Nitrogenase is one of the most fascinating enzymes in nature, being responsible for all biological nitrogen reduction. Despite decades of research it is among the enzymes in bioinorganic chemistry whose mechanism is the most poorly understood. The MoFe protein of nitrogenase contains an iron-molybdenum-sulfur cluster, FeMoco, where N 2 reduction takes place. The resting state of FeMoco has been characterized by crystallography, multiple spectroscopic techniques and theory (broken-symmetry density functional theory) and all heavy atoms are now characterized. The cofactor charge, however, has been controversial, the electronic structure has proved enigmatic and little is known about the mechanism. While many computational studies have been performed on FeMoco, few have taken the protein environment properly into account. In this study, we put forward QM/MM models of the MoFe protein from Azotobacter vinelandii, centered on FeMoco. By a detailed analysis of the FeMoco geometry and comparing to the atomic resolution crystal structure we conclude that only the [MoFe 7 S 9 C] 1-charge is a possible resting state charge. Further we find that, of the 3 lowest energy broken-symmetry solutions of FeMoco the BS7-235 spin isomer (where 235 refers to Fe atoms that are "spin-down") is the only one that can be reconciled with experiment. This is revealed by a comparison of the metal-metal distances in the experimental crystal structure, a rare case of spin-coupling phenomena being visible through the molecular structure. This could be interpreted as the enzyme deliberately stabilizing a specific electronic state of the cofactor, possibly for tuning specific reactivity on specific metal atoms. Finally, we show that the alkoxide group on the Mo-bound homocitrate must be protonated under resting state conditions; the presence of which has implications regarding the nature of FeMoco redox states as well as for potential substrate reduction mechanisms.3

Molybdenum nitrogenase is one of the most intriguing metalloenzymes in nature, featuring an exotic ironmolybdenum-sulfur cofactor, FeMoco, whose mode of action remains elusive. In particular, the molecular and electronic structure of the N 2 -binding E 4 state is not known. In this study we present theoretical QM/ MM calculations of new structural models of the E 4 state of molybdenum-dependent nitrogenase and compare to previously suggested models for this enigmatic redox state. We propose two models as possible candidates for the E 4 state. Both models feature two hydrides on the FeMo cofactor, bridging atoms Fe 2 and Fe 6 with a terminal sulfhydryl group on either Fe 2 or Fe 6 (derived from the S2B bridge) and the change in coordination results in local lower-spin electronic structure at Fe 2 and Fe 6 . These structures appear consistent with the bridging hydride proposal put forward from ENDOR studies and are calculated to be lower in energy than other proposed models for E 4 at the TPSSh-QM/MM level of theory. We critically analyze the DFT method dependency in calculations of FeMoco that has resulted in strikingly different proposals for this state. Importantly, dinitrogen binds exothermically to either Fe 2 or Fe 6 in our models, contrary to others, an effect rationalized via the unique ligand field (from the hydrides) at the Fe with an empty coordination site. A low-spin Fe site is proposed as being important to N 2 binding. Furthermore, the geometries of these states suggest a feasible reductive elimination step that could follow, as experiments indicate. Via this step, two electrons are released, reducing the cofactor to yield a distorted 4-coordinate Fe 2 or Fe 6 that partially activates N 2 . We speculate that stabilization of an N 2 -bound Fe(I) at Fe 6 (not found for Fe 2 model) via reductive elimination is a crucial part of N 2 activation in nitrogenases, possibly aided by the apical heterometal ion (Mo or V). By using protons from the sulfhydryl group (to regenerate the sulfide bridge between Fe 2 and Fe 6 ) and the nearby homocitrate hydroxy group, we calculate a plausible route to yield a diazene intermediate. This is found to be more favorable with the Fe 6 -bound model than the Fe 2 -bound model; however, this protonation is uphill in energy, suggesting protonation of N 2 might occur later in the catalytic cycle or via another mechanism. Electronic supplementary information (ESI) available: Further information on all broken-symmetry solutions calculated for all models. Details of the QM/MM model preparation. Spin populations of all models. Localized orbital analysis of selected models. Geometry analysis of E 0 state calculated with different functionals and electronic structure analysis. Cartesian coordinates for the QM regions of all optimized structures available as XYZ les. See

A new 1.2 Å crystal structure of vanadium nitrogenase, isolated under turnover conditions, recently revealed a light atom ligand (OH or NH) replacing the bridging S2B sulfide of the FeV cofactor. QM/MM calculations on the new structure now reveal the light-atom ligand to be a bridging hydroxo group, probably derived from water binding to the cofactor.

The open-shell electronic structure of iron–sulfur clusters presents considerable challenges to quantum chemistry, with the complex iron–molybdenum cofactor (FeMoco) of nitrogenase representing perhaps the ultimate challenge for either wavefunction or density functional theory. While broken-symmetry density functional theory has seen some success in describing the electronic structure of such cofactors, there is a large exchange–correlation functional dependence in calculations that is not fully understood. In this work, we present a geometric benchmarking test set, FeMoD11, of synthetic spin-coupled Fe–Fe and Mo–Fe dimers, with relevance to the molecular and electronic structure of the Mo-nitrogenase FeMo cofactor. The reference data consists of high-resolution crystal structures of metal dimer compounds in different oxidation states. Multiple density functionals are tested on their ability to reproduce the local geometry, specifically the Fe–Fe/Mo–Fe distance, for both antiferromagnetically coupled and ferromagnetically coupled dimers via the broken-symmetry approach. The metal–metal distance is revealed not only to be highly sensitive to the amount of exact exchange in the functional but also to the specific exchange and correlation functionals. For the antiferromagnetically coupled dimers, the calculated metal–metal distance correlates well with the covalency of the bridging metal–ligand bonds, as revealed via the corresponding orbital analysis, Hirshfeld S/Fe charges, and Fe–S Mayer bond order. Superexchange via bridging ligands is expected to be the dominant interaction in these dimers, and our results suggest that functionals that predict accurate Fe–Fe and Mo–Fe distances describe the overall metal–ligand covalency more accurately and in turn the superexchange of these systems. The best performing density functionals of the 16 tested for the FeMoD11 test set are revealed to be either the nonhybrid functionals r2SCAN and B97-D3 or hybrid functionals with 10–15% exact exchange: TPSSh and B3LYP*. These same four functionals are furthermore found to reproduce the high-resolution X-ray structure of FeMoco well according to quantum mechanics/molecular mechanics (QM/MM) calculations. Almost all nonhybrid functionals systematically underestimate Fe–Fe and Mo–Fe distances (with r2SCAN and B97-D3 being the sole exceptions), while hybrid functionals with >15% exact exchange (including range-separated hybrid functionals) overestimate them. The results overall suggest r2SCAN, B97-D3, TPSSh, and B3LYP* as accurate density functionals for describing the electronic structure of iron–sulfur clusters in general, including the complex FeMoco cluster of nitrogenase.

The nitrogenase enzymes are responsible for all biological nitrogen reduction. How this is accomplished at the atomic level, however, has still not been established. The molybdenum-dependent nitrogenase has been extensively studied and is the most active catalyst for dinitrogen reduction of the nitrogenase enzymes. The vanadium-dependent form, on the other hand, displays different reactivity, being capable of CO and CO 2 reduction to hydrocarbons. Only recently did a crystal structure of the VFe protein of vanadium nitrogenase become available, paving the way for detailed theoretical studies of the iron–vanadium cofactor (FeVco) within the protein matrix. The crystal structure revealed a bridging 4-atom ligand between two Fe atoms, proposed to be either a CO 3 2– or NO 3 – ligand. Using a quantum mechanics/molecular mechanics model of the VFe protein, starting from the 1.35 Å crystal structure, we have systematically explored multiple computational models for FeVco, considering either a CO 3 2– or NO 3 – ligand, three different redox states, and multiple broken-symmetry states. We find that only a [VFe 7 S 8 C(CO 3 )] 2– model for FeVco reproduces the crystal structure of FeVco well, as seen in a comparison of the Fe–Fe and V–Fe distances in the computed models. Furthermore, a broken-symmetry solution with Fe2, Fe3, and Fe5 spin-down (BS7-235) is energetically preferred. The electronic structure of the [VFe 7 S 8 C(CO 3 )] 2– BS7-235 model is compared to our [MoFe 7 S 9 C] − BS7-235 model of FeMoco via localized orbital analysis and is discussed in terms of local oxidation states and different degrees of delocalization. As previously found from Fe X-ray absorption spectroscopy studies, the Fe part of FeVco is reduced compared to FeMoco, and the calculations reveal Fe5 as locally ferrous. This suggests resting-state FeVco to be analogous to an unprotonated E 1 state of FeMoco. Furthermore, V–Fe interactions in FeVco are not as strong compared to Mo–Fe interactions in FeMoco. These clear differences in the electronic structures of otherwise similar cofactors suggest an explanation for distinct differences in reactivity.

The biological conversion of N2 to NH3 is accomplished by the nitrogenase family, which is collectively comprised of three closely related but unique metalloenzymes. In the present study, we have...

Inventing new ways of exploring the materials phase space accelerates functional materials discovery. For new breakthroughs materials, faster exploration of bigger phase spaces is a key goal. High throughput computational screening (HTCS) is widely used to quickly search for materials with the right functional property. In this article we redefine HTCS methods to combine deep learning and physics based model to explore much large chemical spaces than possible by pure physics driven HTCS. Deep generative models are used to autonomously create materials libraries with high likelihood of desired properties, inverting the design paradigm. Additionally machine learnt surrogates enable next layer of screening to prune the set further such as high quality quantum mechanical simulations can be performed. With organic photovoltaic (OPV) molecules as test bench, we show the power of this redesigned HTCS approach in inverse design OPV molecules with very limited computational expense.