosphate (TBP) in supercritical CO₂ (SC-CO₂) is investigated by monitoring the absorption spectra of Ln(III) in UV-Vis region. For light lanthanides Nd and Pr, the extracted Nd(III) and Pr(III) complexes with TBP in SC-CO₂ increase proportionally with the concentration of TBP and the pressure, and the separation factor, β_Nd/Pr does not change obviously. In contrast, for the pair of one light and one heavy lanthanides, Nd and Ho, Nd(III) is preferentially extracted when TBP is inadequate for completely extracting both Nd(III) and Ho(III). In addition, the extraction of heavier Ho(III) is preferred over that of the lighter Nd(III) at higher pressures and temperatures, resulting in an increasing separation factor β_Ho/Nd. A noteworthy observation should be emphasized, that is the loading capacity of Ln(III) complexes in SC-CO₂ is much bigger for mixed Ln(III) ions than for single one Ln(III). The different competition extraction behavior for the light–light Ln(III) pair and for the light–heavy Ln(III) pair, and the enhanced loading capacity of mixed Ln(III) complexes found here provide useful fundamental information for selective extraction of Ln(III) ions in SC-CO₂.

---

1. Introduction

The application of supercritical CO₂ (SC-CO₂) as solvent in the extraction, chemical synthesis and formation of fine particles is attracting more attention, due to its environmentally benign properties.^1–4 By taking the advantages of gas-like viscosity and liquid-like solubility, extraction processes using SC-CO₂ as solvent have not only been developed as normal solvent–solvent extraction method, but also been deployed to extract interested substance directly from various solid matrices. Among these applications, a very active area is the extraction of metal ions using SC-CO₂ system. In the processes, metal ions themselves are not soluble in SC-CO₂ duo to their polar nature, but they can be transformed into SC-CO₂ soluble metal complexes with a variety of suitable organic chelating ligands, and then can be effectively extracted.⁵ The application of process based on SC-CO₂ to nuclear waste treatment is very appealing to nuclear industry. In traditional liquid–liquid extraction processes, solid nuclear wastes containing lanthanide and actinide oxides, such as UO₂, are usually dissolved into nitric acid to provide feeding solutions, and a large volume of liquid solution with high radioactivity is generated causing serious public concern. In contrast, the direct extraction of lanthanides and actinides from solid wastes using SC-CO₂ containing suitable chelating ligands avoids the nitric acid dissolution procedure, and the amount of the second-hand liquid waste generated in the processes can be sharply reduced. In the past two decades, the extraction of lanthanides and actinides from solid matrices,^6–9 metal oxides^10–12 and ionic liquid^13,14 by using SC-CO₂ system have been extensively studied, and application in industry scale has been successfully demonstrated. For instance, a pilot plant has been built and operated by AREVA NP, aiming to recover high enriched uranium from incineration ash by SC-CO₂ containing TBP–HNO₃ reagent.¹⁵ Recovery of uranium and other key actinides from real irradiated nuclear fuel has been successfully tested too.^16,17 The extraction of uranium from particle pellets coated with carbon and SiC that used in high temperature gas-cooled reactor fuel has also been achieved.^18,19

Although great progress has been made in the extraction of lanthanides and actinides by using SC-CO₂ system, the separation (selective extraction) of lanthanides and actinides or from each other is still a challenging task, and seldom studied.²⁰ They are complicated system, involving complexation reaction of multiple components, and the temperature, pressure and solubility all have significant effect on the complexation reactions. Some researchers reported the complexation reaction and solubility of single component in SC-CO₂. Fox et al.^21,22 studied the stoichiometry of some lanthanide complexes with TBP using UV-Vis and luminescence spectroscopies. Both praseodymium nitrate and neodymium nitrate form 4 : 1 (of TBP : Ln) complexes in SC-CO₂, while holmium nitrate could form 2 : 1 and 4 : 1 Ho(III) complexes. The lower ratio of TBP : Ln in stoichiometry, the complex is less soluble in SC-CO₂.²³ The solubilities of trialkyl phosphates in SC-CO₂ were reported by Pitchaiah et al.,²⁴ and it was found that the solubility of TBP was approximately 0.1 mol mol⁻¹ at the temperature range of 50–70 °C and the pressure of 15–25 MPa. The solubility of UO₂(NO₃)₂·2TBP was determined to be 0.0025–0.5 mol L⁻¹ at 40 °C and 10–22.5 MPa.²⁵ The solubilities of lanthanide complexes with fluorinated β-diketonates were determined to be 0.06–0.2 mol L⁻¹ at 40 °C and 25 MPa, and heavy lanthanide complexes were found more soluble than the light ones.²⁶

Despite those fruitful results from the fundamental studies, an application of SC-CO₂ as solvent in a nuclear industry process may need a comprehensive understanding to further optimize the system. Taking the inspiration from the successes in our previous projects on extracting single lanthanides and actinides with TBP in SC-CO₂, in this work, we look at a more complicated system, the competition extraction of two lanthanide nitrates with TBP in SC-CO₂. At thermodynamic equilibrium state, the extracted Ln(III) complexes are investigated by using UV-Vis absorption spectroscopy. The effects of the amount of TBP, the oxidation states of lanthanides, the temperature and pressure on the competition extraction have been discussed.

2. Experimental

2.1. Chemicals

Tri-n-butyl phosphate (TBP) of analytical reagent was used as purchased from Alfa Aesar Chemical Co. Ltd. Ultra-pure carbon dioxide (CO₂, 99.995%) was provided by Beijing Bei Temperature Gas Factory, China. Lanthanide nitrates Nd(NO₃)₃·6H₂O, Pr(NO₃)₃·6H₂O, Ho(NO₃)₃·6H₂O, Ce(NO₃)₃·6H₂O and (NH₄)₂Ce(NO₃)₆·6H₂O, all with a purity above 99.9%, were bought from LNcreative Ltd., China. 0.05 mol L⁻¹ ethylenediaminetetraacetic acid disodium (EDTA) standard solution from Alfa Aesar Chemical Co. Ltd., was used to determine the concentrations of Ln(III) by titration method in a hexamethylenetetramine buffer solution of pH 5–6 with 0.1% xylenol orange as indicator.

2.2. High-pressure cell

A high-pressure cell with optical windows was designed and machined for performing the extraction of lanthanide nitrates with TBP in SC-CO₂ at 60 °C and 25 MPa. The reaction cell consisted of two blocks of stainless-steel (17-4PH) as shown in Fig. 1. Two blocks were tightly sealed by a threaded cap with an O-ring (ø 45 × 2.4 mm) resistant to high pressure CO₂. On the lower block, a chelating ligand inlet, a CO₂ outlet and a blind well for thermocouple probe were bored, and a concave bottom was adopted to facilitate the complexation reaction. On the upper block, two cylindrical platforms and a CO₂ inlet were bored, two cylindrical quartz windows (ø 15 × 12 mm) were sealed on the platforms by backing nuts with PTFE gaskets and an O-ring as the sealing accessories. A threaded hole in the backing nut was machined to accommodate the fiber probe. The optical path length between the two quartz windows is 1.6 cm. The inner volume of the reactor vessel is 41.2 mL.

Fig. 1 High-pressure cell with optical windows. (1) CO₂ outlet, (2) magnetic stirring bar, (3) threaded cap, (4) optical fiber adaptor, (5) backing nut, (6) quartz window, (7) O-ring, (8) CO₂ inlet, (9) PTFE gasket, (10) PEEK buffer, (11) O-ring, (12) reagent inlet, (13) thermocouple well, (14) magnetic stirrer.

2.3. Apparatus

The scheme of the apparatus used for competition extraction was shown in Fig. 2. The high-pressure cell was placed in a thermostat and connected with three functional sections, namely, CO₂ supply, chelating ligand supply, and spectroscopy measurement section. The CO₂ supply section consisted of a CO₂ cylinder, a cooling bath, a CO₂ pump, a one-way valve, a pressure sensor and a safety valve. The chelating ligand supply section consisted of an electronic weighing balance, a TBP vial, and a high pressure reciprocating plunger pump. The electronic weighing balance was used to precisely record the amount of chelating ligand TBP pumped into the cell. The spectroscopy measurement section consisted of a deuterium–halogen light source (Avalight-DH-S-Bal, Avantes, Apeldoorn, The Netherlands), optical fibers, a UV-Vis spectrometer (Avaspec-dual, Avantes, Apeldoorn, The Netherlands) and a control computer.

Fig. 2 Scheme of the apparatus used for competition extraction. (1) CO₂ cylinder, (2) needle valve, (3) cooling bath, (4) CO₂ pump, (5) one-way valve, (6) pressure sensor, (7) safety valve, (8) high-pressure cell, (9) magnetic stirrer, (10) thermostat, (11) thermometer, (12) electronic weighing balance, (13) TBP vial, (14) reciprocating plunger pump, (15) light source, (16) optical fiber, (17) UV-Vis spectrometer, (18) computer, (19) collection vial.

2.4. Procedure

The molar absorption coefficients of the lanthanide complexes with TBP (TBP–Ln(III) complexes) in SC-CO₂ was first determined. Each TBP–Ln(III) complex was prepared by mixing lanthanide nitrate with TBP in a centrifugal test tube, then shaking the test tube for 30 min, and centrifuging to remove excess lanthanide nitrate and free water. The concentration of lanthanide in the TBP–Ln(III) sample was determined by EDTA titration. To begin with, the dead space of tubes and valves in the experimental system was prefilled with a TBP–Ln(III) complex. The high-pressure cell was heated to designed temperature, then CO₂ was pumped in to purge the system. The system was gradually pressurized to supercritical state after closing the CO₂ outlet valve. Subsequently, the absorbance spectrum of SC-CO₂ was recorded as reference when the system reached stable state. After the reference spectrum was recorded, several portions of weighed TBP–Ln(III) complex were introduced into the cell, absorption spectra were collected accordingly. The relationship between the absorbance at selected wavelength and the amount of the added TBP–Ln(III) complex was plotted, thus the molar absorption coefficient was obtained by Beer–Lambert lawwhere, A is the absorbance at a selected wavelength; I and I₀ are the intensities of transmitted light and incident light, respectively; c (mol L⁻¹) is the concentration of TBP–Ln(III) complex in SC-CO₂; l (cm) is the optical path length; and ε (L mol⁻¹ cm⁻¹) is the molar absorption coefficient.

To study the effect of the concentration of TBP on the competition extraction, two lanthanide nitrates of approximately 0.1 g each were placed in the cell. After the cell reached designed temperature and pressure, a reference spectrum was recorded. Then weighed amount of TBP was pumped into the cell, a spectrum containing the characterized absorption of the two TBP–Ln(III) complexes was recorded at the equilibrium state. The concentrations of the lanthanide complexes in SC-CO₂ were calculated from the absorbance and the molar absorption coefficients, and the relationship between the concentrations of TBP–Ln(III) complexes and added TBP was obtained.

To investigate the effect of the pressure on the competition extraction, two lanthanide nitrates of about 0.1 g each were placed in the cell, and the system was brought to supercritical state. Then a quantitative amount of TBP was added, and absorption spectra were collected accordingly at varying pressures of the system controlled by pumping more CO₂ into the cell. The relationship between the TBP–Ln(III) complexes concentrations and pressures was obtained by following the same procedure described above. The effect of the temperature was studied in a similar way, by increasing the temperature of the system instead of pumping CO₂.

The competition extraction behavior of lanthanide nitrates A and B with TBP in SC-CO₂ was evaluated by a separation factor β_A/B defined as following equation:

where, [A]_scf and [B]_scf are the molar quantities of A and B complexes in SC-CO₂, which can be calculated by the absorbance and according molar absorption coefficients; [A]_solid,0 and [B]_solid,0 are the molar quantities of A and B initially added in the cell.

3. Results and discussion

3.1. Molar absorption coefficients

The spectrum of TBP–Nd(III) complex in SC-CO₂ is similar to that of neodymium nitrate in aqueous solution, as shown in Fig. 3. Its maximum absorption is at 582 nm. A strong absorption around 300 nm is noticed, increasing with the addition of TBP–Nd(III) complex. This peak is attributed to the absorption of free TBP, which is confirmed by only pumping TBP to the cell. However, the attempt to quantitatively analyzes the concentration of free TBP in SC-CO₂ is not reliable, as the incident intensity of the light source is not strong enough around 300 nm. The spectra of TBP–Ho(III) and TBP–Pr(III) complexes in SC-CO₂ are also similar with their nitrates in aqueous solution, the maximum absorption of TBP–Ho(III) and TBP–Pr(III) are at 452 and 446 nm, respectively.

Fig. 3 The spectrum of TBP–Nd(III) complex in SC-CO₂ at 49.7 ± 0.2 °C and 14.8 ± 0.05 MPa.

The maximum absorbance of the TBP–Ln(III) complexes increase linearly with their molar quantities introduced into the cell as shown in Fig. 4. The slopes of the plots of TBP–Nd(III) and TBP–Ho(III) are larger than that of TBP–Pr(III). Taking into account the volume of the cell and the optical path length in the system, the molar absorption coefficients of TBP–Nd(III), TBP–Ho(III) and TBP–Pr(III) in SC-CO₂ are calculated to be 10.41, 8.36 and 4.03 L mol⁻¹ cm⁻¹, respectively at 582, 452, and 446 nm. Cautions should be taken, when the concentration is measured by absorption spectra in SC-CO₂, because the molar absorption coefficients of some solutes might change with CO₂ densities.²⁷ On the contrary, the molar absorption coefficient of UO₂(NO₃)₂·2TBP complex was reported independent of the temperature and pressure,^28,29 or the influence was slight.³⁰ Here, the effect of the temperature and the pressure on the molar absorption coefficients of the TBP–Ln(III) complexes are examined, by monitoring the absorbance intensity of the fixed amount of TBP–Ln(III) sample at different pressure and temperature. The relative standard deviation from average is ±0.7% in the pressure range of 14–19 MPa at 50 °C, and ±1.4% in the temperature range of 45–60 °C. No clear trend is found in the deviation, and no obvious absorption position change is observed. The effects of the temperature and the pressure on the molar absorption coefficients of TBP–Ln(III) complexes are slight, thus, no further calibration is made for following measurements.

Fig. 4 The relationship between the maximum absorbance and the molar quantity of TBP–Ln(III) complexes. TBP–Nd(III): 0.218 mol L⁻¹, 49.5 ± 0.3 °C, 16.00 ± 0.05 MPa; TBP–Ho(III): 0.221 mol L⁻¹, 49.2 ± 0.2 °C, 16.05 ± 0.05 MPa; TBP–Pr(III): 0.223 mol L⁻¹, 49.4 ± 0.2 °C, 15.51 ± 0.05 MPa.

3.2. Effect of the amount of TBP

The effect of the relative amount of TBP on the concentrations of the TBP–Ln(III) complexes is shown in Fig. 5. For the pair of the two light lanthanides (Nd–Pr), the concentrations of the TBP–Ln(III) complexes in SC-CO₂ increase proportionally with TBP, and the separation factor is at around 1.0. When the molar ratio of [TBP]/[Pr + Nd]_total exceeds 10 : 1, their concentrations reach approximately 6.0 mmol L⁻¹. For the pair of light-heavy lanthanides (Nd–Ho), their concentrations also increase with the amount of TBP, but Nd(III) complex is preferentially formed with TBP when [TBP]/[Ho + Nd]_total is less than 3, and the separation factor β_Ho/Nd is lower than 1. The separation factor is about 1.3 when [TBP]/[Ho + Nd]_total is 4, then slowly decreases to about 1.1 as adequate amount of TBP is added. As reported, Ho(III) forms 2 : 1 TBP–Ho(III) complex at low TBP concentration, and forms 4 : 1 complex at high TBP concentration,²² and lower TBP : Ln stoichiometry complex is less soluble in SC-CO₂.²³ Our observation in the Nd–Ho competition extraction shows that the concentration of TBP–Nd(III) is larger than that of TBP–Ho(III) at low [TBP]/[Ho + Nd]_total ratio, which agrees well with literatures.

Fig. 5 Effect of the amount of TBP on the competition extraction of lanthanide nitrates. (a) Nd(NO₃)₃·6H₂O was 0.24 mmol, Pr(NO₃)₃·6H₂O was 0.24 mmol, 50.0 ± 0.1 °C, 14.82 ± 0.02 MPa; (b) Nd(NO₃)₃·6H₂O was 0.22 mmol, Ho(NO₃)₃·6H₂O was 0.23 mmol, 50.0 ± 0.2 °C, 14.90 ± 0.05 MPa.

3.3. Ce(III) and Ce(IV)

It is interesting to investigate the competition extraction behavior of lanthanides in different oxidation states. Among lanthanide nitrates, cerium has two stable oxidation states, Ce(III) and Ce(IV), however, neither Ce(III) nor Ce(IV) could give a satisfactory UV-Vis absorption spectra in our system. Thus, the behavior of Ce(III) and Ce(IV) are evaluated by monitoring the TBP–Nd(III) concentration when Ce(III) or Ce(IV) presents in the system, as shown in Fig. 6. The concentration of the TBP–Nd(III) complex in SC-CO₂ increases rapidly with the amount of TBP when only neodymium nitrate presents, and its ultimate concentration reaches about 3.3 mmol L⁻¹ at [TBP]/[Nd] above 20 : 1. The concentration of the TBP–Nd(III) increases a little slower when Ce(III) co-exists, and the ultimate concentration of TBP–Nd(III) reaches about 5.8 mmol L⁻¹, much higher than in the case that only neodymium nitrate presents. The concentration of the TBP–Nd(III) complex also increases much slower in the present of Ce(IV), and the ultimate concentration reaches about 4.5 mmol L⁻¹. The results suggest that both Ce(III) and Ce(IV) form complexes with TBP, resulting in the slower increment and the higher ultimate concentration of TBP–Nd(III) complex. It is difficult to make a further analysis on the different extraction behavior of Ce(III) and Ce(IV) with TBP, because the information about the complexation reaction and solubilities of the cerium complexes is not available.

Fig. 6 Concentration profiles of TBP–Nd(III) complex in SC-CO₂ in the presence of Ce(III) and Ce(IV). (■) Nd(NO₃)₃·6H₂O was 0.23 mmol, 49.7 ± 0.3 °C, 15.07 ± 0.03 MPa (●) Nd(NO₃)₃·6H₂O was 0.23 mmol, Ce(NO₃)₃·6H₂O was 0.24 mmol, 50.2 ± 0.2 °C, 15.10 ± 0.05 MPa (▲) Nd(NO₃)₃·6H₂O was 0.23 mmol, Ce(NH₄)₂(NO₃)₆·6H₂O was 0.22 mmol, 49.9 ± 0.3 °C, 15.11 ± 0.03 MPa.

It is noteworthy to point out that the ultimate concentration of each lanthanide complex in SC-CO₂ is higher when other lanthanide co-presents as shown in Fig. 5 and 6. To understand this behavior, the dissolution mechanism of the TBP–Ln(III) complexes in SC-CO₂ and the effect of free TBP concentration on their solubilities are discussed.

Metal ions in lanthanide complexes are shielded by chelating ligand TBP, thus the dissolution mechanism of lanthanide complexes in SC-CO₂ is similar to that of chelating ligand TBP, which is the interaction of organic phosphate group with CO₂, such as Lewis acid–Lewis base interaction and weak C⋯H–O hydrogen bonding.^31–33 Usually, the solubility of a metal complex is much lower than that of the chelating agent.³⁴ The solubility of TBP is 1.3 mol L⁻¹ at a density of 700 g L⁻¹ (50 °C, 15 MPa), while that of UO₂(NO₃)₂·​2TBP is only 0.025 mol L⁻¹ at the same condition.^25,35 Thus, the solubility of TBP–Ln(III) complexes are estimated much lower than that of TBP. In addition, the solubilities of solutes are reported to decrease in a co-solutes system over its pure component system, due to the repulsive interactions between the solutes.^36,37 The solubilities of lanthanide complexes are likely affected by the concentration of free TBP, since they are dissolved in a similar manner, but with much lower solubilities than TBP.

The concentration of free TBP is analyzed. For a system only one lanthanide nitrate presents, the general reaction can be written as following:

Because the solubility of H₂O in SC-CO₂ below 100 °C is low,^38,39 H₂O and solid lanthanide nitrate are treated as heterogeneous phase. Then the equilibrium of complexation reaction is written as following:

The relationship between TBP initially added ([TBP]₀) and the free TBP ([TBP]_free) could be written as following:

In the competition extraction system, the relationship between the initial TBP and the free TBP is as following:

[TBP]₀ = [TBP]_free + mk′₁[TBP]_free^m + nk′₂[TBP]_freeⁿ (6)

Because all the constants in above equations are positive, we can conclude that [TBP]_free increases nonlinearly with [TBP]₀ from eqn (5), even some TBP is consumed in the complexation reaction. In addition, by comparing eqn (6) with eqn (5), it can be suggested that the increase of free TBP is slower in the competition extraction system.

Based on above discussion, our results suggest that the solubility of TBP–Nd(III) complex is restricted much earlier by the high concentration of free TBP, if only neodymium nitrate presents. Thus, the local concentration of TBP–Nd(III) near the surface of neodymium nitrate is increased, which preventing further complexation reaction, resulting in a low ultimate concentration in SC-CO₂. In contrast, the concentration of free TBP increases in a modest way in the competition extraction system, and much higher ultimate concentrations of TBP–Ln(III) complexes in SC-CO₂ can be obtained. In general, free TBP plays a dual role in the system, promoting the complexation reaction of lanthanide nitrates, at the same time restricting the solubilities of lanthanide complexes.

3.4. Effect of pressure

The effect of the pressure on the competition extraction was investigated by pumping more CO₂ into the cell at 50 °C. The TBP amount was fixed at a 3 : 1 [TBP]/[Ln]_total ratio, which ensured a sufficient absorbance intensity, meanwhile, maintained the competition relationship between lanthanide nitrates.

The relationship between the solubility of a solute and the density is described by Chrastil model:^24,40

where, S (g L⁻¹) is the solubility of a solute in supercritical fluid; ρ (g L⁻¹) is the density of the supercritical fluid; k is the association number that describes the number of solvent molecules associated with the complex; T is the temperature in K; A and B are empirical parameters which can be achieved by experimental data.

The density of CO₂ increases as more CO₂ is pumped into the system, and the solubility of solute is expected to increase. The concentrations of the TBP–Ln(III) complex in the Nd–Pr competition extraction system indeed increase with the pressure at first, from approximately 2.5 to 3.1 mmol L⁻¹, as the pressure increases from 14 to 15 MPa, as shown in Fig. 7. But their concentrations keep constant when the pressure is further increased. Because initially the concentrations of TBP–Ln(III) are controlled by solubilities as Section 3.3 discussed, the increase of dissolving capacity allows further complexation of Nd(III) and Pr(III) with TBP; but later the concentrations are controlled by complexation reaction since TBP becomes inadequate, even pressure is further increased. The separation factor is near 1.0, suggesting that the concentrations of TBP–Nd(III) and TBP–Pr(III) increase proportionally with the pressure.

Fig. 7 Effect of the pressure on the competition extraction of lanthanide nitrates. (a) Nd(NO₃)₃·6H₂O was 0.24 mmol, Pr(NO₃)₃·6H₂O was 0.24 mmol, [TBP]/[Nd + Pr]_total = 3, 49.5 ± 0.2 °C; (b) Nd(NO₃)₃·6H₂O was 0.23 mmol, Ho(NO₃)₃·6H₂O was 0.22 mmol, [TBP]/[Nd + Ho]_total = 3, 49.7 ± 0.2 °C.

However, quite different behavior is observed for the Nd–Ho competition extraction system. TBP–Ho(III) complex concentration increases from 2.4 to 4.0 mmol L⁻¹, while TBP–Nd(III) concentration decreases from 2.4 to 2.0 mmol L⁻¹ as the pressure is increased from 14 to 18.5 MPa, and the separation factor increases from 1.0 to 2.1. Metal complexes are reported can undergo ligand exchange reaction in organic phase^41,42 and in SC-CO₂.⁴³ The effect of the pressure on the Nd–Ho competition extraction suggests that ligand exchange reaction between TBP–Nd(III) and TBP–Ho(III) might occur, as the dissolving capacity of SC-CO₂ increases with density, resulting in the formation of TBP–Nd(III) with low TBP : Nd stoichiometry, and TBP–Ho(III) with higher TBP : Ho stoichiometry. Lower TBP : Ln stoichiometry complex is less soluble in SC-CO₂,²³ therefore, the concentration of TBP–Ho(III) increases, while that of TBP–Nd(III) decreases. The inner driven force might due to the higher solubilities of heavy Ln(III) complexes in SC-CO₂.²⁶

3.5. Effect of temperature

Temperature has effect on the density ρ and A/T term in eqn (7), and the term A/T is often much smaller than B, thus the effect of the temperature on solubility is mainly attributed to its effect on the density.^24,44 In a dynamic extraction system, the increase of temperature results in the decrease of density when the pressure is kept constant, thus the solubility of solute decreases with temperature.⁴⁵

However, in present static extraction system, the density of CO₂ does not change as the temperature is increased from 45 to 60 °C, and the pressure is increased accordingly from 15 to 19 MPa, thus there is actually no apparent change in the dissolving capacity. In the Nd–Pr competition extraction system, the concentrations of the complexes and separation factor are constant as shown in Fig. 8. In the Nd–Ho competition extraction system, a strong temperature effect is observed. TBP–Ho(III) concentration increases from 2.8 to 4.3 mmol L⁻¹, while TBP–Nd(III) concentration decreases from 3.1 to 2.8 mmol L⁻¹ as temperature is increased from 45 to 60 °C, and the separation factor increases from 0.9 to 1.6. Like the effect of the pressure on the Nd–Ho competition system, this might also be attributed to the ligand exchange reaction that is induced by the changing temperature. The equilibrium of the ligand exchange reaction is rebalanced at the elevated temperature, leading to the formation of TBP–Ho(III) complex with higher TBP : Ho stoichiometry, and TBP–Nd(III) complexes with lower TBP : Nd stoichiometry. And higher TBP : Ln stoichiometry complex is more soluble, thus the concentration of TBP–Ho(III) complex increases and that of TBP–Nd(III) decreases.

Fig. 8 Effect of the temperature on the competition extraction of lanthanide nitrates (a) Nd(NO₃)₃·6H₂O was 0.23 mmol, Pr(NO₃)₃·6H₂O was 0.24 mmol, [TBP]/[Nd + Pr]_total = 3; (b) Nd(NO₃)₃·6H₂O was 0.23 mmol, Ho(NO₃)₃·6H₂O was 0.23 mmol, [TBP]/[Nd + Ho]_total = 3.

4. Conclusions

Competition extraction of lanthanides with TBP in SC-CO₂ is investigated. For the Nd–Pr competition extraction system, the concentrations of the lanthanide complexes response simultaneously to the amount of TBP, the pressure and the temperature. However, for the Nd–Ho competition extraction system, Nd(III) is preferentially extracted when TBP is deficient. In addition, the concentration of TBP–Ho(III) complex increases with the pressure and the temperature, while that of TBP–Nd(III) complex decreases. The results may help understanding those competition extraction system, providing new fundamental information for the selective extraction of metal ions in SC-CO₂. Further work on competition extraction of uranyl nitrate and lanthanide nitrates with TBP in SC-CO₂ is undergoing.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (No. 21301192) and the China Scholarship Council.

References

1. E. J. Beckman, J. Supercrit. Fluids, 2004, 28, 121–191.
2. X. X. Zhang, S. Heinonen and E. Levanen, RSC Adv., 2014, 4, 61137–61152.
3. P. T. Anastas and J. B. Zimmerman, Innovations in Green Chemistry and Green Engineering, Springer, New York, 2013.
4. T. Clifford, Fundamentals of supercritical fluids, Oxford University, Oxford, 1999.
5. K. E. Laintz, C. M. Wai, C. R. Yonker and R. D. Smith, Anal. Chem., 1992, 64, 2875–2878.
6. S. M. Ghoreishi, A. Hedayati and K. Ansari, J. Supercrit. Fluids, 2016, 117, 131–137.
7. K. Park, W. Jung and J. Park, Metals, 2015, 5, 1788–1798.
8. P. Prabhat, A. Rao, P. Kumar and B. S. Tomar, Hydrometallurgy, 2016, 164, 177–183.
9. G. X. Tian, W. S. Liao, C. M. Wai and L. F. Rao, Ind. Eng. Chem. Res., 2008, 47, 2803–2807.
10. M. D. Samsonov, T. I. Trofimov, Y. M. Kulyako and B. F. Myasoedov, Radiochemistry, 2007, 49, 246–250.
11. W. H. Duan, P. J. Cao and Y. J. Zhu, J. Rare Earths, 2010, 28, 221–226.
12. L. Y. Zhu, W. H. Duan, J. M. Xu and Y. J. Zhu, Chin. J. Chem. Eng., 2009, 17, 214–218.
13. J. S. Wang, C. N. Sheaff, B. Yoon, R. S. Addleman and C. M. Wai, Chem.–Eur. J., 2009, 15, 4458–4463.
14. A. Rao and B. S. Tomar, Sep. Purif. Technol., 2016, 161, 159–164.
15. C. M. Wai, in Advanced Separation Techniques for Nuclear Fuel Reprocessing and Radioactive Waste Treatment, Woodhead Publishing, 2011, pp. 414–435.
16. A. Shadrin, A. Murzin, A. Lumpov and V. Romanovsky, Solvent Extr. Ion Exch., 2008, 26, 797–806.
17. M. Kamiya, S. Miura, Y. Koma, T. Koyama, K. Aoki, S. Yamada and K. Sawada, Development of Actinides Co-extraction System with Direct Extraction Process Using Supercritical Fluid, Global, 2011.
18. W. H. Duan, L. Y. Zhu and Y. J. Zhu, Prog. Nucl. Energy, 2011, 53, 664–667.
19. L. Y. Zhu, W. H. Duan, J. M. Xu and Y. J. Zhu, J. Hazard. Mater., 2012, 241–242, 456–462.
20. D. L. Quach, B. J. Mincher and C. M. Wai, J. Hazard. Mater., 2014, 274, 360–366.
21. R. V. Fox, R. D. Ball, P. D. Harrington, H. W. Rollins, J. J. Jolley and C. M. Wai, J. Supercrit. Fluids, 2004, 31, 273–286.
22. R. V. Fox, R. D. Ball, P. D. Harrington, H. W. Rollins and C. M. Wai, J. Supercrit. Fluids, 2005, 36, 137–144.
23. D. L. Baek, R. V. Fox, M. E. Case, L. K. Sinclair, A. B. Schmidt, P. R. McIlwain, B. J. Mincher and C. M. Wai, Ind. Eng. Chem. Res., 2016, 55, 7154–7163.
24. K. C. Pitchaiah, N. Sivaraman, N. Lamba and G. Madras, RSC Adv., 2016, 6, 51286–51295.
25. M. J. Carrott, B. E. Waller, N. G. Smart and C. M. Wai, Chem. Commun., 1998, 373–374.
26. D. Hwang, T. Tsukahara, N. Miyamoto and Y. Ikeda, J. Supercrit. Fluids, 2016, 110, 251–256.
27. J. K. Rice, E. D. Niemeyer and F. V. Bright, Anal. Chem., 1995, 67, 4354–4357.
28. M. J. Carrott and C. M. Wai, Anal. Chem., 1998, 70, 2421–2425.
29. R. S. Addleman and C. M. Wai, Phys. Chem. Chem. Phys., 1999, 1, 783–790.
30. T. Sasaki, Y. Meguro and Z. Yoshida, Talanta, 1998, 46, 689–695.
31. K. H. Kim and Y. Kim, Bull. Korean Chem. Soc., 2007, 28, 2454–2458.
32. P. Raveendran, Y. Ikushima and S. L. Wallen, Acc. Chem. Res., 2005, 38, 478–485.
33. W. H. Teoh, R. Mammucari and N. R. Foster, J. Organomet. Chem., 2013, 724, 102–116.
34. W. Cross, A. Akgerman and C. Erkey, Ind. Eng. Chem. Res., 1996, 35, 1765–1770.
35. Y. Meguro, S. Iso, T. Sasaki and Z. Yoshida, Anal. Chem., 1998, 70, 774–779.
36. S. N. Reddy and G. Madras, J. Supercrit. Fluids, 2012, 63, 105–114.
37. Z. Huang, M. Feng, Y. H. Guo, J. F. Su, L. J. Teng, T. Y. Liu and Y. C. Chiew, Fluid Phase Equilib., 2008, 272, 8–17.
38. R. Oparin, T. Tassaing, Y. Danten and M. Besnard, J. Chem. Phys., 2005, 123, 224501.
39. N. Sabirzyanov, A. P. Il'in, A. R. Akhunov and F. M. Gumerov, High Temp., 2002, 40, 203–206.
40. J. Chrastil, J. Phys. Chem., 1982, 86, 3016–3021.
41. S. T. H. and S. W. E., Inorg. Nucl. Chem. Lett., 1967, 3, 279–284.
42. T. H. Siddall III and W. E. Stewart, Solvent Extr. Ion Exch., 2000, 18, 809–820.
43. Y. Kachi, Y. Kayaki, T. Tsukahara, T. Ikariya and Y. Ikeda, Inorg. Chem., 2008, 47, 349–359.
44. N. Lamba, R. C. Narayan, J. Raval, J. Modak and G. Madras, RSC Adv., 2016, 6, 17772–17781.
45. N. Lamba, R. C. Narayan, J. Modak and G. Madras, J. Supercrit. Fluids, 2016, 107, 384–391.

---