Introduction
Mössbauer spectroscopy is of central importance for studying the electronic structures of iron in diverse environments: from mononuclear coordination complexes in frozen solution, to inorganic cofactors in enzymes, to iron sites in bulk materials and catalytically active sites on surfaces or amorphous materials.1-9 The experimental study of iron in such systems is often complemented with a computational analysis, chiefly via density functional theory (DFT). As a result, the oxidation and spin states of the iron ion, its coordination number and the composition and symmetry of its immediate environment can be pinpointed. To translate the parameters predicted by DFT into those derived experimentally and assess the reliability of the DFT prediction, calibration studies for various combinations of density functionals, basis sets, grid sizes, relativistic options and solvation and dispersion corrections exist alongside very early work based on Hartree–Fock theory.10-22,23 Given this wealth of information, the reasons for presenting another calibration study may not be obvious. Our motivation to calibrate computational Mössbauer spectroscopy is threefold: (i) several technical advances have been made that are not included in previous calibration studies, e.g. newer basis sets, approximations and corrections; (ii) some of the more recent studies do not report on the quadrupole splitting;19,24 (iii) the emergence of single-atom catalysts, where the catalytically active iron site is embedded in an ill-defined carbonaceous environment with significant π-character,8 presents a challenging problem for computational Mössbauer spectroscopy and hence warrants a dedicated assessment of its predictive power for this specific coordination environment.
Single-atom catalysts (SACs) are materials synthesized from metal, nitrogen and carbon precursors with at least one pyrolysis step. Typically, the resulting MeNC catalysts are amorphous, carbonaceous materials with multiple phases and contain only small amounts of metal (< 5 wt%) in various forms. The synthesis is thought to result in MeNC active sites that contain individual metal ions coordinated by nitrogen.25-32 The catalytic transformations these SACs achieve include CO2conversion,33-35 selective C-H oxidation,36 hydrogen evolution and water oxidation as half-reactions of the water splitting reaction,37 and oxygen reduction.38-41 A specific subclass are so-called FeNC catalysts, where an iron center is most likely coordinated by several nitrogen donor atoms embedded in a graphene-like matrix.8,42-44 FeNC catalysts show high activity in the oxygen reduction reaction (ORR), a key reaction for fuel cells that are of central importance for green mobility applications. In fact, the ORR activity of recent FeNC catalysts is on par with that of low-platinum content electrodes, the current state of the art.45 However, the stability and hence the long-term usability of FeNC materials is still lacking.46-48 An important step towards a better understanding of this promising substitution material for platinum electrodes is to clarify the structure and electronic properties of the active site.39,43,49,50
The current consensus in the literature is that the catalytically active iron ion is surrounded by two to four nitrogen donor atoms embedded in a graphene sheet that may have structural or electronic defects and possibly an additional axial ligand.39,49,51,52 The sketch shown in Figure 1A attempts to summarize the types of environment discussed in the literature. It can be seen that several aspects are unclear: the electronic character of the N-donors, the presence and nature of axial ligands, the type and abundance of defects close to the active site, and the position of the active site within a sheet, at the edge or positioned between two sheets.39,44,53-58While other spectroscopies fall short due to the amorphous nature of the SAC material, Mössbauer spectroscopy is ideally equipped to study the coordination environments and electronic structures of iron sites within the amorphous material. However, the definitive assignment of structural characteristics proves difficult purely by comparison with reference data from model complexes. Therefore, a broad variety of active site candidates derived from density functional theory are needed to develop a deeper understanding of the influence that specific structural and electronic aspects of the active site may have.43,44,58-60 The long-term goal is to decipher the composition of the active site in its resting state and the changes it undergoes during catalysis in a joint effort of experimental and computational Mössbauer spectroscopy.