Ki Chul
Park
,
Haruka
Tateno
and
Takehiko
Tsukahara
*
Laboratory for Advanced Nuclear Energy, Institute of Innovative Research in Tokyo Institute of Technology, 2-12-1-N1-6 O-okayama, Meguro-ku, Tokyo, 152-8550, Japan. E-mail: ptsuka@lane.iir.titech.ac.jp
First published on 13th December 2017
The present study investigates the applicability of poly(N-isopropylacrylamide) (PNIPAAm) as an extraction medium for trivalent lanthanide (Ln(III)) ions in high concentrations of nitric acid solutions by using octyl(phenyl)-N,N-diisobutylcarbamoylmethylphosphine oxide (CMPO) as an extracting agent. The extraction is based on the hydrophobic, co-aggregative binding of the Ln-CMPO nitrate complex formed in situ onto the PNIPAAm aggregates growing above the phase-transition temperature. The resulting co-aggregates enabled easy separation of the liquid and solid phases. The extraction of Ln(III) ions was affected by the efficiency and extent of complexation and the stability of the complex. Furthermore, the difference in the stability of Ln-CMPO complex species allowed for the stripping of Ln(III) ions in hydrochloric acid solutions.
Solid phase extraction (SPE) based on a clouding phenomenon (phase transition) of surfactant micelles has been extensively studied as a green, safe and economical process for separation of metal ions. Cloud point extraction (CPE) using mainly non-ionic (or zwitterionic) surfactants is based on the trapping of metal complexes into the hydrophobic cores of micelles and subsequent phase separation by controlling the temperature above (or below) the cloud points.7–10 The great advantage of CPE lies in the high preconcentration of solutes from aqueous solutions to a far smaller volume of micellar phases, compared with conventional liquid phase extraction. The solubilization of hydrophobic solutes in aqueous solutions is attained by the interaction with micellar binding sites, forming an equilibrium between free solutes and micelle-bound solutes. The equilibrium (binding) constant11 is directly related to the extraction efficiency of CPE. Therefore, various experimental factors affecting the equilibrium, as well as those related to micellization and cloud points, are required to be optimized to enhance the extractability of CPE.7
Another approach similar to CPE includes SPE utilizing the phase transition of poly(N-isopropylacrylamide) (PNIPAAm, Fig. 1) as the extraction medium.12–14 PNIPAAm is a well-known smart polymer exhibiting a thermo-responsive property of reversibly switching hydrophilicity and hydrophobicity at a lower critical solution temperature (LCST) of about 32 °C.15–17 Thus far, a wide variety of PNIPAAm-based materials including homo- and copolymers have been applied to the adsorption and detection of metal ions, i.e., Ag(I),18 Cu(I),18,19 Cs(I),20 Li(I),21 Fe(II),12,13 Cu(II),22–24 Pd(II),18 Co(II),19 Ni(II),25 Cd(II),26 Hg(II),27 Pb(II),27,28 Au(III),18,29 In(III),14 Ln(III).30 The application of PNIPAAm as an extraction medium was first reported for the extraction of a water-soluble cationic chelate, tris(1,10-phenanthroline)iron(II), as an ion pair with perchlorate anions.12,13 On the other hand, Tokuyama and Iwama reported the first application of PNIPAAm to the direct extraction of a water-insoluble complex species, the In(III)-organophosphate complex.14 The SPE of In(III) ions is composed of two main processes, that is, the complexation of In(III) ions with organophosphate ligands in the presence of dissolved PNIPAAm below the LCST, and hydrophobic binding of the metal complex onto the PNIPAAm precipitates formed above the LCST. The extraction of In(III) ions was successfully achieved at a high extractability of 89.1% (0.854 mmol per g dry polymer). Furthermore, based on the dependence of extractability on the initial ligand amount and pH, it was indicated that the amount of In(III) extracted to PNIPAAm precipitates was determined by the degree of complex formation, which was enhanced with increasing organophosphate anions as coordination species in the complexation equilibrium. Inversely, the stripping of In(III) ions was achieved by decreasing the anionic species of ligands in more acidic solutions.
We have explored the applicability of PNIPAAm-based SPE to the extraction of metal ions in highly acidic solutions, aiming to develop a new technology for spent nuclear fuel reprocessing. In this study, we investigate its applicability to the extraction of Ln(III) ions in high concentrations of nitric acid (1.0–3.0 mol L−1 HNO3). PNIPAAm-based SPE is more advantageous in respect that the extraction process is simply based on the hydrophobic binding between PNIPAAm and metal complexes, compared with CPE with complicated experimental parameters. Above the LCST, PNIPAAm dissolved in an aqueous solution initiates intramolecular hydrophobic interaction to form globular particles by a conformational change from the coil state,16 further growing to larger aggregates of hydrophobic particles. From another perspective, an inherently difficult molecular-level dispersion of hydrophobic macromolecules is considered to be achieved in the early stage of phase transition. Therefore, we expect that a high dispersion of PNIPAAm molecules realizes a highly efficient binding with the metal complex coexisting in the system. In this study, we report the PNIPAAm-based extraction of Ln(III) ions using CMPO ligands in nitric acid solutions of mixed ions (La3+, Nd3+, Eu3+, Lu3+, Cs+ and Sr2+).
The stripping of metal ions was carried out using the same procedure as the extraction process, except for the additional washing operations. The glutinous, gel-like aggregates of PNIPAAm left in the jacketed beaker at 40 °C were washed four times with nitric acid (1.0 mol L−1, 30 mL × 3 and 15 mL × 1) pre-heated at the temperature. After washing, the aggregates were re-dissolved in hydrochloric acid (1.0 mol L−1, 15 mL) under vigorous stirring at 20 °C. The stirring was continued for 1 h after the dissolution of the aggregates. The subsequent phase separation was carried out at 40 °C under vigorous stirring for 1 h. After the recovery of the liquid phase, the same stripping operation was repeated twice. The recovered solutions were also diluted using ultrapure water and nitric acid for ICP-MS analysis.
According to the above-mentioned procedure, the extraction was conducted basically by varying the metal concentrations (0.0378–1.21 mmol L−1 for each metal ion, solution volume: 15 mL) under constant amounts of PNIPAAm (30.8 mg, NIPAAm-monomer unit: 0.272 mmol) and CMPO (11.1 mg, 0.0272 mmol). Each batch experiment was carried out three times in total, and the average values of the metal amounts determined by ICP-MS analysis were used to plot the extraction and stripping percentages. The error bar in each plot represents the standard deviation of three batch experiments.
As shown in Fig. 3, the extractability was dependent on the nitric acid concentration. The extraction was conducted at 20 °C for 1 h, followed by 40 °C for 1 h (open symbols in Fig. 3, [Ln] = [Cs] = [Sr] = 0.0755 mmol L−1). The higher extractability of La3+, Nd3+ and Eu3+ was observed at 1.0 mol L−1 nitric acid, while it declined at higher acid concentrations. The LCST of PNIPAAm at nitric acid concentrations of more than 2.0 mol L−1 was observed below 10 °C (Fig. 4). Therefore, the decrease in extractability is attributed mainly to the decrease in extraction efficiency of the Ln-CMPO complex, due to the insolubility of PNIPAAm at 20 °C. Actually, the extractability at 2.0 mol L−1 nitric acid was improved in the experiment conducted at 4 °C below the LCST (filled symbols in Fig. 3), although it stayed lower than that at 1.0 mol L−1 nitric acid. The abrupt decrease of LCST in nitric acid solutions indicates that nitrate anions would more remarkably reduce the hydrogen-bond energy around the amide groups of PNIPAAm, compared with chloride anions.30
The effect of polymer amount on the extraction efficiency was examined in the presence of an excess amount of Ln ions against CMPO. The molar ratios of CMPO against Ln ions and total metals correspond to [CMPO]/[Ln]total = 0.375 and [CMPO]/[M]total = 0.250, respectively ([Ln] = [Cs] = [Sr] = 1.21 mmol L−1, [HNO3] = 1.0 mol L−1). As shown in Fig. 5, the electroneutral CMPO exhibited almost no extractability to the alkali and alkaline earth metal ions. In contrast, Ln(III) ions were extracted to PNIPAAm (30.8 mg) at the total extraction percentage of 12.3% (22.0, 14.9, 10.8, 1.4% for La3+, Nd3+, Eu3+ and Lu3+, respectively). The total extraction capacity calculated from the extraction percentage corresponds to the high value of 41.8 mg per g dry PNIPAAm (18.0, 12.6, 9.6 and 1.5 mg per g dry PNIPAAm for La3+, Nd3+, Eu3+ and Lu3+, respectively). The presence of excess Ln ions induced the selective coordination of CMPO, probably due to the competitive complexation among Ln ions with CMPO. The lighter La3+, Nd3+ and Eu3+ exhibited much higher extractability than the heavy Lu3+. This tendency is consistent with that of conventional SPE.31,32 In general, the extraction of Ln(III) ions increases for lighter Ln, and then, with a break at Gd(III), decreases with increasing charge density.33 Therefore, the observed tendency appears to suggest that the coordination of CMPO becomes harder as the hydration energy of Ln(III) ions increases with increasing charge density. In particular, the lower extractability for Lu3+ could result from its more stable hydration which hindered the coordination of CMPO. This appears to be supported by Fig. 3, where only the extractability for Lu3+ increased with increasing concentration of nitric acid. A higher concentration of nitric acid would be required for the disruption of the stable hydration shells of Lu3+ to facilitate the coordination of CMPO. As seen from Fig. 5, three times the amount of PNIPAAm exhibited no significant difference in extractability. This indicates that almost all the Ln-CMPO complex was extracted with a smaller amount of PNIPAAm. The calculated molar ratios of CMPO against the extracted Ln ions were [CMPO]/[Ln]ext = 3.06 and 3.04 for the small and large amounts of PNIPAAm, respectively. It has been reported that Ln(III) ions in nitrate solutions are extracted to organic phases in the form of a 1:3 complex with CMPO, as presented in eqn (i).31,32,34–36
Ln3+ + 3CMPOorg + 3NO3− ⇄ Ln(NO3)3·(CMPOorg)3 | (i) |
Therefore, the obtained values of [CMPO]/[Ln]ext which are consistent with the stoichiometry of eqn (i) support that almost all CMPO formed a 1:3 complex with Ln(III) ions, which was extracted to the PNIPAAm solid phase. Furthermore, the extractability was almost constant, independent of the stirring time (0.5–15 h) in the complexation process (see Fig. S2 in the ESI†). This suggests that the complexation reaction proceeds fast enough.
The variation of extractability with increasing total molar ratios of [CMPO]/[Ln]total in the lower concentration region of each Ln ion ([Ln]: 0.0378–0.604 mmol L−1) is shown in Fig. 6. The molar ratios were controlled by adjusting the total metal concentration ([M]total) to the constant amount of CMPO (0.0272 mmol, 11.1 mg) in order to ensure the quantitative sufficiency of PNIPAAm for the extraction of all the Ln-CMPO complex species. The increase in extractability with increasing [CMPO]/[Ln]total ratio is attributed to the enhancement of complex formation. However, the extractabilities of La3+, Nd3+ and Eu3+ were almost saturated at molar ratios of more than 6. The maximum extraction percentages were ca. 74% for La3+, 76% for Nd3+, 68% for Eu3+ and 14% for Lu3+. Furthermore, the extraction selectivity of Ln ions became ambiguous compared with the extraction conducted in a large excess molar amount of Ln ions (Fig. 5). The results in Fig. 5 and 6 suggest that the extractability is largely affected by the concentration range of Ln ions. In the present extraction method, the complex formation proceeds in a heterogeneous system due to the water insolubility of CMPO, where the reaction efficiency of complexation is a significant factor to determine the extractability. Therefore, increasing not only the high relative molar ratios but also the large absolute amounts of metal ions or ligands would be required to increase the efficiency of complex formation. In particular, in terms of increasing the extraction percentage of Ln ions, a large excess amount of ligands should be required to enhance the complexation efficiency.
The stripping of Ln(III) ions was achieved by the complete re-dissolution of PNIPAAm co-aggregates (ion-adsorbing species) at 20 °C for 1 h in the presence of hydrochloric acid (1.0 mol L−1) and the subsequent phase separation between PNIPAAm co-aggregates (ion-desorbing species) and bulk solutions at 40 °C above the LCST. The extraction was carried out at [CMPO]/[Ln]total ratios of 3 and 6 ([HNO3] = 1.0 mol L−1). As seen from Fig. 7(a) ([CMPO]/[Ln]total = 6), no significant stripping was observed in the nitric acid solutions used for washing the co-aggregates after the extraction process. In contrast, hydrochloric acid achieved the stripping percentage of ca. 44% for La3+, 58% for Nd3+, 51% for Eu3+ and 5.7% for Lu3+ against the initial concentration of the test solution. The stripping for the extraction conducted at a [CMPO]/[Ln]total ratio of 3 also afforded Ln recoveries of ca. 36%, 49%, 43% and 4.6%, respectively, by using 1.0 mol L−1 hydrochloric acid (Fig. 7(b)). The stripping percentages converted against the extracted amounts are shown in Fig. 7(c). Except for Lu3+, the stripping effect of hydrochloric acid on each Ln ion was almost the same, independent of the extraction amounts, exhibiting the inverse tendency to the order of extraction and adsorption selectivity shown in Fig. 5. It has been reported that nitrate ions form an inner-sphere Ln-complex in moderate concentrations (2 mol L−1) of nitric acid solutions, while chloride ions coordinate to Ln(III) ions in the outer sphere even in concentrated hydrochloric acid (6 mol L−1).37 In hydrochloric acid solutions, therefore, the stability of the Ln-CMPO chloride complex would be lower than that of the nitrate complex for higher hydration energies of the Ln(III) ions (Eu3+ > Nd3+ > La3+). In addition, more hydrophilic chloride ions than nitrate ions might facilitate the exchange reaction between the complex and water. However, only Lu3+ with the highest hydration energy among the Ln ions considered deviated from the above tendency. This might be attributed to the small difference in the stability constants of nitrate and chloride complexes (although the stability constant of the Lu nitrate complex is larger than that of the chloride complex).38 The almost the same stripping efficiency observed for each Ln ion of different extraction amounts could result from the concentrations of the Ln ions stripped in hydrochloric acid solutions being not high enough to cause the quantitative difference in the formation of the Ln-CMPO chloride complex in the complexation equilibrium (note that the concentration region of the extracted Ln ions was almost comparable to that in the [CMPO]/[Ln]total ratio ranging from 6 to 12 in Fig. 6, where the extractability of each Ln ion has no large difference). From the above discussion, the stripping amount of Ln ions would reflect the difference in the stability and amounts between the nitrate complex formed in the extraction process and the chloride complex formed in the stripping process.
Footnote |
† Electronic supplementary information (ESI) available. See DOI: 10.1039/c7re00060j |
This journal is © The Royal Society of Chemistry 2018 |