Homoleptic Fe(III) and Fe(IV) Complexes of a Dianionic C3-Symmetric Scorpionate

Abstract

High-valent iron species have been implicated as key intermediates in catalytic oxidation reactions, both in biological and synthetic systems. Many heteroleptic Fe(IV) complexes have now been prepared and characterized, especially using strongly π-donating oxo, imido, or nitrido ligands. On the other hand, homoleptic examples are scarce. Herein, we investigate the redox chemistry of iron complexes of the dianonic tris-skatylmethylphosphonium (TSMP^2-) scorpionate ligand. One-electron oxidation of the tetrahedral, bis-ligated [(TSMP)₂Fe^II]^2- leads to the octahedral [(TSMP)₂Fe^III]^-. The latter undergoes thermal spin-cross-over both in the solid state and solution, which we characterize using superconducting quantum inference device (SQUID), Evans method, and paramagnetic nuclear magnetic resonance spectroscopy. Furthermore, [(TSMP)₂Fe^III]^- can be reversibly oxidized to the stable high-valent [(TSMP)₂Fe^IV]⁰ complex. We use a variety of electrochemical, spectroscopic, and computational techniques as well as SQUID magnetometry to establish a triplet (S = 1) ground state with a metal-centered oxidation and little spin delocalization on the ligand. The complex also has a fairly isotropic g-tensor (g_iso = 1.97) combined with a positive zero-field splitting (ZFS) parameter D (+19.1 cm^-1) and very low rhombicity, in agreement with quantum chemical calculations. This thorough spectroscopic characterization contributes to a general understanding of octahedral Fe(IV) complexes.

Chart 1. Reported Homoleptic Fe(IV) Complexes^,,,−

Scheme 1. Synthesis of Metal Complexes 2a,b, 3a–c, and 4

Figure 1

Molecular structure of complexes 3a and 4 derived from single crystal X-ray diffraction. Displacement ellipsoids are drawn at 30% probability level. Fused benzene rings are shown in a wireframe style for clarity. Counterions, solvent molecules, and hydrogens are omitted for clarity. Selected bond distances (Å) and angles (degrees): 3a: two molecular fragments in an asymmetric unit, fragment 1: Fe1–N11 1.9885(12), Fe1–N21 2.0019(12), Fe1–N31 1.9805(12), N11^∧Fe1^∧N21 91.11(5), N21^∧Fe1^∧N31 90.70(5), N31^∧Fe1^∧N11 90.73(5); fragment 2: Fe2–N12 1.9889(12), Fe2–N22 1.9885(12), Fe2–N32 1.9552(12), N12^∧Fe2^∧N22 91.56(5), N22^∧Fe2^∧N32 90.44(5), N32^∧Fe2^∧N12 91.16(5); 4: molecule has C_i symmetry, Fe1–N1 1.966(5), Fe1–N2 1.975(6), Fe1–N3 1.966(6), N1^∧Fe1^∧N2 91.6(2), N2^∧Fe1^∧N3 91.0(2), N3^∧Fe1^∧N1 90.7(2).

Figure 2

Cyclic voltammograms of compound 2b (ca. 8 mM solution) in 0.1 M ⁿBu₄NPF₆ acetonitrile electrolyte. Potentials are referenced with respect to the Fc/Fc⁺ redox couple. Left panel: overview scans at the rate of 100 mV/s; the full scan starts from an open-circuit potential of −2.12 V). Middle panel: UV–vis of independently synthesized compounds 2b, 3a (in MeCN), and 4 (in DCM) and time-dependent UV–vis spectra of controlled electrolysis at points A (−0.64 V) and B (−0.37 V) indicated on the CV on the left panel. Absorption in spectrum B was truncated due to detector saturation. The asterisks indicate isosbestic points that support a clean conversion between 3a and 4. Right panel: the quasi-reversible redox pair A–C centered at E_1/2 = −0.41 V; the inset shows linear dependence of the peak current vs square root of the scan rate.

Figure 3

SQUID measurement of 3a under an applied magnetic field of 1 T. The dots represent experimental data, and the solid red line the fit with the following parameters: S = ¹/₂, g_iso = 2.30, zJ = −6 cm^–1; S = ⁵/₂, D = E = 0, g_iso = 2.00, and TIP = 0.

Figure 4

Zero-field Mössbauer spectrum of 3a measured on a powder sample at 80 K. The fitted parameters are δ = 0.25 mm/s and |ΔE_Q| = 1.63 mm/s.

Figure 5

SCO curve for the [(TSMP)₂Fe^III]⁻ (3) complex as obtained by the Evans method in different solvents. Data points represent experimental measurements, whereas smooth curves are fits based on the regressive model in eq 1. Explored temperature ranges are limited by the freezing and boiling points of the respective solvents or precipitation of the compound at low temperatures.

Figure 6

Overlay of CVs measured for 3a and 5a. The measurements were performed in ca. 8 mM solution in 0.1 M ⁿBu₄NPF₆ acetonitrile electrolyte. Potentials are referenced with respect to the Fc⁺/Fc redox couple. The onsets of oxidation of 3a and 5a (E_B and E_B′, respectively) are defined as the points of intersection between extrapolated baseline and a tangent to the oxidation feature B/B′.

Figure 7

Zero-field Mössbauer spectrum of 4 measured on a powder sample at 80 K. The fitted parameters are δ = 0.04 mm/s and |ΔE_Q| = 1.96 mm/s.

Figure 8

Experimental (top panel) and simulated (bottom panel) L_2,3 edge XAS spectra of the Fe(III) (3a/3) and Fe(IV) (4) complexes.

Figure 9

VT μ_eff at 0.1 T (top) and VTVH magnetization at 1, 4, and 7 T (bottom) SQUID measurements of 4, fitted using the following spin Hamiltonian parameters: g_iso = 1.76, D = +15 cm^–1, E/D = 0.

Figure 10

THz-EPR spectra of 4. Relative absorbance spectra (black lines) are offset for the magnetic field B₀ at which they were measured. Simulations using D = +19.1 cm^–1, E = 0, and an isotropic g-value of 1.97 are shown in red. Calculated transition energies for magnetic fields applied parallel and perpendicular to the main anisotropy axis (z) are shown as solid green and dashed blue lines, respectively. Branch II corresponds to the formally forbidden transitions between the excited m_S = ±1 sublevels.

Figure 11

Quasi-restricted frontier orbitals (isocontour = 0.05) of the ground state of 4 calculated at the B3LYP-D3BJ/def2-TZVP (CP(PPP) for Fe) level of theory using a geometry optimized at the BP86-D3BJ/def2-TZVP level. Orbital energies are given in parentheses.

Chart 2. Comparison of the Quasi-Restricted Orbital (QRO) Energies of 4 and Its Si-tethered Analogue G Calculated at the B3LYP-D3BJ/def2-TZVP (CP(PPP) for Fe) Level of Theory for a Geometry Optimized at the BP86-D3BJ/def2-TZVP Level

Figure 12

¹H (400 MHz) and ²H (61 MHz) NMR spectra of 4 and its deuterated analogue 4-d₆ in DCM at 298 K. Only paramagnetic signals are assigned. The integrals are given in blue and were rounded to the nearest integer.

Figure 13

Top panel: variable-temperature ¹H NMR (400 MHz) δ^HFT products of 4 in DCM-d₂. Dots show experimental values, straight lines show linear fits (Section S13.3.2). Bottom panel: correlation plots of experimental vs calculated observed chemical shifts for 4 (Section S13.3.3).

Figure 14

Top panel: UV–vis–NIR spectra of complexes 2b, 3b, and 4 in solution at 298 K. Compound 2b was measured in acetonitrile,3b and 4 were measured in DCM. Bottom panel: experimental and TD-DFT-calculated (first 50 excitations) optical spectra of 4. Calculations were performed at the TPSSh-D3BJ/def2-SVP (def2-TZVP for Fe) level of theory in DCM. Gaussian broadening with FWHM of 170 nm was applied.