Calibration of computational Mössbauer spectroscopy to unravel active sites in FeNC catalysts for the oxygen reduction reaction

Abstract

Single-atom catalysts with iron ions in the active site, known as FeNC catalysts, show high activity for the oxygen reduction reaction and hence hold promise for access to low-cost fuel cells. Because of the amorphous, multiphase structure of the FeNC catalysts, the iron environment and its electronic structure are poorly understood. While it is widely accepted that the catalytically active site contains an iron ion ligated by several nitrogen donors embedded in a graphene-like plane, the exact structural details, such as the presence or nature of axial ligands, are unknown. Computational chemistry in combination with Mössbauer spectroscopy can help unravel the geometric and electronic structures of the active sites. As a first step toward this goal, we present a calibration of computational Mössbauer spectroscopy for FeN 4-like environments. The uncertainty of both the isomer shift and the quadrupole splitting prediction is determined, from which trust regions for the Mössbauer parameter predictions of computational FeNC models are derived. We find that TPSSh, B3LYP, and PBE0 perform equally well; the trust regions with B3LYP are 0.13 mm s −1 for the isomer shift and 0.36 mm s −1 for the quadrupole splitting. The calibration data is made publicly available in an interactive notebook that provides predicted Mössbauer parameters with individual uncertainty estimates from computed contact densities and quadrupole splitting values. We show that a differentiation of common FeNC Mössbauer signals by a separate analysis of isomer shift and quadrupole splitting will most likely be insufficient, whereas their simultaneous evaluation will allow the assignment to adequate computational FeNC models.

Active Site Identification in FeNC Catalysts and Their Assignment to the Oxygen Reduction Reaction Pathway by In Situ57Fe Mössbauer Spectroscopy

“…B 5 : δ iso = 0.30 mm s −1 , Δ E Q = 2.10 mm s −1 oxy , δ iso = 0.74 mm s −1 , Δ E Q = 1.34 mm s −1 (de‐oxy)). [ 69 ] As mentioned earlier, enforcing an Fe(II) high‐spin electronic structure is possible and leads to Mössbauer parameters in better agreement with the D3 signal measured in all preparation routes (calc. : δ iso = 1.00 mm s −1 , Δ E Q = 4.63 mm s −1 ).…”

“…Based on the general definition of isomer shift and quadrupole splitting in combination with ligand field theory, some basic considerations regarding the expected changes can be made. It should be noted, as illustrated in Gallenkamp et al, [ 69 ] that neither isomer shift nor quadrupole splitting alone can be used for an unambiguous assignment of iron speciation. Only their combination allows for a robust assignment.…”

“…The ferrous high‐spin case is always easy to identify based on a significantly larger isomer shift compared to other ferrous and ferric environments. [ 35,69 ] Moreover, the electric field gradient (EFG) at the iron nucleus is expected to be largest for a ferric intermediate‐spin state, yielding the largest quadrupole splitting (of all three discussed environments), which is typically 3 mm s −1 or even larger. It should be noted that these general trends in the relation between Mössbauer parameters and electronic state apply to any iron environments independent of being (pseudo‐)molecular or inorganic.…”

Electrocatalysis with Single‐Metal Atom Sites in Doped Carbon Matrices

“…These doublets have initially been interpreted by comparison to doublets experimentally measured for unpyrolysed FeN4 macrocycles with known oxidation and spin states. Recently, DFT methods have been developed to calculate the QS and the isomer shift (IS) of different Fe-Nx hypothetical sites embedded in (disordered) graphene sheets, which rationalizes the interpretation of Mössbauer spectra of pyrolysed Fe-N-C SMAC[20,121,122].The DFT results nevertheless showed the ambiguity of the spectroscopic response. For example, three different sites (namely Fe(II)N4C12 site with spin state 0 or 1 but also Fe(III)N4C12 with spin state 5/2 or Fe(III)N4C10 with spin 5/2) all lead to QS-values around 1 mm s -1 , compatible with the experimental D1[20].…”

Germany’s Future in Theoretical and Computational Chemistry: a Special Issue Celebrating DEAL

“…Giovanni Li Manni et al [3] present an in‐depth analysis of strong correlation effects in a ferrous porphyrin model based on a set of multireference calculations with large active spaces. Addressing similar iron environments, Vera Krewald and coworkers [4] present a calibration study for Mössbauer spectroscopy including an interactive, web‐based notebook by Jonny Proppe for uncertainty estimates of DFT predictions.…”