Computational Modeling of Octahedral Iron Oxide Clusters: Hexaaquairon(III) and Its Dimers

Abstract

Octahedral monomeric and dimeric iron oxide clusters represent the basic units in many iron oxide and oxide-hydroxide minerals. In this paper, we provide a detailed theoretical analysis of the structural and optical properties of the most important of these clusters in a vacuum and in an aqueous environment. An evaluation of various computational methods was performed on the experimentally well-known monomer [Fe(H2O)6]3+, and it is found that all methods provide similar and reliable structures. Most density functional theory (DFT) methods reasonably reproduce the spin-forbidden sextet–quartet d–d transition energy, which also resembles the lowest transition energies in many infinite octahedral iron oxide systems. On the other hand, Hartree–Fock (HF) and MP2 methods significantly overestimate this energy. The ligand-to-metal charge transfer (LMCT) energy is highly sensitive to the method employed, with the closest agreement with experiment provided by the BHandHLYP functional. Thermodynamic property calculations suggest that dimerization reactions starting from [Fe(H2O)6]3+ are highly exothermic in a vacuum. In contrast, these reactions have insignificant energy changes in solution, though the singly μ-oxo bridged dimer is slightly favored. The electrostatic repulsion between two charged monomers hinders their close contact. The singly μ-oxo bridged dimer suffers less from this because of its maximal Fe–Fe distance, which is consistent with the existence of stable crystal structures for this dimer. A comparison between the calculated structures and experimental results suggests that several dimer species coexist in solution. The calculated ferromagnetic and antiferromagnetic states of the dimers are found to have comparable energies and structures. While the singly μ-oxo and doubly μ-hydroxo bridged dimers have spin states that are well separated in energy, the spin states in the triply μ-hydroxo bridged dimer pack closely. The single excitation d–d transition in the dimer structure is comparable in energy to the d–d transition in the monomer, while the double excitation d–d transition, i.e., simultaneous excitation of two iron centers, has a higher excitation energy that is 1.6–2.6 times the single excitation energy but below the LMCT energy. This means that doubly excited states can be populated during the non-radiative relaxation of iron oxide clusters following initial photoexcitation of the LMCT state.