Characterization of the ALSEP Process at Equilibrium: Speciation and Stoichiometry of the Extracted Complex.

We have determined the identity of the complexes extracted into the ALSEP process solvent from solutions of nitric acid. The ALSEP process is a new solvent extraction separation designed to separate americium and curium from trivalent lanthanides in irradiated nuclear fuel. ALSEP employs a mixture of two extractants, 2-ethylhexyl phosphonic acid mono-2-ethylhexyl ester (HEH[EHP]) and N,N,N',N'-tetra(2-ethylhexyl)diglycolamide (TEHDGA) in n-dodecane, which makes it difficult to ascertain the nature of the extracted metal complexes. It is often asserted that the weak acid extractant HEH[EHP] does not participate in the extracted complex under ALSEP extraction conditions (2-4 M HNO3). However, the analysis of the Am extraction equilibria, Nd absorption spectra, and Eu fluorescence emission spectra of metal-loaded organic phases argues for the participation of HEH[EHP] in the extracted complex despite the high acidity of the aqueous phases. The extracted complex was determined to contain fully protonated molecules of HEH[EHP] with an overall stoichiometry of M(TEHDGA)2(HEH[EHP])2·3NO3. Computations also demonstrate that replacing one TEHDGA molecule with one (HEH[EHP])2 dimer is likely energetically favorable compared to Eu(TEHDGA)3·3NO3, whether the HEH[EHP] dimer is monodentate or bidentate.

Introduction

Implementation of advanced nuclear fuel cycles is critical to plans for sustainable use of nuclear energy. Proposed closed nuclear fuel cycles based on partitioning and transmutation will recycle hemical">uranium hemical">and hemical">an class="Chemical">plutonium from used nuclear fuel and then separate and transmute minor actinide elements to short-lived nuclides in fast reactors. The separation and recycle of uranium and plutonium by the plutonium uranium reduction extraction (PUREX) process is well-studied, but efficient chemical separation of the trivalent minor actinides, americium and curium, from the lanthanide-rich matrix the PUREX process leaves behind is difficult due to the physical and chemical similarities of the trivalent lanthanide (Ln) and actinide (An) ions.[1,2]

In the past 60 years, a number of solvent extraction systems for hemical">actinide–hemical">an class="Chemical">lanthanide separations have been proposed.[3−9] Recently, a three-step solvent extraction separation, the ALSEP (actinide lanthanide separation) process, has been proposed to simplify isolation of americium and curium from the lanthanides. In ALSEP, the An(III) and Ln(III) cations are extracted together from aqueous nitric acid solutions into an n-dodecane organic phase containing two potential ligands, an acidic dialkylorganophosphorus extractant, and a neutral diglycolamide extractant.[10,11] Although several ALSEP formulations have been considered, the most effective formulation uses a mixture of the extractants 2-ethylhexyl phosphonic acid mono-2-ethylhexyl ester (HEH[EHP]) and N,N,N′,N′-tetra(2-ethylhexyl)diglycolamide (TEHDGA) in n-dodecane (Figure ).[12] Following extraction of the trivalent lanthanide and actinide cations from ca. 3 M HNO3, the organic phase is contacted with a scrub solution to adjust the acidity and remove minor impurities. Then, the americium and curium are selectively stripped from the scrubbed organic phase by the addition of a polyaminocarboxylic acid such as diethylenetriaminepentaacetic acid (DTPA) or N-(2-hydroxyethyl)ethylenediaminetriacetic acid (HEDTA) to the aqueous phase. The ALSEP process greatly simplifies the separation of trivalent actinides by combining the partitioning of trivalent f-elements and the An/Ln separation processes into a single separation cycle, allowing direct use of PUREX raffinate solutions, exhibiting fast extraction rates, and performing robustly under a broad range of process conditions.[13,14]

Figure 1

Chemical structures for (A) 2-ethylhexyl phosphonic acid mono-2-ethylhexyl ester (HEH[EHP]) and (B) N,N,N’,N’-tetra(2-ethylhexyl) diglycolamide (TEHDGA), the two extractants used in combination in the ALSEP organic phase. Structures of additional extractants used in the computational analyses of this study and in related liquid–liquid extraction processes are presented in Figure S1.

Chemical structures for (A) hemical">2-ethylhexyl hemical">an class="Chemical">phosphonic acid mono-2-ethylhexyl ester (HEH[EHP]) and (B) N,N,N’,N’-tetra(2-ethylhexyl) diglycolamide (TEHDGA), the two extractants used in combination in the ALSEP organic phase. Structures of additional extractants used in the computational analyses of this study and in related liquid–liquid extraction processes are presented in Figure S1.

Our research aims to develop a kinetic model of the processes that underlie hemical">an class="Chemical">americium and curium stripping in ALSEP, but probing the kinetics of this complex multiextractant system requires an understanding of the metal complexes present in the bulk phases at equilibrium. This work begins that process by dissecting the speciation of the organic phase metal complexes in the ALSEP extraction step.

The complexes formed in solutions containing only a hemical">diglycolamide or hemical">an acidic hemical">an class="Chemical">dialkylorganophosphorus extractant have been studied extensively,[12,15−29] but only a few published works have investigated the bulk-phase chemical speciation of trivalent actinide or lanthanide complexes in the presence of both diglycolamide and acidic dialkylorganophosphorus extractants.[12,18,30−32] These studies suggest ternary metal–extractant complexes are present in the organic phase under certain conditions, but they either employ different extractants (for example TODGA rather than TEHDGA or the phosphoric acid HDEHP rather than the phosphonic acid HEH[EHP]), examine pH ranges that are not relevant to extracting conditions, or simulate organic phase conditions using solutions that are not in equilibrium with an aqueous phase. Unsurprisingly, these studies do not entirely agree. Some hypothesized that a ternary metal–diglycolamide–organophosphorus complex is present under all conditions[33,34] while others reported that a mixed-extractant complex exists only under specific conditions.[12,31,32]

Understhemical">anding the speciation of the hemical">an class="Chemical">ALSEP process extraction step is important to the accurate interpretation of kinetic data and thus to pinpointing the underlying processes that lead to extraction and ultimately limit the actinide stripping rate. To resolve the ambiguities of the earlier reports, we explore the composition and stoichiometry of the equilibrium f-element complexes extracted into TEHDGA/HEH[EHP]/n-dodecane from nitric acid solutions under conditions directly relevant to the ALSEP process. The equilibrium organic phase species are characterized through equilibrium analysis, UV–vis and time-resolved laser-induced fluorescence spectroscopy (TRLFS), and density functional theory (DFT) calculations to probe the metal–ligand geometries and orientation and identify the thermodynamically stable metal-containing species present at equilibrium in the ALSEP metal extraction step.

Results

Optical Spectroscopy of Loaded Organic Phases

The coordination environment of the organic phase complexes formed by extraction into the hemical">ALSEP process solvent was probed by optical spectroscopy. hemical">an class="Chemical">Lanthanides were loaded into organic phases for spectroscopic measurements from aqueous phases containing 0.01 M Nd or Eu. Neodymium is a particularly useful surrogate for Am because the ionic radii of Am(III) and Nd(III) are nearly identical,[35] and the hypersensitive optical transitions of Nd can be useful for identifying changes in the metal coordination sphere.[36] Europium(III), on the other hand, is isoelectronic with Am(III) (4f6 vs 5f6) and displays strong and sensitive fluorescence in the visible region.[37] Extractions were performed as closely as possible to ALSEP process relevant conditions (3–4 M HNO3 and 0.05 M HEH[EHP]/0.75 M HEH[EHP]), but the need to balance adequate extraction while avoiding third-phase formation meant using acid concentrations lower than the ideal extraction conditions for ALSEP in some measurements.

hemical">Neodymium absorption spectra for the 4I9/2 → 4G5/2, 2G7/2 trhemical">ansition between 560 hemical">and 620 nm, the 4I9/2 → 4S3/2, 4F7/2 trhemical">ansition between 720 hemical">and 770 nm, hemical">and the 4I9/2 → 4F5/2, 2H9/2 trhemical">ansition between 775 hemical">and 825 nm are summarized in Figure . The spectra represent Nd extracted by 0.75 M hemical">an class="Chemical">HEH[EHP]/n-dodecane, 0.1 M TEHDGA/n-dodecane, or the ALSEP organic phase, 0.05 M TEHDGA/0.75 M HEH[EHP]/n-dodecane. The extraction of Nd into the ALSEP organic phase from 2 M HNO3 (not shown), 3 M HNO3 (not shown), and 4 M HNO3 produced identical spectra, but these spectra of the ALSEP organic phase are different from those of the organic phases containing only 0.1 M TEHDGA/n-dodecane or 0.75 M HEH[EHP]/n-dodecane. The most notable differences between the Nd-TEHDGA and the Nd-ALSEP spectra are diminished absorbances at 575 and 591.5 nm in the TEHDGA sample relative to the ALSEP sample and enhanced absorbance at 735 nm for the TEHDGA sample. Furthermore, a general narrowing of the peaks occurs in the TEHDGA sample, particularly those centered around 802 and 587 nm, which is accompanied by a slight blue shift at 582 and 587 nm.

Figure 2

UV–vis spectra of the neodymium 4I9/2→ 4G5/2, 2G7/2 (560–620 nm), 4I9/2 → 4S3/2, 4F7/2, and 2H9/2, 4F5/2 (700–850 nm) transitions after extraction into 0.75 M HEH[EHP]/n-dodecane from 0.001 M HNO3/1 M NaNO3 (short dash), 0.1 M TEHDGA/n-dodecane from 3.5 M HNO3 (short dot), and 0.75 M HEH[EHP]/0.05 M TEHDGA/n-dodecane (ALSEP) from 4 M HNO3 (solid line).

UV–vis spectra of the hemical">neodymium 4I9/2→ 4G5/2, 2G7/2 (560–620 nm), 4I9/2 → 4S3/2, 4F7/2, hemical">and 2H9/2, 4F5/2 (700–850 nm) trhemical">ansitions after extraction into 0.75 M hemical">an class="Chemical">HEH[EHP]/n-dodecane from 0.001 M HNO3/1 M NaNO3 (short dash), 0.1 M TEHDGA/n-dodecane from 3.5 M HNO3 (short dot), and 0.75 M HEH[EHP]/0.05 M TEHDGA/n-dodecane (ALSEP) from 4 M HNO3 (solid line).

While the 0.1 M hemical">TEHDGA hemical">and 0.05 M hemical">an class="Chemical">TEHDGA/0.75 M HEH[EHP] spectra are similar, each organic phase composition investigated exhibits a unique spectrum, indicating that different average Nd coordination environments are encountered in each organic phase. The principal component analysis of the spectra and target transformation using the program SIXpack[38] indicate the Nd absorption spectrum of the two-extractant TEHDGA/HEH[EHP] ALSEP system does not arise from a linear combination of the Nd spectra observed for the single-extractant TEHDGA/n-dodecane and HEH[EHP]/n-dodecane systems. Moreover, peaks at 570 and 606 nm are absent in the spectrum of the ALSEP system, indicating that Nd{H(EH[EHP])2}3 is not present. Although the presence of an equilibrium mixture of multiple Nd complexes in the ALSEP system cannot be excluded based on these absorption spectra, the results of the principal component analysis imply the presence of at least one Nd complex in the ALSEP organic phase that is not found in either of the two single-extractant systems.

The TRLFS emission spectra of Eu extracted in the three different extraction systems (Figure ) more clearly demonstrate that the hemical">metal coordination environment of Eu in the hemical">an class="Chemical">ALSEP system is distinctly different from the complexes in the solutions containing only TEHDGA or HEH[EHP]. Each solution displays distinctly different fluorescence spectra from the other two systems, and the spectrum of the ALSEP system is obviously not related to the spectra of either of the single extractant systems.

Figure 3

TRLFS spectra of europium extracted from 0.001 M HNO3/1 M NaNO3 into 0.75 M HEH[EHP]/n-dodecane (dashed line), from 3.5 M HNO3 into 0.1 M TEHDGA/n-dodecane (dotted line), and from 4 M HNO3 into 0.05 M TEHDGA/0.75 M HEH[EHP]/n-dodecane (solid line), respectively.

TRLFS spectra of hemical">europium extracted from 0.001 M hemical">an class="Chemical">HNO3/1 M NaNO3 into 0.75 M HEH[EHP]/n-dodecane (dashed line), from 3.5 M HNO3 into 0.1 M TEHDGA/n-dodecane (dotted line), and from 4 M HNO3 into 0.05 M TEHDGA/0.75 M HEH[EHP]/n-dodecane (solid line), respectively.

The emission spectra of all three complexes were collected over a wavelength range spanning the 5D0 → 7F transitions (J = 0–4), and three key differences are apparent in the spectra. First, a significant intensity for the 5D0 → 7F0 transition at 579.8 nm is only observed for the Eu-hemical">ALSEP system, hemical">and only one peak is seen for this nondegenerate trhemical">ansition. Second, while the 5D0 → 7F1 trhemical">ansitions at 593 nm are similar for both the hemical">an class="Chemical">ALSEP and TEHDGA/n-dodecane systems, as expected for this magnetic dipole transition, the intensities of the hypersensitive 5D0 → 7F2 manifold (618 nm) are substantially different for each of the solutions, with the Eu-ALSEP system displaying a more intense 5D0 → 7F2 emission than the Eu-TEHDGA/n-dodecane system. Third, the band shapes and barycenters of the 5D0 → 7F4 transition at ca. 700 nm are distinct for each organic phase composition. Together the spectral differences in the TRLFS data indicate that the Eu inner-sphere coordination is different in each extraction system and that the ALSEP organic phase contains a single Eu coordination environment that does not match the Eu coordination in the organic phases containing only TEHDGA or HEH[EHP].

Probing Water Coordination by TRLFS

hemical">Water molecules have been proposed as inner-sphere lighemical">ands in the orghemical">anic phase complexes of Eu in a mixed hemical">an class="Chemical">malonamide/HDEHP extraction system.[39] The possible presence of water in the inner coordination sphere of the extracted complexes in the ALSEP process was investigated by measuring the lifetimes of Eu3+ samples prepared in light water and identical heavy water samples made with deuterated reagents. The presence of OH oscillators directly coordinated with europium decreases the emission decay lifetime (τ = 0.11 ms, aqueous Eu solution) significantly when compared to heavier OD oscillators (τ = 3.3 ms).[40] The fluorescence lifetime of the major fluorescence peaks in each solution could be fit to a single exponential decay with χ2 < 1 and fluorescence lifetimes on the order of milliseconds (Table ). As expected, Eu3+ in the aqueous stock solutions was found to contain nine water molecules directly coordinated in the inner coordination sphere.[41] In contrast, all the organic-phase Eu–ligand complexes we studied exhibited minimal inner-sphere water molecules as shown in Table .

Table 1

Fluorescence Lifetimes in H2O and D2O for Eu Complexesa

complex	τ in H2O (ms)	τ in D2O (ms)	number of H2O
aqueous Eu solution	0.11	3.30	9.20
0.05 M TEHDGA/0.75 M HEH[EHP]	2.17	2.26	0.02
0.1 M TEHDGA	2.10	2.35	0.05
0.75 M HEH[EHP]	3.03	3.27	0.03

Samples prepared by extracting 0.01 M Eu from 0.001 M HNO3/1 M NaNO3, 3.5 M HNO3, and 4 M HNO3 into 0.75 M HEH[EHP]/n-dodecane, 0.1 M TEHDGA/n-dodecane, and 0.05 M TEHDGA/0.75 M HEH[EHP]/n-dodecane, respectively. Number of inner-sphere water molecules calculated based on eq , with an absolute uncertainty of ±0.5 H2O.

Samples prepared by extracting 0.01 M Eu from 0.001 M hemical">HNO3/1 M NaNO3, 3.5 M hemical">an class="Chemical">HNO3, and 4 M HNO3 into 0.75 M HEH[EHP]/n-dodecane, 0.1 M TEHDGA/n-dodecane, and 0.05 M TEHDGA/0.75 M HEH[EHP]/n-dodecane, respectively. Number of inner-sphere water molecules calculated based on eq , with an absolute uncertainty of ±0.5 H2O.

Liquid–Liquid Extraction Equilibria

hemical">An hemical">analysis of the partitioning behavior of hemical">an class="Chemical">Am3+ or Nd3+ between aqueous solutions of nitric acid and organic phases consisting of TEHDGA, HEH[EHP], and n-dodecane was used to define the stoichiometries of the organic-phase complexes in the ALSEP system.

TEHDGA Dependence

hemical">An hemical">an class="Chemical">americium extraction from either 2 or 4 M HNO3 into organic phases containing 0.75 M HEH[EHP] and varying amounts of TEHDGA in n-dodecane was examined. For comparison, additional measurements of the Am extraction from 2 M HNO3 into n-dodecane containing only variable concentrations of TEHDGA were also evaluated. The extraction of Am from 4 M HNO3 into TEHDGA/n-dodecane in the absence of HEH[EHP] could not be studied due to the formation of a third phase under these conditions.[42]

The results of the three hemical">TEHDGA dependence experiments are summarized in Figure hemical">and Table . One set of extractions was conducted in the absence of hemical">an class="Chemical">HEH[EHP], while two sets were conducted in the presence of 0.75 M HEH[EHP]. The addition of 0.75 M HEH[EHP] to the TEHDGA solution caused a substantial increase in Am extraction, and the highest DAm values were always observed in the 4 M HNO3–TEHDGA/0.75 M HEH[EHP]/n-dodecane system. In the solutions containing only TEHDGA, the slope analysis of the Am extraction from 2 M HNO3 indicates a third power dependence on the TEHDGA concentration (slope = 2.95 ± 0.09, Table ), implying the formation of a 1:3 Am:TEHDGA complex in the organic phase, as previously reported for the extraction of Eu by TEHDGA[20] and An3+/Ln3+ extraction by TODGA.[19,28] The ALSEP organic phases, on the other hand, contain 0.75 M HEH[EHP] in addition to TEHDGA. In the ALSEP organic phases, the slopes of the extraction curves decrease substantially from the TEHDGA-only case, giving second-power dependence on the TEHDGA concentration at both acidities (2.11 ± 0.07 for 2 M HNO3 and 2.1 ± 0.1 for 4 M HNO3, Table ). This implies an average 1:2 Am:TEHDGA stoichiometry for the extracted complexes at both acidities when 0.75 M HEH[EHP] is also present in the organic phase. Consequently, one TEHDGA molecule is displaced from the equilibrated extracted complex when 0.75 M HEH[EHP] is added to the TEHDGA/n-dodecane organic phase to create the ALSEP extraction system.

Figure 4

TEHDGA dependence for the extraction of Am3+ from 4 M HNO3 into 0.03–0.075 M TEHDGA/0.75 M HEH[EHP]/n-dodecane (filled square), from 2 M HNO3 into 0.02–0.07 M TEHDGA/0.75 M HEH[EHP]/n-dodecane (empty square), and from 2 M HNO3 into 0.02–0.2 M TEHDGA/n-dodecane (filled diamond).

Table 2

Stoichiometric Coefficients Determined by Linear Regression Analysis of Logarithmic Am Extraction Data

aqueous phase	organic phase [TEHDGA]	organic phase [HEH[EHP]]	slope	intercept
2 M HNO3	0.02–0.2 M	0 M	2.95 ± 0.09	3.5 ± 0.1
2 M HNO3	0.02–0.07 M	0.75 M	2.11 ± 0.07	3.6 ± 0.1
4 M HNO3	0.03–0.075 M	0.75 M	2.1 ± 0.1	4.4 ± 0.1
2 M HNO3	0.05 M	0.005–0.75 M	0.43 ± 0.02	0.80 ± 0.02
1–5 M HNO3	0.05 M	0.75 M	3.10 ± 0.10a	4.35 ± 0.05

Slope = 2n – 3 with n = 3.05 ± 0.05 (see eq S15).

hemical">TEHDGA dependence for the extraction of hemical">an class="Chemical">Am3+ from 4 M HNO3 into 0.03–0.075 M TEHDGA/0.75 M HEH[EHP]/n-dodecane (filled square), from 2 M HNO3 into 0.02–0.07 M TEHDGA/0.75 M HEH[EHP]/n-dodecane (empty square), and from 2 M HNO3 into 0.02–0.2 M TEHDGA/n-dodecane (filled diamond).

HEH[EHP] Dependence

Attempts to study Am extraction from 4 M hemical">HNO3 by 0.05 M hemical">an class="Chemical">TEHDGA/n-dodecane with variable HEH[EHP] concentrations resulted in the formation of a third phase at lower HEH[EHP] concentrations, and these investigations were not pursued. However, third phases were not observed during contact with 2 M HNO3, yielding distribution data across the HEH[EHP] concentration range of 0–0.75 M (Figure ). The strong propensity of HEH[EHP] to dimerize in n-dodecane means that the dimer (HEH[EHP])2 is actually the active form of the extractant in our experiments.[20] The HEH[EHP] dependence of the Am extraction derived from linear regression analysis was 0.43 ± 0.02 (Table ), similar to a previous report.[31]

Figure 5

HEH[EHP] dependence for the extraction of 241Am3+ from 2 M HNO3 into 0.05 M TEHDGA/0–0.75 M HEH[EHP]/n-dodecane. The 0 M HEH[EHP] data is represented by a dashed line.

hemical">HEH[EHP] dependence for the extraction of hemical">an class="Chemical">241Am3+ from 2 M HNO3 into 0.05 M TEHDGA/0–0.75 M HEH[EHP]/n-dodecane. The 0 M HEH[EHP] data is represented by a dashed line.

The simplest interpretations of this result suggest an approximate 2:1 Am:(hemical">HEH[EHP])2 stoichiometry in the extracted complex. This condition could be met either by the disruption of hemical">an hemical">an class="Chemical">HEH[EHP] dimer, (HEH[EHP])2,[21] accompanied by the incorporation of one HEH[EHP] molecule into a mononuclear extracted complex, or by the formation of a polynuclear organic phase complex with a 2:1 Am:(HEH[EHP])2 stoichiometry. Neither explanation is satisfying, however. First, the ease of HEH[EHP] monomer formation will decrease with increasing HEH[EHP] concentration and would cause noticeable curvature in the distribution data over the wide range of HEH[EHP] concentrations studied. Second, extraction experiments at higher Am concentrations demonstrate that the distribution ratio is independent of [Am] across the range of Am concentrations studied. This observation is not consistent with the presence of a 2:1 Am:(HEH[EHP])2 complex in the organic phase, as DAm is independent of the metal concentration only if each extracted complex contains a single Am cation.

A reasonable alternate explanation for the nonintegral hemical">HEH[EHP] dependence is that hemical">an class="Chemical">HEH[EHP] acts as a phase modifier as well as an extractant. The addition of HEH[EHP] substantially alters the polarity of the TEHDGA/n-dodecane organic phases, as suggested by its ability to suppress third-phase formation in the ALSEP process.[10,11] This, in turn, will affect the extractability of the organic-phase complex with the effect that the equilibrium constant for the extraction will vary with the organic phase HEH[EHP] concentration, a condition that is incompatible with using the distribution ratio to probe complexation equilibria.

In light of this complication, a different approach for probing the hemical">HEH[EHP] stoichiometry of the extracted complexes was devised. The speciation of the extracted complex was studied by UV–vis spectrophotometry in orghemical">anic phases consisting of 0.05 M hemical">an class="Chemical">TEHDGA in n-dodecane and 0–0.075 M HEH[EHP]. The organic phases were contacted with aqueous phases containing 0.01 M Nd in 2 M HNO3 at 35 °C. The higher temperature was used to discourage third phase formation at these metal concentrations. After measuring the amount of Nd extracted, the resulting organic phases were analyzed by spectrophotometry between 550 and 620 nm to determine changes in the Nd coordination environment at different HEH[EHP] concentrations. The series of organic phase spectra obtained from these experiments are displayed in Figure .

Figure 6

(a) UV–vis spectra of neodymium extraction at 35 °C from 2 M HNO3 into 0–0.075 M HEH[EHP]/0.05 M TEHDGA/n-dodecane. (b) Speciation of extracted Nd calculated from eq , with m = 0.9 ± 0.1.

(a) UV–vis spectra of hemical">neodymium extraction at 35 °C from 2 M hemical">an class="Chemical">HNO3 into 0–0.075 M HEH[EHP]/0.05 M TEHDGA/n-dodecane. (b) Speciation of extracted Nd calculated from eq , with m = 0.9 ± 0.1.

Principal component analysis conducted using both the programs SIXpack[38] and GlobalWorks (Online Instrument Systems, Inc.) indicated that the set of spectra for all 10 hemical">HEH[EHP] concentrations studied was composed of two unique spectral components (Figure S2). The spectrum of the first species determined from the model-linked singular value decomposition in GlobalWorks matched the spectrum of Nd(hemical">an class="Chemical">TEHDGA)3·3NO3 extracted from 2 M HNO3. The spectrum of the second species matched that of the Nd-loaded ALSEP organic phase depicted in Figure . Using the spectra of these two Nd complexes, the ratio of the two complexes in the organic phase could be calculated for each solution and used to derive the number of HEH[EHP] dimers, m, coordinated in the ALSEP system for the following generalized equilibrium for trivalent metal, M3+.where species in the organic phase are indicated by overbars, h represents the number of acidic hydrogens in each metal-complexed HEH[EHP] dimer (h = 0, 1, or 2), and n represents the number of nitrate ions in the product complex, with n = 3 – m(2 – h). Defining the equilibrium constant of this reaction to be Kexchange, the equilibrium constant expression can be written in the following form:with b = log Kexchange + m(2 – h) log [HNO3]. Between 0.005 and 0.050 M HEH[EHP], an analysis of the spectroscopic data with eq yields a value of m = 0.9 ± 0.1 (Figure ), implying that under these conditions one HEH[EHP] dimer displaces one TEHDGA molecule from M(TEHDGA)3·3NO3 to form the complex M(TEHDGA)2(H(EH[EHP])2)·nNO3 in the ALSEP organic phase.

Nitric Acid Dependence

The stoichiometry of hemical">nitrate hemical">and hemical">an class="Chemical">hydrogen ions in the extracted complex was probed by examining Am3+ extraction into the standard ALSEP organic phase (0.05 M TEHDGA/0.75 M HEH[EHP]/n-dodecane) as a function of the equilibrium aqueous concentration of nitric acid (Figure ). As previously described, a minimum in the Am extraction is observed at approximately 0.5 M HNO3 due to a change in the organic-phase metal speciation from complexes containing TEHDGA and nitrate at high acidity to homoleptic complexes of monodeprotonated HEH[EHP] dimers, which show an inverse dependence of the metal distribution ratio on the aqueous acidity at low acidities.[11] However, even at aqueous acidities higher than 1 M HNO3, direct slope analysis of the concentration dependence of the distribution ratio is not a reliable indication of the proton or nitrate stoichiometry of the metal complex extracted in the ALSEP process. The large change in ionic strength across the range of aqueous nitric acid concentrations studied causes substantial changes in the activity coefficients of the aqueous solutes, the activity of water, the fraction of aqueous Am present as nitrate complexes, the degree of nitric acid dissociation, the amount of nitric acid extracted into the organic phase, and the fraction of each extractant available to interact with Am. Consequently, a multiequilibrium thermodynamic model for Am extraction from nitric acid into the ALSEP organic phase was developed to interpret the effect of nitric acid on Am extraction and the nitrate and hydrogen ion stoichiometry of the extracted complex using the following general extraction equilibrium:where h = n – 1 and h = 0, 1, or 2. Details of the model and the relationship between the slope of the nitrate dependence and n, the number of nitrate ions incorporated into the extracted complex, are described in the Supporting Information. The evaluation of n, the number of nitrate anions extracted with each Am3+ cation, and Kex, the equilibrium constant for the extraction reaction, between 1 and 5 M HNO3 gave n = 3.05 ± 0.06 and log Kex = 4.35 ± 0.05 (Figure ).

Figure 7

Effect of aqueous nitric acid on Am extraction into 0.05 M TEHDGA/0.75 M HEH[EHP]/n-dodecane. (a) Experimental extraction data (squares) fit to a third-degree polynomial (short dashed line). The filled squares represent data used in the equilibrium activity modeling shown in panel b. (b) Determination of the nitrate stoichiometry, n, in the extracted complex from the correlation of the Am distribution ratio corrected for variations in aqueous activity coefficients, aqueous nitrate complexation, and extraction of nitric acid to the activity of nitrate ions in the aqueous phase using the model described in the Supporting Information.

Effect of aqueous nitric acid on Am extraction into 0.05 M hemical">an class="Chemical">TEHDGA/0.75 M HEH[EHP]/n-dodecane. (a) Experimental extraction data (squares) fit to a third-degree polynomial (short dashed line). The filled squares represent data used in the equilibrium activity modeling shown in panel b. (b) Determination of the nitrate stoichiometry, n, in the extracted complex from the correlation of the Am distribution ratio corrected for variations in aqueous activity coefficients, aqueous nitrate complexation, and extraction of nitric acid to the activity of nitrate ions in the aqueous phase using the model described in the Supporting Information.

Computational Data

Further insight into the inner-sphere coordination of trivalent f-element cations with hemical">an class="Chemical">TEHDGA, HEH[EHP], nitrate anions, and water molecules was gained by comparing the complexation energies of europium–ligand complexes potentially present in the ALSEP organic phase. Many different coordination complexes are possible in the organic phase given the flexibility of the An3+ and Ln3+ coordination spheres and the presence of multiple potential ligands. However, the experimental data rule out many possibilities for the complexes extracted from 2–4 M HNO3, such as complexes containing only singly deprotonated HEH[EHP] dimers (e.g., M{H(EH[EHP])2}3, complexes with inner-sphere water molecules, or complexes containing a single TEHDGA extractant. Consequently, we calculated energies for the following complexes and explored the conformational space of each: TEHDGA only (Eu(TEHDGA)3·3NO3), 2:1 ratio of TEHDGA:deprotonated HEH[EHP] dimer (Eu(TEHDGA)2(H(EH[EHP])2)·2NO3), and 2:1 ratio of TEHDGA:protonated HEH[EHP] dimer (Eu(TEHDGA)2(HEH[EHP])2·3NO3). To enable tractable calculations on these complexes, the ethylhexyl substituents were replaced with ethyl groups, with N,N,N′,N′-tetraethyldiglycolamide (TEDGA) replacing TEHDGA and ethyl phosphonic acid monoethyl ester (HE[EP]) replacing HEH[EHP] for the calculations. The structures of TEDGA and HE[EP] are summarized in Figure S1.

The minimum energy geometries calculated for these three stoichiomehemical">an class="Gene">tries of the TEDGA and HE[EP] complexes are shown in Figure . In the Eu(TEDGA)3·3NO3 complex, the europium ion is coordinated only by tridentate TEDGA ligands, yielding a complex with D3 symmetry, a coordination number of 9, and a calculated minimum energy of −9.7 kcal/mol. The three nitrate anions are positioned in clefts between the alkyl-chains of adjacent TEDGA ligands to balance the charge of europium and produce a neutral complex, as found by Brigham et al. for the closely related extractant TODGA.[19] Replacing one of the TEDGA ligands with a singly deprotonated dialkylphoshonic acid dimer, H(E[EP])2–, yields Eu(TEDGA)2(H(E[EP])2)·2NO3. This complex was observed to have two possible minimum energy geometries within 0.2 kcal/mol (−11.3 and −11.1 kcal/mol). The two geometries differ primarily in the location of the nitrate anions: geometry A contains both nitrate anions in the outer sphere creating an 8-coordinate Eu complex with C2 site symmetry, while geometry B contains one monodentate nitrate anion in the Eu inner coordination sphere, giving europium a coordination number of 9 and C1 symmetry. In geometry A, neither nitrate remains in the cleft between the TEDGA ligands; instead, each nitrate is associated with one of the TEDGA ligands (Figure ). In geometry B, the two nitrates are found in spaces between the TEDGA molecules and the HE[EP] dimer. The movement of the nitrate counteranions can be explained by the increase in inner-sphere free space when one tridentate TEDGA is replaced by a bidentate H(E[EP])2– anion. The extra room allows the remaining two TEDGA ligands to shift, removing the original clefts where the nitrate ions were positioned. Consequently, the nitrate ions reposition, either outer sphere or inner sphere, to stabilize the newly formed Eu complex. The energetic similarity of these complexes is unusual because 9-coordinate Eu complexes are generally expected to be energetically more stable than the 8-coordinate europium complex. However, an analysis of the europium–ligand coordination distances reveals that inner-sphere coordination of the nitrate anion in geometry B causes the Eu–O bond lengths of the TEDGA ligands to increase (Table S2). As a result, the energies of the 8- and 9-coordinate complexes differ by only 0.2 kcal/mol, which suggests that if Eu(TEHDGA)2(H(EH[EHP])2)·2NO3 were to form in the organic phase, both complexes A and B could coexist in solution.

Figure 8

Optimized Eu complexes, Eu(TEDGA)3·3NO3, Eu(TEDGA)2(H(E[EP])2)·2NO3 (A) and (B), and Eu(TEDGA)2(HE[EP])2·3NO3 (C) and (D), investigated investigated as possible species existing during ALSEP extraction. Geometries A and B refer to two different Eu complexes containing deprotonated HE[EP] dimers and geometries C and D refer to two different Eu complexes containing protonated HE[EP] dimers. See the text for further details of the complexes. Europium is teal, phosphorus atoms are orange, oxygen atoms are red, nitrogen atoms are blue, carbon atoms are gray, and hydrogen atoms are white. Dashed lines from Eu highlight the chelating oxygen in every complex. Hydrogen atoms outlined with black circles correspond to the acidic HE[EP] hydrogen. Black dashed lines represent hydrogen bonding while solid black lines represent bonds.

Optimized Eu complexes, Eu(hemical">TEDGA)3·hemical">an class="Chemical">3NO3, Eu(TEDGA)2(H(E[EP])2)·2NO3 (A) and (B), and Eu(TEDGA)2(HE[EP])2·3NO3 (C) and (D), investigated investigated as possible species existing during ALSEP extraction. Geometries A and B refer to two different Eu complexes containing deprotonated HE[EP] dimers and geometries C and D refer to two different Eu complexes containing protonated HE[EP] dimers. See the text for further details of the complexes. Europium is teal, phosphorus atoms are orange, oxygen atoms are red, nitrogen atoms are blue, carbon atoms are gray, and hydrogen atoms are white. Dashed lines from Eu highlight the chelating oxygen in every complex. Hydrogen atoms outlined with black circles correspond to the acidic HE[EP] hydrogen. Black dashed lines represent hydrogen bonding while solid black lines represent bonds.

Five possible minima were located for the Eu(hemical">TEDGA)2(HE[EP])2·hemical">an class="Chemical">3NO3 complex, which forms from a neutral, fully protonated (HE[EP])2 dimer (i.e., H2(E[EP])2) without the loss of a hydrogen ion to the aqueous phase. Two possible complexes were close in energy (−25.7 and −22.9 kcal/mol), while the other three complexes (−12.8, −11.1, and −5.9 kcal/mol) were substantially less stable. Only the two most stable configurations, geometries C and D, are considered here.

In both of these complexes, one P=O--H–O hemical">hydrogen bond in the (HE[EP])2 dimer is disrupted. The critical difference between the two lowest energy conformations of Eu(hemical">an class="Chemical">TEDGA)2(HE[EP])2·3NO3 is the interaction between the (HE[EP])2 dimer and the Eu ion. In geometry C of the Eu(TEDGA)2(HE[EP])2·3NO3 complex (Figure ), the europium sits in a C2 symmetry site, all the nitrate ions remain in the outer coordination sphere, a hydrogen bond forms between one HE[EP] and a nitrate anion, and bidentate coordination between the HE[EP] dimer and Eu is observed. The energy of this complex is calculated to be −22.9 kcal/mol. The PO–H bond of the HE[EP] hydrogen bonding to nitrate elongates from 1.03 to 1.57 Å, while the H--ON hydrogen bond between this HE[EP] and the nitrate anion is 1.02 Å. This indicates that such a complex would take on the character of a deprotonated HE[EP] dimer hydrogen bonded to nitric acid. Alternately, the Eu(TEDGA)2(HE[EP])2·3NO3 complex can distort, leading to the C1 symmetry complex shown in geometry D. (Figure ) This is the most stable Eu(TEDGA)2(HE[EP])2·3NO3 complex studied at −25.7 kcal/mol. In geometry D, the (HE[EP])2 dimer becomes monodentate, as one Eu–OP bond and one P=O--H–O–P hydrogen bond are cleaved, and nitrate anions reposition to stabilize the entire complex. One nitrate anion moves into the Eu inner coordination sphere with a Eu–ON bond length of 2.34 Å while a second nitrate anion hydrogen bonds to the hydrogen on the noncoordinated HE[EP] in the outer coordination sphere. The noncoordinated HE[EP] PO–H bond length remains 1.04 Å while the H--ON hydrogen bond between this HE[EP] and the nitrate anion is 1.48 Å. This implies that complex D would have the character of a fully protonated HE[EP] dimer hydrogen bonded to a nitrate anion. This asymmetric complex featuring inner- and outer-sphere nitrate ions is 2.8 kcal/mol more stable than the Eu(TEDGA)2(HE[EP])2·3NO3 complex with all three nitrates in the outer sphere and 19.0 kcal/mol more stable than Eu(TEDGA)3·3NO3.

Discussion

Together, the equilibrium partitioning experiments, optical spectroscopy, and modeling clearly demonstrate the formation of a new extracted complex when hemical">HEH[EHP] is added to solutions of hemical">an class="Chemical">TEHDGA under metal extraction conditions in acidic biphasic systems. Previous evidence of mixed TEHDGA–HEH[EHP] complexes under ALSEP extraction conditions has been inconsistent. The original work on the ALSEP process postulated that ternary metal–TEHDGA–HDEHP (HDEHP = bis(2-ethylhexyl)phosphoric acid) complexes were likely involved in An3+ and Ln3+ extraction under acidic conditions but also concluded that it was unclear if metal–TEHDGA–HEH[EHP] complexes form.[11] Subsequent EXAFS studies of Nd3+ extracted into mixtures of TEHDGA and HEH[EHP] from 1 M HNO3 suggested the possibility of a Nd–O–P scattering path consistent with “relatively weak interactions between Nd–HEH[EHP]–T2EHGDA when extracted from a mildly acidic aqueous environment”.[12] In contrast to both these studies, a later EXAFS study of Eu–TODGA–HDEHP complexes concluded that only 1:3:3 Eu:TODGA:nitrate complexes are extracted into 0.05 M TODGA/0.75 M HDEHP/n-dodecane from 3 M HNO3.[32] On the other hand, from FTIR experiments, Rama Swami et al. proposed simultaneous coordination of TEHDGA and HEH[EHP] or HDEHP when Eu3+ is extracted into mixtures of these extractants from 3 M HNO3.[18] The observation of multiextractant organic-phase complexes in the ALSEP system is also similar to reports for An3+ and Ln3+ extraction by mixtures of N,N′-dimethyl-N,N′-dioctylhexylethoxymalonamide and dialkylphosphoric acid extractants.[39,43,44] Our experiments and modeling confirm the formation and identity of the mixed metal–TEHDGA–HEH[EHP] complex formed in the ALSEP organic phase under acidic extracting conditions.

The presence of neutral and acidic coextractants at variable concentrations and high aqueous acidities makes the determination of the organic-phase speciation in the hemical">ALSEP process difficult. At low acidity, the hemical">an class="Chemical">ALSEP process’ acidic extractant, HEH[EHP], forms the M{H(EH[EHP])2}3 complex,[12] liberating hydrogen ions to the aqueous phase according to the following equilibrium.Eq is clearly opposed by high acidities, and the extraction of An3+ or lanthanides lighter than terbium by HEH[EHP] is negligible at aqueous acidities of 1 M or greater.[45] By itself, the neutral extractant TEHDGA forms complexes in the organic phase according the following equilibrium.[19,20]If the nitrate anions are supplied by nitric acid, eq will be favored by high nitric acid concentrations and negligible at the low acidities favored by eq . At high nitric acid concentrations, experiments also suggest coextraction of nitric acid, depending on the conditions,[20] with the following equilibrium.Each of these equilibria has a characteristic dependence on the extractant, nitrate, and hydrogen ion activities (Table ). However, in the mixed HEH[EHP]/TEHDGA ALSEP system, a distinctly different stoichiometry is observed, 1 An3+/Ln3+:2.1 TEHDGA:0.9 (HEH[EHP])2:3.05 NO3–, with no aqueous hydrogen ions as reactant or product and no inner-sphere-coordinated water. Furthermore, the optical spectroscopy of the extracted Nd and Eu complexes demonstrate the stoichiometry observed in the ALSEP system solvent does not result from mixtures of the products of eq and eq or eq because the distinct optical signatures of M{H(EH[EHP])2}3, the absorption peaks at 570 and 606 nm in the Nd spectrum (Figure ) and the Eu fluorescence emission band at 611 nm (Figure ), are absent. This disappearance of the homoleptic Nd–HEH[EHP] complex under ALSEP extracting conditions is also consistent with experiments by Gullekson et al., who titrated TEHDGA·HNO3/n-dodecane into isolated organic phases initially containing Nd{H(EH[EHP])2}3.[12]

Table 3

Expected Dependence of Distribution Ratios on the Organic-Phase Concentrations of Extractants and the Activity of Hydrogen and Nitrate Ions in Aqueous Nitric Acid for Proposed Extraction Equilibria

		slope	slope	stoichiometry		slopea
equilibrium	extracted complex	[TEHDGA]	[(HEH[EHP])2]	{H+}	{NO3–}	{NO3–}
3	M(TEHDGA)2(Hh(EH[EHP])2)·nNO3	2	1	n – 3	n	2n – 3
4	M{H(EH[EHP])2}3	0	3	–3		–3
5	M(TEHDGA)3·3NO3	3	0	0	3	3
6	M(TEHDGA)3·3NO3·pHNO3	3	0	p	p + 3	2p + 3

Slope expected in nitric acid solutions when [H+] ≈ [NO3–] (eqs S7c and S15).

Slope expected in an class="Chemical">nitric acid solutions when [H+] ≈ [NO3–] (eqs S7c hemical">and S15).

The change in the hemical">metal ion site symmetry of the extracted complex between orghemical">anic phases containing only hemical">an class="Chemical">TEHDGA and those containing mixtures of TEHDGA and HEH[EHP] is also apparent in the changes in the optical spectra (Figures and 6). The 5D0 → 7F0 transition of Eu3+, which occurs near 580 nm in the condensed phases, is particularly diagnostic. This transition is only allowed for a subset of noncubic space groups without an inversion center, C, C, and C.[46] In other symmetries, for example D or any space group with an inversion center, this transition is forbidden and is generally very weak if observed at all. The extraction of Eu3+ into 0.75 M HEH[EHP]/n-dodecane or 0.2 M TEHDGA/n-dodecane yielded solutions with barely discernible 5D0 → 7F0 emissions (Figure and Figure S6), in agreement with previous reports.[21,47] In contrast, the extraction of Eu3+ from 4 M HNO3 into 0−0.75 M HEH[EHP]/0.05 M TEHDGA/n-dodecane gives a solution with an obvious 5D0 → 7F0 emission band at 579.7 nm. The absence of a significant 5D0 → 7F0 emission for the organic-phase Eu complexes containing only HEH[EHP] or TEHDGA is readily understood based on the Eu site symmetry. Eu{H(EH[EHP])2}3 is believed to possess approximately octahedral symmetry with an inversion center.[21,29] The Eu(TEHDGA)3·3NO3 complex extracted into 0.05 M TEHDGA will have D3 site symmetry,[19] similar to the europium trisoxydiacetato and trisdipicolinato complexes, which also show exceedingly weak 5D0 → 7F0 emission.[48,49] However, replacing one TEHDGA molecule with one mono- or bidentate (HEH[EHP])2 will necessarily yield a lower symmetry complex (C2, C2, or C1) with an allowed 5D0 → 7F0 transition, as we observe. The lowest energy conformations we found in our theoretical calculations (geometries C and D) show these lowered symmetries.

The changes in the 5D0 → 7F4 emission band at approximately 700 nm are also consistent with a decrease in symmetry of the Eu site in the hemical">ALSEP system compared to the orghemical">anic phase containing only hemical">an class="Chemical">TEHDGA. The Eu3+7F4 state is split into 4 levels in crystal fields of D3 symmetry (i.e., for Eu(TEHDGA)3·3NO3), while it splits into 7 levels for C2 symmetry and 9 levels for C2 or C1 symmetry.[50] The broadening of the 5D0 → 7F4 transition in the Eu-ALSEP emission spectrum compared to the Eu-TEHDGA spectrum is accompanied by a transition from 4 discernible bands in the Eu-TEHDGA spectrum to the presence of 6 discernible shoulders on the main peak for the Eu-ALSEP spectrum (Figure S6), suggesting that the 7F4 state splits into at least 7 discrete levels. This change in site symmetry of the extracted complex is consistent with direct coordination of HEH[EHP] in the ALSEP system.

With the hemical">metal:lighemical">and stoichiometry established by equilibrium partitioning hemical">and spectroscopic experiments, the protonation state of hemical">an class="Chemical">HEH[EHP] in the extracted complex remains key to understanding the nature of the extracted complex. The extracted complex must be charge neutral. The charge on the extracted An3+ and Ln3+ cations can be balanced by coextracted nitrate anions (eq ), deprotonated HEH[EHP] molecules (e.g., H(EH[EHP])2– in eq ), or a combination of the two. In addition, HNO3 may be incorporated into the outer sphere of the extracted complex (eq ). The aqueous acidity disfavors the formation and complexation of the H(EH[EHP])2– anion, as the pKa of HEH[EHP] is ca. 4.1,[51] and our nitrate dependence experiments (Figure ) clearly show the coextraction of 3 nitrates per trivalent cation (n = 3) without the loss of hydrogen ions from HEH[EHP] to the aqueous phase according to eq . If H(EH[EHP])2– or 2 equiv of EH[EHP]− was to form by the release of H+ from the HEH[EHP] dimers to the aqueous phase, the experimental nitrate activity dependence in Figure would have a slope of 1 (n = 2) or −1 (n = 1), respectively (see eq , eq S15, and the Supporting Information). Instead, the experimentally verified lack of H+ release to the aqueous phase suggests either that HEH[EHP] coordinates to the metal center as a neutral, protonated species[45] or that HEH[EHP] deprotonates during the formation of the extracted complex with the released hydrogen ion remaining in the organic phase and forming a molecule of nitric acid by reacting with one nitrate anion.

Computational modeling of the possible extracted complexes provides further insight into hemical">HEH[EHP] coordination hemical">and the likely protonation state of these complexes. The complexation energy of Eu(hemical">an class="Chemical">TEDGA)3·3NO3 was calculated to be −9.7 kcal/mol compared to −11.3/–11.1 and −22.9/–25.7 kcal/mol for Eu(TEDGA)2(H(E[EP])2)·2NO3 geometries A/B and Eu(TEDGA)2(HE[EP])2·3NO3 geometries C/D, respectively (Figure and Table ). Although the formation of M(TEDGA)3·3NO3 is exothermic, the substantially greater stability of the ternary metal–diglycolamide–phosphonic acid complexes suggest that fully protonated HEH[EHP] can readily displace a TEHDGA molecule, since HEH[EHP] is available at 15 times the concentration of TEHDGA in the ALSEP organic phase. Indeed, the biphasic spectrophotometric titrations (Figure ) indicate that an appreciable concentration of a 1:2:1 Nd:TEHDGA:(HEH[EHP])2 complex is already present with 0.015 M concentrations of HEH[EHP] and 0.05 M TEHDGA. Consequently, M(TEHDGA)3·3NO3 is unlikely to be present during ALSEP extractions.

Table 4

Calculated Gibbs Free Energies of Complexation for the Optimal Geometries of Eu(TEDGA)·3NO3, Eu(TEDGA)2(H(E[EP])2)·2NO3, and Eu(TEDGA)2(HE[EP])2·3NO3

complexa	complexation energy (kcal/mol)
Eu(TEDGA)·3NO3	–9.7
Eu(TEDGA)2(H(E[EP])2)·2NO3 (A)	–11.3
Eu(TEDGA)2(H(E[EP])2)·2NO3 (B)	–11.1
Eu(TEDGA)2(HE[EP])2·3NO3 (C)	–22.9
Eu(TEDGA)2(HE[EP])2·3NO3 (D)	–25.7

Extractants truncated from 2-ethylhexyl to ethyl chains for computational feasibility.

Extractants truncated from an class="Chemical">2-ethylhexyl to hemical">an class="Chemical">ethyl chains for computational feasibility.

Substitutions of hemical">HEH[EHP] for hemical">an class="Chemical">TEHDGA in the extracted complex of the ALSEP system also appear to affect the nitrate coordination environment profoundly. In the M(TEHDGA)3·3NO3 complex, the nitrate anions required to balance the positive charge of the metal cation sit in the outer coordination sphere in clefts between the coordinated TEHDGA ligands,[19,52] likely interacting with coextracted water molecules present in the outer coordination sphere.[53] Two of these 3-fold interligand nitrate binding clefts are lost when one TEHDGA molecule is replaced with HEH[EHP] to form the ALSEP system’s extracted complex. Consider the hypothetical ALSEP extraction reaction:It would display a nitrate activity dependence slope of 1 with n = 2 (eq S15 and the Supporting Information) and produce complexes with geometries A or B. For these complexes, the computations indicate that replacing one TEHDGA with H(EH[EHP])2– will cause both nitrate anions to migrate out of the one remaining cleft between TEHDGA molecules. In geometry A, the nitrates occupy spaces between the TEHDGA and HEH[EHP] molecules. In the case of geometry B, one of the nitrate anions repositions further to form an inner-sphere complex with the Eu ion (Figure ).

The changes in the hemical">nitrate environment are even more profohemical">an class="Gene">und for the most stable complexes studied: geometries C and D of Eu(TEDGA)2(HE[EP])2·3NO3. Unlike the complexes represented by geometries A and B, the stoichiometry of these complexes C and D matches the experimentally determined stoichiometry and the overall extraction equilibrium.In both geometries C and D, the computations suggest that one nitrate anion will remain in the cleft between the two TEHDGA molecules and a second nitrate will interact with the acidic hydrogen of one of the neutral HEH[EHP] molecules. In geometry C the third nitrate anion resides in the outer coordination sphere between one TEHDGA and one HEH[EHP] molecule. However, the bond lengths suggest that the third nitrate reacts with the acidic hydrogen from the other HEH[EHP] molecule in the complex to form a molecule of HNO3 that hydrogen bonds to the coordinated oxygen of the HEH[EHP] (Figure ). In complex C, one inter-HEH[EHP] hydrogen bond remains intact, and the (HEH[EHP])2 dimer is coordinated to the metal in a bidentate fashion. The equilibrium for this process can be represented asIn geometry D, similar to geometry C, one nitrate remains in the inter-TEHDGA cleft, a second nitrate hydrogen bonds to one fully protonated HEH[EHP] molecule, and one of the hydrogen bonds between HEH[EHP] monomers remains intact. However, in complex D, the HEH[EHP] dimer becomes monodentate. The HEH[EHP] monomer that hydrogen bonds to nitrate is rotated away from the Eu and does not coordinate to the metal center. Instead, the third nitrate anion moves into the inner coordination sphere to give an 8-coordinate complex. Highlighting this change in nitrate coordination, the equilibrium for the formation of complex D can be written asFunctionally, the low-energy complexes with geometries C and D form by sharing an acidic hydrogen between a HEH[EHP] molecule and a nitrate ion in the organic phase. Stoichiometrically, complexes C and D are indistinguishable, they match our experimentally determined speciation of the organic-phase complex, and they present similarly favorable complexation energies compared to the other complexes studied.

Complexation-driven hemical">HEH[EHP] proton trhemical">ansfer to create hemical">an class="Chemical">hydrogen-bonded HNO3 in the organic phase is an intriguing reaction mechanism for trivalent f-element extraction in the ALSEP process. It explains the observed synergistic enhancement to the extraction caused by HEH[EHP] addition, the observed stoichiometry under common extracting conditions, and the eventual loss of HEH[EHP] dependence at very high aqueous acidities,[31] and it is further supported by FTIR studies of Eu extracted from 3 M HNO3 into TEHDGA/HEH[EHP]/n-dodecane.[18] Moreover, the predicted stabilization of HNO3 in the organic phase is not surprising. The ability of HEH[EHP] to extract nitric acid on its own is known (Figure S3), and the coextraction of nitric acid with An(NO3)3 or Ln(NO3)3 has been reported for bifunctional extractants such as octyl(phenyl)-N,N-diisopropylcarbamoylmethanephosphine oxide (CMPO),[54,55] malonamides,[56] and diglycolamides TODGA and TEHDGA.[20,57] In the ALSEP process, the nondissociated acidic hydrogens of HEH[EHP] appear to exert a stabilizing influence on the extracted complexes, enabling the formation of f-element-TEHDGA-HEH[EHP] complexes under conditions where HEH[EHP]’s conventional metal–proton exchange equilibrium is strongly hindered.

Conclusions

Our studies conclusively demonstrate the formation of a previously suspected mixed trivalent f-element–hemical">an class="Chemical">TEHDGA/HEH[EHP] complex in the organic phase under the acidic extracting conditions of the ALSEP process (2–4 M HNO3). The mutual presence of a strong neutral extractant and a weakly acidic extractant in the ALSEP organic phase provides unique opportunities for the synergistic extraction of An3+ and Ln3+ from aqueous molar nitric acid solutions. As a more basic extractant, the dialkylphosphonic acid HEH[EHP] in the ALSEP system displays different behavior than the dialkylphosphoric acid HDEHP in similar mixed-extractant systems.[11,43,58] HEH[EHP] does not appear to undergo deprotonation when it cooperates with TEHDGA to extract trivalent lanthanide or actinide nitrates under the ALSEP process extracting conditions. While the strongly acidic aqueous phases inhibit the deprotonation of HEH[EHP], the relatively high concentration of HEH[EHP], strong PO–M bonds, a smaller coordination footprint (mono or bidentate (HEH[EHP])2 vs tridentate TEHDGA), and an ability to hydrogen bond with nitrate anions work together to boost the stability of the mixed M(TEHDGA)2(HEHEHP)2·3NO3 complex in the ALSEP organic phase.

Methods

Materials

Chemicals were purchased from Sigma-Aldrich except where otherwise indicated. The extractant hemical">HEH[EHP] was obtained from BOC Sciences (95%) hemical">and purified to ≥98% by the third-phase method,[59] as confirmed by hemical">an class="Chemical">31P NMR and acid–base titration in an ethanol–water mixture (80:20). Stock extractant solutions were made by combining weighed amounts of TEHDGA (Eichrom Technologies, > 99%) and the purified HEH[EHP] in n-dodecane (anhydrous, 99%) and then diluting to a known volume.

Aqueous phases were made from hemical">nitric acid (Baker, ULTREX II) hemical">and sthemical">andardized by titration with hemical">an class="Chemical">sodium hydroxide. Solutions of lanthanide nitrates were prepared at the desired acidities from stock solutions of neodymium(III) nitrate hexahydrate (99.9% trace metals basis) or a europium nitrate solution prepared by dissolving a weighed amount of europium oxide (Treibacher Industrie AG, 99.99% REE) in nitric acid. The neodymium stock solution was standardized by titration with a standard EDTA solution (Fisher Scientific), xylenol orange indicator, and a saturated solution of hexamethylenetetramine (>99%) to buffer the titration at pH 5–5.5. The pH was adjusted with 4 M HNO3 and monitored throughout the titration with a ThermoOrion Ross semimicro pH electrode. Aqueous solutions were prepared using Millipore 18 MΩ cm deionized water, while deuterium oxide (99.9 atom % D) and 65 wt % nitric acid-d in D2O (99 atom %) were used for the D2O luminescence experiments.

Absorption Spectroscopy

Aqueous phases of hemical">neodymium were made to 0.01 M hemical">and diluted with the relevhemical">ant titrated nihemical">an class="Gene">tric acid solutions. Acid concentrations were chosen to optimize extraction of Nd while avoiding the formation of a third phase; therefore, the acid concentrations at ALSEP process relevant conditions were not always possible to use. Organic-phase solutions were prepared by weight and pre-equilibrated through contact with twice their volume of the appropriate metal-free aqueous phase and then vortexing for 30 s, followed by centrifugation for phase separation. This process was repeated twice to complete pre-equilibration. Metal extractions were performed by vortexing equal volumes of pre-equilibrated organic phases with aqueous phases containing 0.001–0.01 M Nd(NO3)3 at the desired nitric acid concentration for 5 min followed by 5 min of centrifugation before the phases were separated.

Spectra of both the aqueous and organic phases were obtained on a Varian Cary 300 Spectrophotometer in 1.000 cm quartz cuvettes from 480 to 850 at 0.2 nm resolution. The equilibrium concentrations of an class="Chemical">neodymium in the orghemical">anic phases were calculated from the total amount of Nd in the system hemical">and the difference in the initial hemical">and final spectra of the aqueous phase.

The influence of hemical">HEH[EHP] on the stoichiometry of the orghemical">anic-phase Nd complexes also was studied spectroscopically. Orghemical">anic phases containing 0.05 M hemical">an class="Chemical">TEHDGA and varying amounts of HEH[EHP] were made by combining aliquots of 0.2 M HEH[EHP]/n-dodecane, 0.2 M TEHDGA/n-dodecane, and n-dodecane, all of which had been pre-equilibrated with 2 M HNO3. These solutions were pipetted in appropriate ratios to vary the HEH[EHP] concentration only. In addition, 0.05 M HEH[EHP]/n-dodecane and 0.05 M TEHDGA/n-dodecane organic phases were also tested. Culture tubes containing equal volumes organic phase and 0.01 M neodymium in 2 M HNO3 were initially placed into a 35 °C water bath for 30 min, and the samples were then vortexed for 30 s and placed back in the water bath for several minutes. These intervals were repeated until a total vortexing time of 5 min was reached. Phases were separated, and spectra were collected between 490 and 610 nm in a 1.000 cm jacketed cuvette held at 35 °C by a recirculating water bath.

Time-Resolved Laser-Induced Fluorescence Spectroscopy (TRLFS)

TRLFS measurements were performed on hemical">europium-loaded orghemical">anic solutions prepared by extraction from light hemical">and heavy hemical">an class="Chemical">water. Deuterated samples were made entirely with deuterated chemicals except for the europium nitrate stock solution, which had been prepared in light water and contributed no more than 2 atom % 1H to the aqueous solutions. Organic solutions were pre-equilibrated with appropriate D2O/DNO3 or H2O/HNO3 aqueous phases and then contacted with a fresh aqueous phase containing europium similar to the procedure for the absorption spectroscopy experiments. Fluorescence spectra were obtained using an Edinburgh Instruments LP980 spectrometer with a Continuum Surelite nanosecond Nd:YAG laser as the excitation source. The excitation wavelength was fixed at 394 nm, while the emission spectrum was recorded in the range of 300 to 800 nm. Fluorescence lifetimes were obtained by monitoring the lifetimes of the europium 5D0 → 7F transitions (J = 0–4) at the wavelengths of interest. Kinetic traces at each wavelength were fit as a single exponential decay, and the goodness of fit was evaluated using the reduced chi-squared (χ2). The coordination environment of europium was elucidated based on the fluorescence emission spectra and fluorescence lifetime of the complexes, and the change in lifetime between nondeuterated and deuterated samples. In addition, the number of inner-sphere-coordinated water molecules (NH2O) was calculated using Horrock’s equation:where kH and kD are the fluorescence decay rate constants in ms–1 for samples prepared in H2O and D2O, respectively.[60−63] The equation calculates the number of water molecules to within 0.5 molecules.

Extraction Experiments

The procedures for studying the extraction of hemical">nitric acid are described in the Supporting Information. All radiotracer experiments were performed using radiochemically pure 241Am (Eckert hemical">and Ziegler) dissolved in 2 M hemical">an class="Chemical">HNO3. Extractions were performed by pipetting equal volumes of pre-equilibrated organic and aqueous phases containing the desired concentrations of acid and extractant into glass culture tubes, spiking with 2 μL of the 241Am solution, and vortexing for 5 min, followed by 5 min of centrifugation. The phases were separated, and aliquots of each phase were taken for liquid scintillation counting on a Packard 2500 TR liquid scintillation analyzer using an Ecoscint liquid scintillation cocktail. Distribution ratios (D) were calculated from the following equation:assuming equal counting efficiency of the aqueous and organic phases. All experimental uncertainties are reported at two standard deviations, and error bars are not shown in figures if the uncertainty is smaller than the data points.

Theoretical Calculations

All calculations were performed using Gaussian 09 software.[64] The alkyl chains of the ligands under investigation were truncated from hemical">an class="Chemical">2-ethylhexyl chains to ethyl chains for ease of calculation, as previously reported.[65] Therefore, the computational analysis replaces TEHDGA with TEDGA and HEH[EHP] with HE[EP] (see Figure S1 for structures of the truncated ligands). The europium was treated as a trivalent atom containing a +3 charge and a multiplicity of seven. The geometries of the hypothesized europium–ligand complexes were optimized using Becke’s three parameter exchange functional with Lee–Yang–Parr’s correlation functional (B3LYP) and the 6-31G(d,p) basis sets for all second and third row elements.[66,67] A relativistic effective core pseudopotential and corresponding basis set were employed (Stuttgart RSC 1997 ECP) for treatment of the europium cation.[68] These calculations were performed without explicit or implicit solvation since the europium extraction complexes under investigation will be present in the dodecane organic phase, which has a dielectric constant near that of the gas phase. Calculated stationary points were identified as a minimum by verifying the lack of any imaginary frequencies. Complexation energies were calculated from the difference of the Gibbs Free Energies of the products and reactants for each complex. Even with truncation of the alkyl chains, the number of degrees of freedom of each complex is large. Therefore, the exploration of the conformational spaces was necessarily limited to dozens of conformations for each complex.