Article pubs.acs.org/IC
Thermodynamic Insight into the Solvation and Complexation Behavior of U(VI) in Ionic Liquid: Binding of CMPO with U(VI) Studied by Optical Spectroscopy and Calorimetry Qi Wu,†,§ Taoxiang Sun,†,‡ Xianghai Meng,§ Jing Chen,†,‡ and Chao Xu*,†,‡ †
Collaborative Innovation Center of Advanced Nuclear Energy Technology, Institute of Nuclear and New Energy Technology and Beijing Key Lab of Radioactive Waste Treatment, Tsinghua University, Beijing 100084, China § State Key Laboratory of Heavy Oil Processing, China University of Petroleum, Beijing 102249, China ‡
S Supporting Information *
ABSTRACT: The complexation of U(VI) with octylphenyl-N,N-diisobutylcarbamoylmethylphosphine oxide (CMPO, denoted as L) in ionic liquid (IL) C4mimNTf2 was investigated by UV−vis absorption spectrophotometry and isothermal titration calorimetry. Spectro-photometric titration suggests that three successive complexes, UO2Lj2+ (j = 1−3), formed both in “dry” (water content < 250 ppm) and “wet” (water content ≈ 12 500 ppm) ionic liquid. However, the thermodynamic parameters are distinctly different in the two ILs. In dry IL, the complexation strength between CMPO and U(VI) is much stronger, with stability constants of the respective complexes more than 1 order of magnitude higher than that in wet IL. Energetically, the complexation of U(VI) with CMPO in dry IL is mainly driven by negative enthalpies. In contrast, the complexation in wet IL is overwhelmingly driven by highly positive entropies as a result of the release of a large amount of water molecules from the solvation sphere of U(VI). Moreover, comparisons between the fitted absorption spectra of complexes in wet IL and that of extractive samples from solvent extraction have identified the speciation involved in the extraction of U(VI) by CMPO in ionic liquid. The results from this study not only offer a thermodynamic insight into the complexation behavior of U(VI) with CMPO in IL but also provide valuable information for understanding the extraction behavior in the corresponding solvent extraction system.
■
INTRODUCTION Application of ionic liquid (IL) in the extraction and separation of metal ions, especially the radioactive metals relevant to nuclear fuel reprocessing, has been the subject of intense investigation over the years.1−3 Used as an alternative solvent to replace volatile and toxic organic compounds, IL offers several advantages in solvent extraction, for example, high extraction efficiency, nonvolatility, and low flammability. In combination with certain functional extractants, IL-based solvents have shown extremely high extraction ability to various metals. For example, a distribution ratio higher than 1 × 104 was achieved for the extraction of Sr2+ by a crown ether, dicyclohexanel-18crown-6 (DCH18C6), in the ionic liquid C2mimNTf2, far outperforming that in conventional diluents such as toluene and chloroform.4 The extraction efficiency of actinide and lanthanide ions such as UO22+, Am3+, and Eu3+ by octylphenylN,N-diisobutylcarbamoylmethylphosphine oxide (CMPO) was found to be enhanced greatly by using ionic liquids as the solvent when compared to n-dodecane.5−8 These extraordinary extraction performances highlight vast opportunities for the utilization of IL in separation applications. Apart from studies on the extraction performance, there are also increasing interests on revealing the mechanism associated with the © XXXX American Chemical Society
extraction process. Distinct ion exchange extraction mechanisms involving the exchange of extracted cationic or anionic complexes with the cation or anion of IL have been commonly found in IL-based extraction systems.9−13 Undoubtedly, both the high extraction ability and the distinct extraction mechanism of IL-based extraction system are closely relevant to the unique solvation environment of IL, which is supposed to facilitate the transfer and solvation of charged species. In this context, it is of great importance to study the solvation of metal ions and their complexes in IL. Fundamental knowledge of the solvation and complexation behavior of metal ions in IL would not only advance our understanding of the unique properties of IL but also could help us better reveal the mechanisms involved in IL-based separation systems.14−18 In the past decade, there are already a number of studies concerning the complexation of various metal ions, such as U(VI),19−28 Nd(III),29−31 Eu(III),19,32,33 and Ni(II),34 etc., in IL. A variety of experimental techniques, including optical spectroscopy, extended X-ray absorption fine structure (EXAFS), mass spectroscopy, as well as computational Received: December 22, 2016
A
DOI: 10.1021/acs.inorgchem.6b03132 Inorg. Chem. XXXX, XXX, XXX−XXX
Inorganic Chemistry
■
methods, have been employed to investigate the complexation behavior of metal ions in IL, and informative findings have been obtained. Among them, the complexation of lanthanide and actinide ions, especially the uranyl ion,19−28 with various inorganic anions such as nitrate and chloride has been the subject of most studies. Various complexes formed between U(VI) and nitrate in IL were identified spectro-photometrically, and the coordination mode was characterized by EXAFS. The results from these studies show that the solvation environment of IL facilitates the successive complexation of U(VI) with nitrate, with ionic species up to UO2(NO3)3−,21,23−25 and in a few cases UO2(NO3)42−,26,28 could be formed in IL, which otherwise would be difficult to access in aqueous solutions due to strong hydration effect. However, in contrast to a relatively large number of studies on the complexation of metal ions with simple inorganic anions in IL, fewer studies concern about their complexation with more complicated organic ligands in a homogeneous IL medium.21,27,33 The complexation of metal ion with organic ligands, particularly the so-called extractants used in solvent extraction, is of great significance, since it could provide direct information to help us better understand the corresponding extraction behavior, and thus we can optimize and design more efficient extraction systems. Moreover, there are also quite fewer studies to reveal the complexation behavior of metal ions in IL from a thermodynamic point of view through the determination of thermodynamic parameters such as stability constant, enthalpy, and entropy, all of which are very helpful to energetically understand the complexation and interaction behavior in IL.23,29−31,33,34 In this work, we studied the complexation of CMPO (Scheme 1), a bifunctional organophosphorous compound that
Article
EXPERIMENTAL SECTION
Chemicals. The IL C4mimNTf2 with a purity of greater than 99% was provided by Lanzhou Institute of Chemical Physics, Chinese Academy of Sciences. Two sets of ionic liquids with different content of water were prepared. The first set was regarded as “dry” IL after drying it at 120 °C under vacuum in a heating oven for 24 h and then was kept in a water-free (99%) was purchased from J&K Scientific Ltd., China. Uranyl nitrate hexahydrate (UO2(NO3)2· 6H2O; >98.5%) was obtained from the stockroom at Institute of Nuclear and New Energy Technology, Tsinghua University, China. Uranium trioxide (UO3) was prepared from the thermal decomposition of UO2(NO3)2·6H2O. UO2(NTf2)2 was synthesized from the reaction of HNTf2 with UO3 according to published reports.21,23 Stock solutions of CMPO and uranyl ion in IL were prepared by weighing certain amounts of the respective chemicals and dissolving them in the IL. The chemical structures of C4mimNTf2 and CMPO are shown in Scheme 1. Spectro-photometric Titration. All experiments were performed at 25 °C. The UV−vis absorption spectra of U(VI) complexes were collected over the wavelength range from 350 to 600 nm (0.5 nm interval) on an Agilent Cary6000i UV−vis−NIR spectro-photometer in a rectangular cuvette of 10 mm path length with Teflon stopper. The reference solution was pure C4mimNTf2 solution. The temperature of the sample and reference cell holders was controlled via an external temperature controller, which drives a Peltier heat pump attached to holders. In a typical titration, 2.0 mL of 10 mM U(VI) solution was placed in a 10 mm cuvette, into which appropriate aliquots of 0.14 M CMPO solutions were added and mixed thoroughly for 10 min before the spectrum was collected. Preliminary kinetic experiments showed that the complexation reaction was fast, and the absorbance became stable within a few minutes. Generally, ∼30 additions were made in each titration. The stability constants of U(VI)/CMPO complexes were calculated by the nonlinear regression program HypSpec.36 Solvent Extraction. The extractions of U(VI) from aqueous phase (HNTf2 or HNO3 solutions) by CMPO in C4mimNTf2 and ndodecane were performed, and the UV−vis spectra of the organic phases after extraction were recorded. The experiments were conducted by contacting an aqueous solution of U(VI) and an organic phase of CMPO with equal volumes in a stoppered glass tube. In addition to CMPO, the n-dodecane solution also contained 1 M tributyl phosphate (TBP) as a phase modifier to avoid third phase formation. The two phases were mixed in a thermostated water bath equipped with electromagnetic stirring for 15 min, then centrifuged, and the organic phases were taken for UV−vis absorption measurement. Calorimetric Titration. Calorimetric titrations were conducted at 25 °C to determine the enthalpy of complexation with an isothermal micro-calorimeter (model TAMIII, TA Instruments). The calorimeter was calibrated by the built-in dynamic calibration method and was tested against the IUPAC recommended complexation of BaCl2 with 18-crown-6-ether at 25 °C.37 In all titrations, 0.700 mL of the U(VI) solution was put in a 1.0 mL reaction vessel, and ∼35 additions of 0.005 mL of titrant (CMPO solution) were made automatically through a 0.250 mL titration syringe, with a duration of 12 min between each addition, resulting in 35 experimental values of total heat (Qex,j, j = 1−35). These values were corrected by the heats of titrant dilution (Qdil,j) that were measured in a separate run. The net reaction heat at the jth point (Qr,j) was obtained from the difference: Qr,j = Qex,j − Qdil,j. The value of Qr,j is a function of the concentrations of the reactants (CU and CL), the equilibrium constants, and the enthalpies of the reactions that occurred in the titration. The net reaction heats were used, in conjunction with the equilibrium constants obtained by
Scheme 1. Chemical Structures of C4mimNTf2 and CMPO
is the preferred extractant in TRUEX process for the partitioning of minor actinides from high-level liquid waste,35 with uranyl ion, one of the most important actinide ions in nuclear industry, in a homogeneous IL medium (C4mimNTf2, Scheme 1). The complexes formed between CMPO and U(VI) were identified, and their stability constants were obtained by spectrophotometry. Moreover, the absorption spectra of U(VI)/CMPO complexes formed in the homogeneous IL medium were compared with that of IL samples from biphasic solvent extraction, through which the extracted species was identified. The enthalpies of complexation were determined by isothermal calorimetry, and the entropies were calculated accordingly. Besides, we examined the effect of water content in IL on the solvation and complexation behavior of U(VI). Remarkable influence of water on the thermodynamic parameters was observed, and the role of water in the complexation process was discussed. B
DOI: 10.1021/acs.inorgchem.6b03132 Inorg. Chem. XXXX, XXX, XXX−XXX
Article
Inorganic Chemistry spectrophotometry, to calculate the enthalpies of complexation with the HypDeltaH program.36 For the calorimetric titration with dry IL, special measures were taken to prevent the adsorption of water into the sample in the reaction vessel. The air in the lab was dehumidified by an air conditioner, and the humidity in the environment was maintained at a relatively low level during the titration. A blank titration by adding dry IL into dry IL was conducted, and the water content in the IL sample before and after titration was found to increase by less than 10%, which is supposed to have little influence on the interpretation of the calorimetric data.
increased, which demonstrate the gradual complexation of U(VI) with CMPO. To have a clearer view of the variation of the absorptions, the changes of spectra in Figure 1a were arbitrarily grouped into the following three stages. In the first stage (CL/CU = 0−2, Figure 1b), the absorption bands show the most significant changes. A series of well-defined and sharper absorption peaks appeared in the spectra, and the absorption was intensified and red-shifted simultaneously. In the second stage (CL/CU = 2−3.5, Figure 1c), the absorption bands between 400 and 450 nm were further red-shifted but with a concurrent decrease in the absorbance, while the absorbance of four smaller peaks between 450 and 500 nm was intensified as well as red-shifted. In the third stage (CL/CU = 3.5−5, Figure 1d), fewer changes were observed, and the absorbance just decreased slightly due to the dilution effect, suggesting that there was very few new complexes formed in the solution. For the titration in dry IL, the changes of the spectra (Figure S1 in the Supporting Information) show a similar trend, but lower CL/CU ratio was required for the transition of different stages. Such an observation indicates that stronger complexes might be formed in dry IL than in wet IL, which is confirmed by the higher stability constants of complexes in dry IL as determined by fitting of the spectra. The spectra were analyzed and fitted by the nonlinear regression program HypSpec. The best fit of the spectra was achieved by assuming successive formation of three U(VI)/L complexes, UO2L2+, UO2L22+, and UO2L32+ during the titration. Figure 2 shows the diagram for speciation change of U(VI) in wet IL and the evolving of the absorbance at a representative wavelength (425 nm) during the titration, as well as the corresponding molar absorptivity of each complex. The evolving of the absorbance at 425 nm is in good agreement with the change of speciation. The diagram of speciation change and molar absorptivities of complexes in dry IL are shown in the Supporting Information (Figure S2). The stability constants of the complexes formed both in wet and dry IL are all listed in Table 1.
■
RESULTS AND DISCUSSION Spectrophotometry: Species Identification and Stability Constants Determination. In the visible region, U(VI) has characteristic absorption bands that are sensitive to the change of coordination environment and thus can be used to effectively probe its complexation with CMPO. Figure 1 shows
Figure 1. Spectro-photometric titration of U(VI) with CMPO in wet IL. Initial solution: V0 = 2.0 mL, CU = 10 mM, titrant: CL = 0.14 M, 0.700 mL added. (a) Full titration spectra and (b−d) spectra within different CMPO/U(VI) ratios.
Table 1. Thermodynamic Parameters of the Complexation of U(VI) with CMPO in Ionic Liquid
the representative UV−vis spectro-photometric titrations of U(VI) with CMPO in wet IL. As shown in Figure 1, in the absence of CMPO ligand, U(VI) in the IL has a series of absorption bands, with several main peaks sitting at ∼404, 415, 428 nm, etc. These absorption bands correspond well to those of U(VI) in aqueous perchlorate media, where the uranyl ion is supposed to coordinate with five water molecules in the equatorial plane.38,39 Significant changes in the spectra were observed when CMPO was introduced and the ratio of CL/CU
complex dry IL
wet IL
2+
UO2L UO2L22+ UO2L32+ UO2L2+ UO2L22+ UO2L32+
log β 6.09 10.7 14.1 4.61 8.78 11.8
± ± ± ± ± ±
0.05 0.1 0.2 0.03 0.06 0.2
ΔH (kJ mol−1) −26.1 −46 −61 12.1 31.2 17
± ± ± ± ± ±
0.8 2 2 0.9 0.9 2
ΔS (J mol−1 K−1) 29 51 65 129 273 283
± ± ± ± ± ±
4 8 11 4 5 11
Figure 2. Change of speciation of U(VI) (solid lines) and the absorbance at 425 nm (▲) as a function of the CL/CU ratio in wet IL (left) and calculated molar absorptivity of U(VI) complexes (right). C
DOI: 10.1021/acs.inorgchem.6b03132 Inorg. Chem. XXXX, XXX, XXX−XXX
Article
Inorganic Chemistry As shown in Figure 2 (right), all the three U(VI)/CMPO complexes in wet IL show more intensive and sharper absorption bands when compared to those of the “free” uranyl ion, which has relatively weak absorptions with several smooth peaks in the 350−500 nm region. The better-defined vibronic fine structure and sharper absorption bands of U(VI) are clear evidence that the uranyl ions complexed with CMPO in IL were in an ordered ligand field.21 Besides, there are a couple of other features regarding the respective absorptivity of the complexes worth noting. First, all the absorption peaks of U(VI) were red-shifted upon the successive complexation with CMPO. Fox example, the four main absorption bands at 398.5, 409.7, 422.0, and 434.5 nm for the 1:1 complex UO2L2+ were shifted to 401.8, 413.3, 425.3, and 438.2 nm for UO2L22+ and then to 403.8, 415.4, 427.6, and 440.3 nm for UO2L32+, respectively. Second, the absorption branches from 450 to 500 nm for the three U(VI)/CMPO complexes are quite different from each other. The spectrum of the 1:3 complex UO2L32+ shows four relatively well-defined peaks from 450 to 500 nm, while those of the 1:1 and 1:2 complexes exhibit less featured absorption peaks in the same region. It was reported that the intensity and position of the vibronic spectra of U(VI) are mainly governed by the geometry of equatorial coordination and only to a lesser extent by the chemical nature of the ligands.21,40 Therefore, the changes of the absorption bands discussed above reflect the variation of the symmetry and geometry of the coordination sphere of uranyl ion upon complexation with CMPO. The absorption spectra of U(VI)/ CMPO complexes in dry IL also show similar trends (Figure S2 in the Supporting Information). Since the spectra of U(VI) complexes reflect the coordination environment of U(VI), the well-fitted molar absorption spectra of the individual U(VI)/CMPO complexes could provide useful information to identify the extractive species for the extraction of U(VI) by CMPO in ionic liquids. As mentioned previously, the extractions of U(VI) as well as other actinides by CMPO in IL have been studied, and extremely high extraction efficiencies were obtained.5−8 However, controversy still remains on what kind of U(VI) species are involved in the extraction of U(VI) from nitric acid medium by CMPO in IL.41−43 For example, Visser et al. suggested a net stoichiometry of UO2(NO3)L+ in the extraction systems using C4mimPF6 and C8mimNTf2, compared to UO2(NO3)2L2 in n-dodecane, by EXAFS measurement.41 More recently, another work even proposed a stoichiometry of UO2(NO3)2L6 for the extracted species in C2mimNTf2 by slope analysis.43 To help further reveal the extraction species in the uranyl-CMPO-IL biphasic extraction system, we compared the fitted spectra of U(VI) complexes in wet IL with those of so-called extractive samples (Figure 3). Since the IL phase is also wet in solvent extraction after contacting with the aqueous phase, it would be more meaningful to compare the fitted spectra in wet IL rather than in dry IL with those of the extractive samples. As shown in Figure 3, the spectra (a−c) of three IL phases after extraction of 10 mM U(VI) from nitric acid media by different concentrations of CMPO show almost identical absorption bands in terms of the position and relative intensity of each absorption peak, indicating that same extractive species were formed in the IL phase regardless of the initial CMPO/ U(VI) ratio before extraction. In contrast, the spectrum of the organic phase after extraction of U(VI) by CMPO in ndodecane (g in Figure 3) is very different from these spectra in
Figure 3. Absorption spectra of U(VI) complexes. (a−c) Spectra of the IL phases after extraction of uranyl nitrate ([U]initial = 10 mM) from 0.1 M HNO3 medium by 10, 20, 50 mM CMPO, respectively. (d, e) Fitted spectra of species UO2L22+ and UO2L32+, respectively, in wet IL. (f) IL phase after extraction of UO2(NTf2)2 ([U]initial = 10 mM) from 0.1 M HNTf2 medium by 50 mM CMPO. (g) Organic phase after extraction of uranyl nitrate ([U]initial = 10 mM) by 50 mM CMPO in n-dodecane (with 1 M TBP as phase modifier).
ILs, demonstrating that U(VI)/CMPO complexes in ILs and ndodecane are not equivalent. In fact, the extractive species in ndodecane system has been identified as UO2(NO3)2L2 from previous work.44 The spectra (a−c) of the extractive IL phases were further compared with the fitted absorption spectra of species in the homogeneous wet IL. Clearly, the spectra (a−c) are quite similar to that of UO2L32+ (e in Figure 3), evidenced by the same positions of the main peaks from 400 to 450 nm and the almost identical absorption patterns from 450 to 500 nm, while they are apparently different from that of UO2L22+ (d in Figure 3) and UO2L2+ (not shown in Figure 3). In this respect, we postulate that UO2L32+ is the most likely extractive species in the extraction of U(VI) from nitric acid media by CMPO in C4mimNTf2. However, this is in contrast to the extractive species like UO2(NO3)L+ or UO2(NO3)2L6 as implied in similar extraction systems from previous studies.41,43 We could regard the species UO2(NO3)2L6 as unreasonable, since it incorporates too many and overcrowded CMPO ligands in the complex. But the discrepancy between the extractive species UO2L32+ and UO2(NO3)L+ is still quite pronouncing, even after taking consideration of a few differences in extraction conditions (e.g., type of IL, concentration of HNO3, etc.) among the studies. To further reveal the U(VI)/CMPO speciation involved in the extraction, we conducted another extraction experiment by excluding any nitrate ion in the solutions. The extraction of UO2(NTf2)2 from 0.1 M HNTf2 solution by CMPO in C4mimNTf2 was performed, and the IL phase after extraction was analyzed. Interestingly, the spectrum of this IL phase (f in Figure 3) is almost identical to those after extraction from nitric acid media (a−c in Figure 3) and the spectrum of UO2L32+ (e in Figure 3). Such a similarity suggests that (1) nitrate ion may not be involved directly in the primary coordination sphere of U(VI) in the IL phase for the extraction of U(VI) from HNO3 medium by CMPO in IL as suggested by previous work41,42 and (2) UO2L32+ is the most likely extractive species for the extraction of U(VI) from both HNO3 and HNTf2 media by CMPO in C4mimNTf2. Micro-calorimetry: Determination of Enthalpies. Figure 4 shows representative micro-calorimetric titrations of D
DOI: 10.1021/acs.inorgchem.6b03132 Inorg. Chem. XXXX, XXX, XXX−XXX
Article
Inorganic Chemistry
Figure 4. Micro-calorimetric titrations of U(VI)/CMPO complexation in dry and wet IL. (left) Titration thermograms. (right) Cumulative reaction heat (◇ = experimental and + = calculated) and U(VI)/CMPO species as a function of the volume of titrant added. Initial solution: 0.700 mL of 10 mM UO2(NTf2)2, titrant: 0.14 M CMPO.
content of less than 250 ppm (
