Article pubs.acs.org/IC
Cite This: Inorg. Chem. XXXX, XXX, XXX−XXX
Coordination Chemistry of a Strongly-Donating Hydroxylamine with Early Actinides: An Investigation of Redox Properties and Electronic Structure Alex McSkimming,†,‡ Jing Su,†,§ Thibault Cheisson,‡ Michael R. Gau,‡ Patrick J. Carroll,‡ Enrique R. Batista,*,§ Ping Yang,*,§ and Eric J. Schelter*,‡ ‡
P. Roy and Diana T. Vagelos Laboratories, Department of Chemistry, University of Pennsylvania, 231 S 34th Street, Philadelphia, Pennsylvania 19104, United States § Theoretical Division, Los Alamos National Laboratory, Los Alamos, New Mexico 87545, United States S Supporting Information *
ABSTRACT: Separations of f-block elements are a critical aspect of nuclear waste processing. Redox-based separations offer promise, but challenges remain in stabilizing and differentiating actinides in high oxidation states. The investigation of new ligand types that provide thermodynamic stabilization to high-valent actinides is essential for expanding their fundamental chemistry and to elaborate new separation techniques and storage methods. We report herein the preparation and characterization of Th and U complexes of the pyridyl-hydroxylamine ligand, N-tert-butyl-N-(pyridin-2-yl)hydroxylamine (pyNO−). Electrochemical studies performed on the homoleptic complexes [M(pyNO)4] (M = Th, U) revealed significant stabilization of the U complex upon one-electron oxidation. The salt [U(pyNO)4]+ was isolated by chemical oxidation of [U(pyNO)4]; spectroscopic and computational data support assignment as a UV cation.
■
congeners16 showed that the pyNO− ligands provided unprecedented stabilization of the CeIV cation, as determined by the potential for electrochemical reduction to CeIII. In an effort to extend the coordination chemistry of the pyNO− ligand framework to 5f block elements, we report herein the synthesis and electrochemisty of [M(pyNO)4] (M = Th, U) (Figure 1) and the preparation and structure of the uranyl complex, [UO2(pyNO)3K]n, obtained through an oxidative group transfer reaction with the pyNO− ligand. The pyNO−
INTRODUCTION A significant challenge to the use of nuclear power is the generation of high-level liquid waste (HLLW) which presents severe radiological health risks and storage challenges.1−4 Understanding and control of metal oxidation states is crucial to develop and optimize new separations chemistries and to assess and confer long-term stability to waste forms of HLLW.5−11 For example, the Plutonium Uranium Redox EXtraction (PUREX) process, which is used to reprocess nuclear fuel, makes use of differences in redox chemistry for the separation of Pu and U from lanthanides and heavier actinides and for the separation of Pu and U from one another.12 The PUREX process can also incorporate separation of Np from U by selectively reducing the neptunyl ion (NpVIO22+) to NpVO2+.12 Similarly, in the SESAME process, Am in the post-PUREX waste stream is oxidized to AmIV, which facilitates its separation from CmIII.13,14 We, and others, have engaged in developing ligands that promote f-element redox chemistry and stabilize higher oxidation states.2,6,10,11,15−19 In all of these contexts, it is important to obtain information on the relative redox thermodynamics for f-block elements and to understand how ligands stabilize different oxidation states. Computational chemistry has begun to show predictive power for redox couples of d- and f-block metal complexes, but calibration with experimental data is essential.20−27 Previous reports on the homoleptic complex [Ce(pyNO)4] (pyNO− = N-tert-butyl-N-(pyridin-2-yl)hydroxylamine)15 and © XXXX American Chemical Society
Figure 1. Simplified depiction of [M(pyNO)4]. Heteroatoms are highlighted for clarity. Received: December 31, 2017
A
DOI: 10.1021/acs.inorgchem.7b03238 Inorg. Chem. XXXX, XXX, XXX−XXX
Article
Inorganic Chemistry
Figure 2. Thermal ellipsoid plots (50% probability) of the solid-state structures of (a) [Th(pyNO)4], (b) [U(pyNO)4], and (c) [UO2(pyNO)3K]n. Unlabeled ellipsoids correspond to carbon atoms. Hydrogen atoms are omitted, and tBu groups are shown in wire-frame for clarity.
[UO2(pyNO)3K]n (vide inf ra) suggested O atom abstraction from the pyNO− ligands. As a result of the instability of [U(pyNO)4], all solution phase measurements on this complex were collected as rapidly as possible. [M(pyNO)4] (M = Ce, Th, U) complexes all crystallized in the C2/c space-group with near identical unit cells. The metal coordination environments consisted of four roughly planar pyNO ligands, which bound through the pyridyl and hydroxylamine oxygen donors. The complexes all exhibited distorted trigonal dodecahedron geometries with approximate D2d symmetry. The metal ions each resided on crystallographic C2 axis, which rendered the two sets of trans pyNO− ligands equivalent. Within the pyNO− ligands there were only slight changes in bond metrics between complexes, consistent with the pyNO− ligands being fully reduced and bound to tetravalent cations. The previously reported Ce complex featured the shortest M−O bonds at 2.2351(12) and 2.2361(12) Å, followed by U, 2.250(2) Å, and Th, 2.3013(12) and 2.981(12) Å, reflecting the increasing ionic radii of the metal ions (Ce, U, Th = 0.97, 1.00, 1.05 Å for 8 coordinate complexes).36 Notably, attempts to prepare [U(pyNO)4] from UCl4 and excess K(pyNO) instead gave the uranyl complex [UO2(pyNO)3K]n (Figure 2) as an orange crystalline solid in 62% yield (Scheme 1). Oxidation of UIV, MoV, VIII by O atom abstraction from hydroxamic acids was previously documented;37,38 therefore, [UO2(pyNO)3K]n likely arose from oxygen atom abstraction from pyNO− by the oxophilic uranium cation. Formation of uranyl (UVIO22+) specie(s) likely accounts for the observed instability of
ligands significantly stabilized the pentavalent uranium congener [UV(pyNO)4]+, which was also prepared, relative to other reported UV complexes.28−35 These results expand our knowledge of the redox chemistry of U and Th with electron donating ligands and establish a basis for the application of the pyNO− framework to the preparation and study of high-valent transuranic elements.
■
RESULTS AND DISCUSSION
Synthesis and Characterization of Complexes. To establish the study of 5f elements with the pyNO− ligands, we first sought to prepare [Th(pyNO)4]. We expected the ThIV cation would be redox-silent such that [Th(pyNO)4] would provide baseline expectations for ligand-based redox activity with this framework. The complex [Th(pyNO)4] was prepared as a bright yellow crystalline solid in 82% yield by saltmetathesis of [ThCl4(DME)2] with 4 equiv of K(pyNO). The 1 H NMR spectrum of [Th(pyNO)4] featured four aromatic signals and a singlet at δ 1.40 for the tBu protons of four chemically equivalent pyNO− ligands (see ESI). The 13C{1H} NMR spectrum of the complex revealed the expected 5 aromatic signals and 2 aliphatic signals for the pyNO− ligands. Having established conditions effective for the isolation of [Th(pyNO)4], we next pursued the isostructural uranium congener. The uranium complex [U(pyNO)4] was isolated as a red crystalline solid in 46% yield by a protonolysis reaction between [U(N(SiMe3))3] and 5 equiv of pyNOH at −35 °C. The mechanism by which pyNOH oxidizes U was not studied; however, we note that increasing the amount of pyNOH in the reaction did not increase the yield of [U(pyNO)4]. The 1H NMR spectrum of [U(pyNO)4] revealed five resonances for the equivalent pyNO− ligands, with the chemical shifts and line broadening being consistent with a UIV center. In the absence of air and moisture [Th(pyNO)4] was stable indefinitely at room temperature in the solid state and in THF/ CH2Cl2 solution. Similarly, [U(pyNO)4] proved stable in the solid state with samples stored at −35 °C for months remaining unchanged. Solutions of [U(pyNO)4], however, visibly darkened and became brown ( −0.8 V are typical;28−35 e.g., E1/2(UIV/V) = −0.48 V for [(C5Me5)2U(−NCPh2)].31 Thus, analogous to [Ce(pyNO)4], the pyNO− ligands evidently provide significant stabilization of the higher UV oxidation state. To confirm uranium, rather than pyNO, -centered oxidation, [U(pyNO)4]+ was prepared in 79% yield by addition of 1 equiv of [Fc][BArF4] to [U(pyNO)4]. XRD analysis revealed the complex cation to be structurally analogous to [U(pyNO)4] (see Figure S8). The N−O distances within the pyNO ligands were similar and averaged ∼1.39 Å (c.f. 1.38 Å for [U(pyNO)4]); that is, no diagnostic contraction of the N−O bonds was observed which would indicate localized, ligandcentered oxidation.46,47 The near-infrared (NIR) spectrum of [U(pyNO)4][BArF4] shows a sharp but weak (εmax = 59 M−1 cm−1) band at 1601 nm (Figure S7). Peaks of similar energy and intensity were observed for [(C5Me5)2UV(N−Ar)(X)] complexes and attributed to a characteristic f−f transition.44 This result suggested that for [U(pyNO)4][BArF4], U was indeed in the +5 oxidation state; this hypothesis was probed further by DFT studies (vide inf ra). Computation and Electronic Structure. Density functional theory (DFT) calculations were carried out to gain insight into the electronic structures of [M(pyNO)4] (M = Ce, Th, U) and their oxidation and reduction products. The computed average bond distances of [M(pyNO)4] and [U(pyNO)4]+ as well as the experimental crystal structure data are summarized in Table 1. The computed structural
[U(pyNO)4] in solution. X-ray diffraction (XRD) studies performed on [UO2(pyNO)3K]n revealed a polymeric structure with each [UO2(pyNO)3]− unit bridged by a K cation; a truncated depiction of the polymeric structure is shown in Figure 2. The coordination modes of the pyNO − ligands in [UO2(pyNO)3K]n differed significantly compared to those observed for [U(pyNO)4]. For [UO2(pyNO)3K]n, one pyNO− ligand bound the uranium cation through the pyridyl donor and hydroxylamine O(1) atom, as in [U(pyNO)4]. However, the other two pyNO− groups coordinated in an η2 fashion through the hydroxylamine N and O atoms (N(4), N(6), O(2), and O(3)), with the pyridyl donors bound to K+ cations. The U O bond lengths of 1.803(3) and 1.806(3) Å are typical for uranyl complexes.39−41 As expected, the UO2 fragment is near linear (∠OUO = 177.6(1)°); the slight deviation is possibly a result of a short O−K+ contact (2.803 Å).39−41 [UO2(pyNO)3K]n was insoluble in common organic solvents such as THF and CH2Cl2 as well as strong donor solvents such as pyridine, consistent with the polymeric structure being resistant to dissociation. Electrochemistry. Cyclic voltammograms (CVs) were recorded for the [M(pyNO)4] (M = Th, U) complexes and are presented in Figure 3. For comparison, the CV for [Ce(pyNO)4], which has been reported previously,15 was measured under the same conditions and is also included here.
Table 1. Experimental and Optimized Average M−O and M−N Bond Lengths in Å for [M(pyNO)4]0/+ M−O [Ce(pyNO)4] [Th(pyNO)4] [U(pyNO)4] [U(pyNO)4]+
M−N
Calc
Expt
Calc
Expt
2.25 2.32 2.25 2.16
2.24 2.30 2.25 2.14
2.63 2.67 2.62 2.57
2.54 2.61 2.54 2.53
metrics were in reasonable agreement with XRD data, with M− O and M−N bond distances within 1 and 4% of experiment, respectively, lending validity to the theoretical modeling. The relatively large differences (∼0.04−0.09 Å) between theoretical and experimental M−N bond lengths were within the errors for the applied methods48 and also observed in other lanthanide and actinide compounds.49−52 Detailed geometrical parameters for the one-electron reduction products [M(pyNO)4]− (M = Ce, Th, U) are shown in Table S1. The calculated and formal spin densities on the metal ions for [M(pyNO)4]−/0/+ are shown in Table 2, where the formal charge on the pyNO ligand was assumed to be −1. For
Figure 3. CVs obtained for [M(pyNO)4] (M = Ce, Th, U) in THF with 0.1 M [nPr4N][BArF4].
Sweeps to negative potential in the CV of [Ce(pyNO)4] revealed the CeIII/IV couple at E1/2 = −1.82 V versus Fc/Fc+. Cathodic scans showed poorly reversible, ligand-centered oxidation processes with major waves at Ec = −0.45 and +0.62 V. The CV for [Th(pyNO)4] was similar to that for [Ce(pyNO)4], with poorly reversible ligand-centered oxidation processes observed at Ec −0.57 and +0.51 V. As expected, C
DOI: 10.1021/acs.inorgchem.7b03238 Inorg. Chem. XXXX, XXX, XXX−XXX
Article
Inorganic Chemistry Table 2. Calculated and Formal Spin Densities on the Metal Centers for [M(pyNO)4]−/0 (M = Ce, Th, U) and [U(pyNO)4]+ [M(pyNO)4]− Ce Th U
[M(pyNO)4]+
[M(pyNO)4]
Calc
Formal
Calc
Formal
Calc
Formal
1.01 0.06 2.73
1.0 1.0 3.0
0.00 0.00 2.11
0.0 0.0 2.0
1.22
1.0
[M(pyNO)4] (M = Ce, Th, U), all three metal ions were unambiguously calculated as being in the +4 oxidation state. Upon one-electron reduction, for [M(pyNO)4]− (M = Ce, U) only slight deviations between calculated spin densities and formal ones were observed, indicating metal-centered reduction to the +3 metal oxidation state. In contrast, [Th(pyNO)4]− shows negligible spin density on Th, indicating pyNO−centered reduction. This confirms general chemical expectations given the well-known high stability of the ThIV ion. For the oneelectron oxidation product [U(pyNO)4]+, the spin density of 1.22 on U agreed well with the formal one, indicative of a UV center, consistent with the electrochemical, solid state, and spectroscopic data obtained for [U(pyNO)4][BArF4] (vide supra). Selected calculated redox couples are shown in Table 3 with the absolute redox potentials of the redox couples in Table S3.
Figure 4. Frontier molecular orbital energy level diagram for [M(pyNO)4]− (M = Ce, Th, U). The envelop of singly occupied MOs (SOMO, SOMO+1, and SOMO+2) is given with contour values of 0.03 au with the Z direction labeled with an arrow. Note: only alpha orbital energies for the lowest unoccupied MO (LUMO) and highest occupied MO (HOMO) are shown.
appreciable shortening of the Th−O and Th−N bond distances (see Table S1).
■
Table 3. Predicted Oxidation Potentials in V for [M(pyNO)4] (M = Ce, Th, U) in THF vs Fc+/0 [M(pyNO)4]0/− Ce Th U
Calc
Expt
−2.08 −4.10 −3.79
−1.82
CONCLUSIONS We have prepared the homoleptic complexes [M(pyNO)4] (M = Th, U) of the hydroxylamine pyNO − ligand and characterized these compounds by standard spectroscopic techniques. The uranyl complex [UO2(pyNO)3K]n was isolated from the reaction of UCl4 with K(pyNO). Electrochemical measurements performed on [U(pyNO)4] revealed a high degree of stabilization upon one-electron oxidation. The salt [U(pyNO)4][BArF4] was subsequently prepared; XRD and NIR measurements suggested a UV center. Computational data further supported this assignment. These results emphasize the utility of the pyNO ligand in stabilizing high-valent metal ions. We expect that these results will find utility in redox-based separations of f-block elements.
[M(pyNO)4]+/0 Calc
Expt
−0.98
−1.00
The predicted potentials for the CeIII/IV and UIV/V couples are −2.08 and −0.98 eV, respectively, in agreement with the experimental values of −1.82 and −1.00 V. The differences of ≤0.26 V were within typical errors for the methods applied and may be attributed in part to insufficient consideration of solvation effects, neglect of multiplet effects of the fn complexes, and lack of counterions.16,22−27 One-electron reduction potentials for [M(pyNO)4] (M = U, Th) were predicted as −3.79 and −4.10 V, respectively. These values are well outside the electrochemical discharge limits of THF, consistent with the absence of any reduction processes in the CVs of these complexes (Figure 3). The frontier molecular orbital energy level diagram for [M(pyNO)4]− (M = Ce, Th, U) is displayed in Figure 4. For the CeIII complex, [Ce(pyNO)4]−, the singly occupied molecular orbital (SOMO) was calculated as localized entirely on Ce and was of 4fδ character. For U(pyNO)4]−, calculations revealed three unpaired 5f electrons of 5fπ, 5fσ, and 5fϕ character. The electron density was located exclusively on U for the SOMO and SOMO+1, while for the SOMO+2, slight metal-to-ligand backbonding interactions resulted in a small spin density distribution on the pyNO− ligands. This resulted in the slightly lower than expected spin density value of 2.73 (see Table 2). In contrast to [M(pyNO)4]− (M = Ce, U), the SOMO for [Th(pyNO)4]− was distributed predominantly on the pyridine rings of the pyNO− ligands, which resulted in
■
EXPERIMENTAL SECTION
General Methods. All manipulations were carried out in an N2 atmosphere glovebox. Glassware was oven-dried for at least 3h at 150 °C prior to use. NMR spectra were recorded on Bruker Avance 300 and 500 MHz spectrometers. 1H and 13C chemical shifts are reported in ppm relative to tetramethylsilane using residual solvent as an internal standard. Elemental analyses were performed using a Costech ECS 4010 analyzer. Caution! Natural thorium (primary isotope 232Th) is a weak α-emitter (4.012 MeV) with a half-life of 1.41 × 1010 years and depleted uranium (primary isotope 238U) is a weak α-emitter (4.197 MeV) with a half-life of 4.47 × 109 years; manipulations and reactions should be carried out in monitored f ume hoods or in an inert atmosphere drybox in a radiation laboratory equipped with counting equipment. Materials. Solvents were sparged for 20 min with argon and dried using a commercial two-column solvent purification system comprising of columns packed with Q5 reactant (neutral alumina for hexanes) and stored over 4 Å molecular sieves. Deuterated solvents were purchased from Cambridge Isotope Laboratories Inc., degassed, and dried over activated 4 Å molecular sieves for at least 24 h prior to use. All reagents were purchased from commercial suppliers and used without further purification unless otherwise noted. D
DOI: 10.1021/acs.inorgchem.7b03238 Inorg. Chem. XXXX, XXX, XXX−XXX
Article
Inorganic Chemistry [ThCl4(DME)2],53 [U(N(SiMe3)2)3],54 UCl4,55 pyNOH,56 and K(pyNO),15 were synthesized according to literature procedures. Electrochemistry. Cyclic Voltammetry (CV) experiments were performed using a CH-Instruments 620D potentiostat. Cell resistances were all
