31P Nuclear Magnetic Resonance Spectroscopy as a Probe of Thorium–Phosphorus Bond Covalency: Correlating Phosphorus Chemical Shift to Metal–Phosphorus Bond Order

We report the use of solution and solid-state 31P Nuclear Magnetic Resonance (NMR) spectroscopy combined with Density Functional Theory calculations to benchmark the covalency of actinide-phosphorus bonds, thus introducing 31P NMR spectroscopy to the investigation of molecular f-element chemical bond covalency. The 31P NMR data for [Th­(PH2)­(TrenTIPS)] (1, TrenTIPS = {N­(CH2CH2NSiPri 3)3}3–), [Th­(PH)­(TrenTIPS)]­[Na­(12C4)2] (2, 12C4 = 12-crown-4 ether), [{Th­(TrenTIPS)}2(μ-PH)] (3), and [{Th­(TrenTIPS)}2(μ-P)]­[Na­(12C4)2] (4) demonstrate a chemical shift anisotropy (CSA) ordering of (μ-P)3– > (PH)2– > (μ-PH)2– > (−PH2)1– and for 4 the largest CSA for any bridging phosphido unit. The B3LYP functional with 50% Hartree–Fock mixing produced spin–orbit δiso values that closely match the experimental data, providing experimentally benchmarked quantification of the nature and extent of covalency in the Th–P linkages in 1–4 via Natural Bond Orbital and Natural Localized Molecular Orbital analyses. Shielding analysis revealed that the 31P δiso values are essentially only due to the nature of the Th–P bonds in 1–4, with largely invariant diamagnetic but variable paramagnetic and spin–orbit shieldings that reflect the Th–P bond multiplicities and s-orbital mediated transmission of spin–orbit effects from Th to P. This study has permitted correlation of Th–P δiso values to Mayer bond orders, revealing qualitative correlations generally, but which should be examined with respect to specific ancillary ligand families rather than generally to be quantitative, reflecting that 31P δiso values are a very sensitive reporter due to phosphorus being a soft donor that responds to the rest of the ligand field much more than stronger, harder donors like nitrogen.