Ryan R.
Langeslay
,
Megan E.
Fieser
,
Joseph W.
Ziller
,
Filipp
Furche
* and
William J.
Evans
*
Department of Chemistry, University of California, Irvine, California 92697-2025, USA. E-mail: wevans@uci.edu; filipp.furche@uci.edu; Fax: +1-949-824-2210; Tel: +1-949-824-5174
First published on 3rd November 2014
Reduction of the Th^{3+} complex Cp^{′′}_{3}Th, 1 [Cp′′ = C_{5}H_{3}(SiMe_{3})_{2}], with potassium graphite in THF in the presence of 2.2.2-cryptand generates [K(2.2.2-cryptand)][Cp^{′′}_{3}Th], 2, a complex containing thorium in the formal +2 oxidation state. Reaction of 1 with KC_{8} in the presence of 18-crown-6 generates the analogous Th^{2+} compound, [K(18-crown-6)(THF)_{2}][Cp^{′′}_{3}Th], 3. Complexes 2 and 3 form dark green solutions in THF with ε = 23000 M^{−1} cm^{−1}, but crystallize as dichroic dark blue/red crystals. X-ray crystallography revealed that the anions in 2 and 3 have trigonal planar coordination geometries, with 2.521 and 2.525 Å Th–(Cp′′ ring centroid) distances, respectively, equivalent to the 2.520 Å distance measured in 1. Density functional theory analysis of (Cp^{′′}_{3}Th)^{1−} is consistent with a 6d^{2} ground state, the first example of this transition metal electron configuration. Complex 3 reacts as a two-electron reductant with cyclooctatetraene to make Cp^{′′}_{2}Th(C_{8}H_{8}), 4, and [K(18-crown-6)]Cp′′.
Nevertheless, it was recently discovered that the +2 oxidation state is accessible in soluble molecular complexes for all the elements in the lanthanide series except promethium, eqn (1).^{1} Previously, it was thought that only the traditional six Ln^{2+} ions of Eu, Yb, Sm, Tm, Dy, and Nd were obtainable in solution on the basis of calculated reduction potentials^{2} and solid state chemistry.^{3}
(1) |
Extension of this reductive chemistry to uranium was not initially tried since it is well known that the redox chemistry of uranium, which includes multiple oxidation states, +3, +4, +5, and +6, is quite different from that of the rare earths. Although it was likely that uranium would be different, an analogous synthesis was eventually attempted and the first fully characterizable U^{2+} complex, [K(2.2.2-cryptand)][Cp^{′}_{3}U] (Cp′ = C_{5}H_{4}SiMe_{3}), was isolated according to eqn (2).^{4}
(2) |
Synthesis of a Th^{2+} complex viaeqn (1) or (2) seemed even more unlikely for several reasons. Complexes of Th^{3+} are already difficult to obtain. The Th^{4+}/Th^{3+} redox potential is estimated to be −3.0 and −3.8 V vs. NHE^{5} and a Th^{3+}/Th^{2+} redox potential of −4.9 V vs. NHE is in the literature.^{6} Reduction to metallic thorium would be predicted to be favored before formation of a Th^{2+} species.^{6} Many studies have been reported to find oxidation states lower than +4 for thorium,^{7} but only five Th^{3+} complexes have ever been structurally characterized.^{7k–o} An analog of eqn (2) was not possible since Cp^{′}_{3}Th has not yet been synthesized. Despite these issues, thorium reduction chemistry was examined using Cp^{′′}_{3}Th [Cp′′ = C_{5}H_{3}(SiMe_{3})_{2}-1,3],^{7k} prepared by Lappert et al. in 1986, and the results are described here.
Addition of potassium graphite to a dark blue solution of Cp^{′′}_{3}Th, 1, and 2.2.2-cryptand in THF immediately forms a green solution from which dichroic dark blue/red crystals of [K(2.2.2-cryptand)][Cp^{′′}_{3}Th], 2, can be isolated and crystallographically characterized, Fig. 1, eqn (3). The analogous reaction with 18-crown-6 instead of 2.2.2-cryptand as the potassium chelator provides [K(18-crown-6)(THF)_{2}][Cp^{′′}_{3}Th], 3, which was also crystallographically characterized [see (ESI†)], eqn (3). Elemental analysis was consistent with the structures determined crystallographically. The ^{1}H and ^{13}C NMR spectra of 2 and 3 gave resonances in the diamagnetic region with a Me_{3}Si ^{1}H NMR resonance shifted about 0.4 ppm from that of KCp′′. A resonance was observed in the ^{29}Si NMR spectrum of 3 at −6 ppm in the region close to the −8 and −15.5 ppm signals of diamagnetic Cp^{′′}_{3}ThBr and KCp′′, respectively. Evans method measurements^{8} on both 2 and 3 and SQUID analysis^{9} at low temperature suggest the [Cp^{′′}_{3}Th]^{1−} anion is diamagnetic. No EPR spectra were observed for 2 and 3. Decomposed samples showed the EPR spectrum of Cp^{′′}_{3}Th.^{7m}
(3) |
Fig. 1 Molecular structure of [K(2.2.2-cryptand)][Cp^{′′}_{3}Th], 2. Thermal ellipsoids are drawn at the 50% probability level and hydrogen atoms are omitted for clarity. |
The structures of the anions in 2 and 3 are very similar to the structure of Cp^{′′}_{3}Th. All three structures have a trigonal planar arrangement of the three Cp′′ rings around thorium with a sum of (ring centroid)–Th–(ring centroid) angles of 360°. The structure of 2, however, is not isomorphous with the lanthanum complex of the same formula, [K(2.2.2-cryptand)][Cp^{′′}_{3}La].^{1a} The average Th–(Cp′′ ring centroid) distances of 2.521 Å in 2 and 2.525 Å in 3 are equivalent to the 2.520 Å distance in Cp^{′′}_{3}Th. The negligible differences in the Th–(ring centroid) distances between the Th^{3+} precursor and the formally Th^{2+} complexes 2 and 3 are similar to the small differences between the Cp^{′}_{3}Ln and Cp^{′′}_{3}Ln Ln^{3+} complexes and the (Cp^{′}_{3}Ln)^{1−} and (Cp^{′′}_{3}Ln)^{1−} complexes, respectively, of all the new Ln^{2+} ions that have 4f^{n}5d^{1} ground states^{1} instead of the 4f^{n+1} configurations expected by reduction of a 4f^{n} Ln^{3+} ion. Similarly, the 2.521 Å U–(ring centroid) distance in the U^{2+} complex, [K(2.2.2-cryptand)][Cp^{′}_{3}U], which appears to have a 5f^{3}6d^{1} ground state, is only slightly larger than the 2.508 Å value in the U^{3+} analog, Cp^{′}_{3}U.^{4} These small changes in M–(ring centroid) distances match the small changes in radial size commonly seen in transition metal complexes,^{10} but contrast with the 0.10–0.20 Å differences generally seen for complexes of 4f^{n+1} Ln^{2+} complexes compared to their 4f^{n} Ln^{3+} counterparts.^{11}
The UV-Vis spectra of 2 and 3 in THF, Fig. 2, contain absorptions at 650 nm with extinction coefficients of 23000 M^{−1} cm^{−1}, that are significantly larger than those of Cp^{′′}_{3}Th, 5000 M^{−1} cm^{−1}. This is similar to the larger intensities observed for the +2 complexes, [K(2.2.2-cryptand)][Cp^{′}_{3}Ln]^{1c,d} and [K(2.2.2-cryptand)][Cp^{′}_{3}U],^{4} compared to their +3 analogs, Cp^{′}_{3}Ln and Cp^{′}_{3}U, respectively. However, the absorptions of the Th^{2+} complexes are even more intense and the solutions look like ink.
Density functional theory (DFT) using the TPSSh functional^{12} was used to examine the (Cp^{′′}_{3}Th)^{1−} anion in 2 and 3. Calculations using scalar-relativistic effective core potentials^{13} and triple-zeta valence basis sets, def-TZVP, for thorium^{14} predicted trigonal planar structures for Cp^{′′}_{3}Th and (Cp^{′′}_{3}Th)^{1−} that match the crystallographic data. The calculated Th–Cp′′(centroid) lengths of 2.538 Å for Cp^{′′}_{3}Th and 2.526 Å for (Cp^{′′}_{3}Th)^{1−} are similar to the experimentally determined distances of 2.52 Å. It is interesting to note that the calculations for the Th^{3+} complex show a slightly longer metal ligand distance than for the Th^{2+} complex. The calculations indicate a spin-paired ground state of 6d^{2} for (Cp^{′′}_{3}Th)^{1−} and a 6d^{1} ground state for Cp^{′′}_{3}Th; the latter is consistent with previous analyses of Cp^{′′}_{3}Th,^{7g,m} (C_{5}Me_{5})_{2}[^{i}PrNC(Me)N^{i}Pr]Th^{7n} and [K(DME)_{2}]{[C_{8}H_{6}(Si^{t}BuMe_{2})_{2}]_{2}Th}.^{7l} Gas-phase studies of Th^{2+} indicate a ground state of 5f^{1}6d^{1}, but the 6d^{2} configuration is just 63 cm^{−1} higher and the 5f^{1}7s^{1} is 2527 cm^{−1} higher than the ground state.^{15} For (Cp_{3}Th)^{1−} the triplet 5f^{1}6d^{1} state is computed to be 9–14 kcal mol^{−1} higher in energy than the singlet 6d^{2} ground state.
The 6d^{2} singlet ground state can arise in this case due to stabilization of a d_{z2} orbital by the trigonal ligand environment as found in DFT calculations on (Cp^{′}_{3}Ln)^{1−} and (Cp^{′}_{3}U)^{1−} complexes^{1c,d,4} and noted earlier in the literature for tris(cyclopentadienyl) metal complexes.^{7g,m,16} Indeed, both the lowest unoccupied molecular orbital (LUMO) of Cp^{′′}_{3}Th and the highest occupied molecular orbital (HOMO) of (Cp^{′′}_{3}Th)^{1−} have d_{z2} character, Fig. 3. Complexes 2 and 3 provide the first examples of the 6d^{2} configuration since stable transition metal ions are only known with the 5d^{n} configurations of the third row transition metals. The 6d^{2} configuration is that predicted for ions like Rf^{2+} and Db^{3+}.^{17}
Fig. 3 Contour plots of (a) the LUMO of Cp^{′′}_{3}Th and (b) the HOMO of the (Cp^{′′}_{3}Th)^{1−} anion in 3. Contour value is 0.05. |
Time-dependent density functional theory was used to simulate the UV-Vis spectra for the (Cp^{′′}_{3}Th)^{1−} anion as shown in Fig. 2 (see ESI† for a description of the predicted excitations). The maxima in the calculated spectra are lower in energy than those observed experimentally, but this is often the case with such calculations.^{18} Analysis of the calculated low energy peak shows that it arises from metal-to-metal transitions that have d → f and d → p character. The high energy peaks arise from metal-to-ligand charge transfer transitions similar to those found in the spectral analysis of (Cp^{′}_{3}Ln)^{1−}^{1b–d} and (Cp^{′}_{3}U)^{1−}.^{4} However, the d → f transitions found for (Cp^{′′}_{3}Th)^{1−} were not apparent in the analysis of the spectra of (Cp^{′}_{3}Ln)^{1b–d} and (Cp^{′}_{3}U)^{1−}.^{4}
The rate of decomposition of [K(18-crown-6)(THF)_{2}][Cp^{′′}_{3}Th], 3, at room temperature was studied by ^{1}H NMR spectroscopy since monitoring by UV-Vis spectroscopy is complicated by the formation of highly colored Cp^{′′}_{3}Th, as identified by X-ray crystallography.^{7k} The rate of decomposition of 3 is much slower than that of the U^{2+} complex, [K(2.2.2-cryptand)][Cp^{′}_{3}U], which has a half-life of 1.5 h in THF at room temperature.^{4} Complex 3 decomposed only 8% after 8 days at 298 K and a sample kept in the dark showed even less decomposition. This suggests that the formally Th^{2+} species are significantly more stable than the other newly discovered +2 ions.^{1d,4}
Complexes 2 and 3 were treated with H_{2} to determine if a Th^{3+} hydride complex such as “[K(2.2.2-cryptand)][Cp^{′′}_{3}ThH]” would form in analogy to the complex formed by reaction of [K(2.2.2-cryptand)][Cp^{′}_{3}U] with H_{2}.^{4} Analogous chemistry is not observed with either H_{2} or KH. Complexes 2 and 3 react in solution within minutes with 1 atm of H_{2} and also over several hours at 60 psi in the solid state^{19} to make EPR active new crystalline complexes that appear to be bimetallic, but suitable models for the crystallographic data on the products have not been obtainable. The reactivity of 2 and 3 with H_{2} contrasts with that of the Th^{3+} complex, Cp^{′′}_{3}Th, which does not react under analogous conditions.
The (Cp^{′′}_{3}Th)^{1−} anion displays net two-electron reduction chemistry in its reaction with 1,3,5,7-cyclooctatetraene (C_{8}H_{8}). The Th^{4+} complex Cp^{′′}_{2}Th(C_{8}H_{8}), 4, is formed as shown in eqn (4) and was characterized by X-ray crystallography, Fig. 4. The (C_{8}H_{8})^{2−} ring in 4, like that of (C_{5}Me_{4}H)_{2}U(C_{8}H_{8}),^{20} displays considerable distortion from the normal planar geometry with several atoms 0.095 Å out of the best plane of the eight carbon atoms. This is reflected by a large range of Th–C(C_{8}H_{8}) distances: 2.736(4) to 2.841(4) Å. This 0.105 Å range is similar to the 0.123 Å range in (C_{5}Me_{4}H)_{2}U(C_{8}H_{8}).^{20}
(4) |
The isolation of the formally Th^{2+} ion in (Cp^{′′}_{3}Th)^{1−} is likely aided by the stabilization of the potassium counter-cation by the 18-crown-6 and 2.2.2-cryptand ligands. This was also observed with U^{2+} in the (Cp^{′}_{3}U)^{1−} anion^{4} and in the (Cp^{′}_{3}Ln)^{1−} complexes of the new Ln^{2+} ions.^{1} In the absence of these potassium-stabilizing chelates, isolation of Th^{2+} appears to be more difficult as described in a 2001 paper by Lappert and co-workers on the formation of Cp^{′′}_{3}Th by Na–K reduction of Cp^{′′}_{3}ThCl.^{7m} In that paper, Lappert reports that treatment of Cp^{′′}_{3}ThCl with excess Na–K alloy caused the initially blue solution (presumably Cp^{′′}_{3}Th) to change to dark green. They isolated a diamagnetic green compound they postulated to be “[K(THF)_{x}][ThCp^{′′}_{3}] and/or ThCp^{′′}_{2}(THF)_{x}” but they could not characterize it or obtain reproducible analytical results. Hence, the (Cp^{′′}_{3}Th)^{1−} anion was probably generated over 10 years ago, but could not be isolated in pure form as a simple [K(THF)_{x}]^{1+} salt.
In summary, although it is difficult to obtain Th^{3+} complexes, further reduction is still possible with thorium: the +2 formal oxidation state of this metal is accessible in soluble molecular complexes. The Th^{2+} complexes provide the first examples of an isolable ion with a 6d^{2} electron configuration, the configuration possible for fourth row transition metal congeners of Hf^{2+} or Ta^{3+}. The synthesis of these complexes demonstrates the power of specific ligand fields to generate new ground states with actinides. The identification of Th^{2+} is more evidence that the oxidation state diversity for the f elements is still increasing. Stabilization of higher-lying d orbitals by the ligand field appears to be a key factor in isolating these new ions and provides a new option in expanding the oxidation state chemistry of these elements. This approach should be pursued further as attempts are made to synthesize soluble molecular complexes of +1 ions of these metals.
Footnote |
† Electronic supplementary information (ESI) available: Experimental and computational details; crystallographic data collection, structure solution, and refinement; and crystallographic data and complete bond distances and angles for compounds 1–4. CCDC 1018011–1018014. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c4sc03033h |
This journal is © The Royal Society of Chemistry 2015 |