Interpretation of the electronic spectra of the confacial bioctahedral nonahalides [Ru2Cl9]3– and [Ru2Br9]3– using the SCF-Xα-SW method: location of the σ→σ* transition

Abstract

Considerable progress has been made in assigning the UV/VIS/near-IR electronic spectra (5000–50 000 cm^–1) of [Ru₂Cl₉]^3– and [Ru₂Br₉]^3–, with the aid of SCF-Xα-SW calculations. A conceptual framework within which band energies and relative intensities can be discussed was constructed by reference to the corresponding [MX₆]^z– monomers. First, an empirical correction (δ_c.t.) of 7500 cm^–1 has been established for Xα-computed energies of frankly charge-transfer (c.t.) transitions, based on the discovery that the calculated X → M charge-transfer (x.m.c.t.) transitions are faithfully linearly correlated (unity gradient) with observed band energies for the congeneric t_2g⁵ complexes [RuCl₆]^3–, [RuBr₆]^3–, [IrCl₆]^2–, and [IrBr₆]^2–. Secondly, the calculated oscillator strengths for c.t. transitions of monomeric hexahalides successfully model observed band intensities, and can be understood in terms of the mutual overlap of ligand-based components of the donor and acceptor orbitals. For binuclear [Ru₂Cl₉]^3– and [Ru₂Br₉]^3– the intense absorption bands above 30 000 cm^–1 are readily assigned to x.m.c.t. transitions to the {7e″, 9e′} levels, derived from the single-ion e_g orbitals. In contrast, the anticipated c.t. to the t_2g-derived levels (i.e. to unfilled 5a₂″) loses intensity in the trigonal field of the confacial complexes, and the prominent near-UV/VIS features are assigned instead to transitions within the metal-metal orbital manifold. In particular, the binuclear σ→σ* transition is located at 22 500 cm^–1 for [Ru₂Cl₉]^3– and 20 000 cm^–1 for [Ru₂Br₉]^3–, well above the calculated values. The discrepancy between observed and calculated σ→σ* transition energies arises through the neglect of electron-correlation effects in the Xα calculations. The implied value of the two-electron exchange term K is about 7800 cm^–1 for both complexes. Thus, two major empirical adjustments (δ_c.t. and K) are required in the present analysis due to the limitations of the Xα method, and the status of these measures is examined.