Computational Mechanistic Study of [MoFe3S4] Cubanes for Catalytic

Article pubs.acs.org/IC

Cite This: Inorg. Chem. XXXX, XXX, XXX−XXX

Computational Mechanistic Study of [MoFe3S4] Cubanes for Catalytic Reduction of Nitrogenase Substrates Albert Th. Thorhallsson†,‡ and Ragnar Bjornsson*,†,‡ †

Science Institute, University of Iceland, Dunhagi 3, Reykjavik 107, Iceland Department of Inorganic Spectroscopy, Max-Planck-Institut für Chemische Energiekonversion, Stiftstrasse 34-36, Mülheim an der Ruhr 45470, Germany

‡

Downloaded via UNIV OF WINNIPEG on January 17, 2019 at 05:18:55 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

S Supporting Information *

ABSTRACT: Molybdenum-dependent nitrogenase is the most active biological catalyst for dinitrogen reduction. This reaction is catalyzed by a [MoFe7S9C] cofactor (FeMoco). FeMoco can be described as a double-cubane, with [MoFe3S3] and [Fe4S3] parts, bound via an interstitial carbide and three bridging sulfides. Model compounds have been synthesized since early studies of the enzyme and Coucouvanis and co-workers demonstrated that [MoFe3S4] cubanes are active catalysts for many substrates catalyzed by nitrogenase. These reactions include hydrazine reduction to ammonia and cis-dimethyldiazene reduction to methylamine. Experiments implicated molybdenum as the binding site but the mechanisms have not been studied by theoretical calculations before. Here we present a DFT study of the catalytic reaction mechanisms of hydrazine and cis-dimethyldiazene reduction with a [MoFe3S4] cubane. Like in the experiments, molybdenum is revealed as the likely substrate binding site, likely due to the labile ligand on Mo. For the hydrazine mechanism, a reduction event is centered on Fe, specifically on the Fe antiferromagnetically coupled to the mixed-valence pair. After protonation of the distal hydrazine nitrogen, the N−N bond can be cleaved to yield NH3 and a Mo-bound -NH2 intermediate. This is followed by another protonation/reduction step to give an -NH3 intermediate, and finally substituted by the substrate to complete the cycle. The computed mechanisms shed light on the bimetallic cooperativity in these systems where the reduction steps are localized on Fe while the substrate binds to Mo and the reductions require only a free coordination site (on Mo) and a favorable reduction event (to Fe). Although both substrates easily displace the weakly bound acetonitrile ligand, one reduction event is required for hydrazine activation and N−N bond cleavage to give an integer-spin -NH2 intermediate. An integer-spin -NH2 intermediate has been observed as a common intermediate for the enzyme reduction of hydrazine and diazene, suggesting a possible link to the enzyme chemistry.

■

INTRODUCTION Nitrogenases are fascinating metalloenzymes in bacteria that are responsible for all dinitrogen reduction in biology. The more active molybdenum-dependent nitrogenase features an iron−molybdenum cofactor (FeMoco) where dinitrogen binds and is reduced to ammonia via a mechanism that is not well understood.1,2 The structure of the cofactor was the subject of controversy due to the first crystal structure in 19923 showing a cavity in the cluster while the 2002 crystal structure4 indicated the presence of a light atom (carbon, nitrogen, or oxygen) inside the cavity. The interstitial atom question was finally resolved by Fe X-ray emission spectroscopy and crystallography5,6 and revealed to be carbon, fully revealing the cofactor as having a [MoFe7S9C] double-cubane structure, albeit with an unusual carbide ligand acting as glue between the [Fe3S3] and [MoFe3S3] cubane-like clusters, as well as 3 sulfur bridges. The vanadium-dependent nitrogenase is only expressed in a molybdenum-deficient environment and is catalytically less active. The iron−vanadium cofactor (FeVco) consists of a similar iron−sulfur structure with vanadium in © XXXX American Chemical Society

place of molybdenum and also curiously another ligand (either carbonate or nitrate according to a recent crystal structure)7 instead of a bridging sulfide but does contain an interstitial carbide like FeMoco.8 Much less is known about the iron-only nitrogenase; the catalytic activity is only a fraction of the Modependent nitrogenase, but recent work suggests they share some mechanistic aspects.9 Despite important progress in structural characterization of the proteins and cofactors of nitrogenase, much less is known about the actual mechanism of dinitrogen reduction or the mechanism for reduction of alternative substrates (such as hydrazine, diazene, acetylene, protons etc.), also catalyzed by nitrogenase. The binding site of N2 remains controversial with both Mo and Fe being suggested as binding sites10 and same applies to the binding of alternative substrates. The mechanism of proton reduction, the main side reaction of the enzyme, remains unclear as well, but bridging hydrides (likely on Fe Received: September 19, 2018

A

DOI: 10.1021/acs.inorgchem.8b02669 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

Figure 1. Overview of the substrate reductions catalyzed by FeMoco and [MoFe3S4] cubanes (the cubane shown, 1, has a tetrachlorocatecholate ligand and bound acetonitrile). Both systems catalyze the same reactions with the exception of the 8-electron N2 reduction, which is catalyzed only by FeMoco and results in both NH3 and H2 as products.

atoms) have been suggested as intermediates.2,11 The H2 evolution that is accompanied by N2-binding to the E4 state has been proposed to result from reductive elimination of bridging hydrides.2 As discussed in a recent review by one of the authors,12 structural studies of the enzymes have from the beginning been accompanied by synthetic studies of model compounds, mostly from the laboratories of Richard Holm13−15 and Dimitri Coucouvanis.16−18 The model compounds include single and double heterometallic (Mo,V) cubane compounds that share many of the structural features of FeMoco and FeVco.19,20 Not only are these model compounds able to mimic parts of the cofactor molecular structure but it has also been demonstrated that metal oxidation states and the presence of mixed-valence delocalized Fe pairs are shared by the enzyme cofactors and [MoFe3S4]/[VFe3S4] cubanes.12 [MoFe3S4] and [VFe3S4] cubanes have recently been used in X-ray spectroscopic studies to shed light on metal oxidation states in FeMoco and FeVco.19,20 Coucovanis and co-workers demonstrated that the single-cubane [MoFe3S4] and [VFe3S4] compounds acted as catalysts for certain reactions.21−23 Catalytic reduction of hydrazine, dimethyldiazene, acetylene and protons were all demonstrated by [MoFe3S4] cubanes; all of these substrates are also reduced by Mo-dependent nitrogenase (see Figure 1). Hydrazine was demonstrated to be a substrate for nitrogenase in the eighties 24,25 and Thorneley and Lowe demonstrated that acid or base quenching of nitrogenase under turnover conditions releases hydrazine.26 Furthermore, it is a minor reaction product for V nitrogenase.27 Hydrazinederived intermediates have been studied by EPR and ENDOR for a mutant-form of MoFe protein and were characterized as cofactor-bound NH2 and NH3 structures.28−31 In the alternating mechanism of N2 reduction as proposed by Hoffman, Seefeldt, Dean and co-workers,32,2 a hydrazine-like intermediate is present during the mechanism of N2 reduction. Mutation experiments have also suggested that both hydrazine and acetylene interact with a specific Fe−S face of FeMoco in a study33 where α-70Val was substituted for either a smaller residue (alanine) resulting in enhanced activity or a larger residue (isoleucine) which resulted in lower activity.

The reactive molecule diazene was demonstrated to be a substrate for nitrogenase via in situ formation.34,35 A diazenelike intermediate has been proposed to be present in an alternating mechanism for N2 reduction.32 Substituted diazenes such as methyldiazene and dimethyldiazene have also been demonstrated to be substrates for nitrogenase.28,34 Common reaction intermediates have been found for hydrazine, methyldiazene and diazene35,30,31 and based on EPR and ENDOR data these intermediates likely correspond to cofactor-bound −NH2 (with spin of S = 2) and cofactorbound -NH3 (with spin of S = 1/2) structures. In these reaction intermediates the N−N bond is always cleaved and no earlier intermediate has been characterized, suggesting that the N−N bond is easily cleaved by the cofactor (though substrates do not bind to the resting state, E0). Coucouvanis and co-workers demonstrated that [MoFe3S4] cubanes were active catalysts for hydrazine, acetylene, protons and dimethyldiazene reductions36,37 and by varying the ligands on the complexes it was proposed that Mo was the most likely binding site for all substrates. Interestingly, polycarboxylate ligands on Mo such as citrate were found to result in the most active catalysis for hydrazine reduction, suggesting a possible connection to the Mo-bound homocitrate ligand in the enzyme. In this study, we present density functional theory calculations of reaction mechanisms for reduction of the substrates hydrazine and cis-dimethyldiazene using the [MoFe3S4] cubane catalysts (mechanisms for acetylene and proton reduction will be reported later). We demonstrate that Mo is the likely binding site for both substrates despite Mo not being the metal atom being reduced in the redox steps. At the end, we discuss the implications of our results for the unknown enzyme mechanism for N2 reduction.

■

COMPUTATIONAL DETAILS

The ORCA quantum chemistry code (version 3.0.3)38 was used for all calculations. Broken-symmetry solutions were found by converging first to a high-spin ferromagnetic solution of the cluster (MS = 17/2 for [MoFe3S4]3+ and MS = 16/2 for [MoFe3S4]2+ charge states), then flipping the spin on specific Fe atoms and converging to the brokenB


Article

Inorganic Chemistry

Figure 2. Electronic structure of the [MoFe3S4]3+ (left) and [MoFe3S4]2+ (right) cubanes. Fe1 is here assigned as a ferric ion in the [MoFe3S4]3+ cubane, antiferromagnetically coupled to a delocalized mixed-valence pair (Fe2 and Fe3) with the Mo ion in an unusual non-Hund configuration (implying SMo = 1/2). Upon reduction of the S = 3/2 [MoFe3S4]3+ core to S = 2 [MoFe3S4]2+, the Fe(III) ion Fe1 is reduced to Fe(II). symmetry state with MS = 3/2 (for [MoFe3S4]3+) and MS = 2 (for [MoFe3S4]2+). The BP86 density functional with D3BJ dispersion correction39,40 was used in combination with the triple-ζ def2-TZVP Ahlrichs basis set41 on all atoms. The COSMO-SMD implicit solvation model implemented in ORCA was used to describe acetonitrile solution.42 Additionally, single-point calculations with the hybrid meta-GGA functional TPSSh using the same basis set and solvation model were carried out, with the results documented in the Supporting Information. As has recently been discussed in the nitrogenase literature,43,44 only GGA functionals (like BP86) and TPSSh (a 10% hybrid) give calculated molecular structures for FeMoco (a [MoFe7S9C] double-cubane cluster) that are in reasonable agreement with the high-resolution crystal structure. Hybrids such as B3LYP (20% hybrid) give errors as large as 0.4 Å for Fe−Fe distances and 0.2 Å in Fe−S distances44 while hybrids like M06-2X (54% hybrid) give errors as high as 0.9 Å for Fe−Fe distances. Such strong deviations in molecular structures points to a dramatic failure of most hybrid functionals (≥20% HF exchange) to adequately describe the complex electronic structure in these systems. Previous DFT studies by Szilagyi et al.45,46 on iron−sulfur clusters like [Fe2S2], [Fe4S4] and [MoFe3S4] cores also revealed that hybrid functionals with higher HF exchange percentage are not recommended and suggested a 5% hybrid functional as optimal. In this study we have thus restricted our functional choices to the GGA functional BP86 and the 10% hybrid meta-GGA functional TPSSh even though we note that GGA functionals (and to a lesser extent TPSSh) are known to underestimate reaction barriers.47,48 Chemshell version 3.749,50 using the DL-FIND51 program was used for NEB calculations in order to locate saddlepoints using a modified Chemshell-ORCA interface. Hessian calculations were performed to confirm all minima and saddle points and to calculate free energy corrections. Reduction and protonation steps were performed by computing the redox/protonation step relative to the calculated redox energy of cobaltocene or the deprotonation energy of lutidinium acid, i.e., [M] + [CoCp2] → [M]− + [CoCp2]+ and [M] + [LutH]+ → [MH]+ + [LutH] at the same level of theory (where [M] is the full complex). Cobaltocene and lutidinium acid are the same reduction and protonation agents that were used in the experiments.

ocatecholate) bidentate ligand and acetonitrile ligand on Mo, hereafter labeled as 1 (see Figure 1). This cubane was demonstrated to be an active catalyst for all reactions studied by Coucouvanis and co-workers. As discussed in previous articles12,20 by one of us, the electronic structure of these [MoFe3S4]3+ cubanes is similar to half of FeMoco (that includes a [MoFe3S3C]+ cubane instead) and features a formal metal oxidation state assignment of Mo(III)2Fe(III)1Fe(II) but there is covalency in the cluster, both in the Fe−S bonds (resulting in spin density on the sulfides) and also in Fe−Fe and Mo−Fe interactions. Mössbauer studies (with and without applied magnetic field)13,52 have revealed the average Fe oxidation state assignment of [MoFe3S4]3+ as Fe2.67+ but also reveal that the electronic structure is better described as a delocalized mixed-valence Fe(2.5)−Fe(2.5) pair (with SMV = 9/2) and an antiferromagnetically coupled SFerric = 5/2 Fe(III) ion as shown in Figure 2. The molybdenum ion in the cubanes was assigned as Mo(III) based on the 57Fe Mössbauer data (i.e., via the Fe oxidation states and the total charge) and theoretical calculations20,53,54 revealed it to be best described as as a spin-coupled Mo(III) d3 ion, remarkably in a non-Hund d3 configuration (↑↓↓), apparently due to strong interaction with the 3 Fe ions. The Mo(III) oxidation state was also revealed in FeMoco by a combination of Mo XAS spectroscopy and theoretical calculations. 20 The non-Hund d 3 configuration (↑↓↓) both in synthetic cubanes and FeMoco suggests a local excited spin-state of SMo = 1/2 (a spin-canted SMo = 3/2 ion is also a possibility).20 The spin-state of the [MoFe3S4]3+ cubane is experimentally found to be S = 3/255 and broken-symmetry DFT calculations (that do not give pure spin states unfortunately) find the lowest energy brokensymmetry state to be MS = 3/2 where the spin of the S = 5/2 Fe(III) ion is opposite to the SMV = 9/2 mixed-valence pair.20 The simplest spin-coupling in the cubane is then STot = SMV(9/ 2) − SFerric(5/2) − SMo(1/2) = 3/2. As discussed in a recent review,12 [MoFe3S4]3+ cubanes and FeMoco thus not only have similar molecular structures, they have very similar electronic structures as well, potentially relevant to their reactivity. Upon reduction of [MoFe3S4]3+ cores to [MoFe3S4]2+, the total spin state changes to S = 213. Further, Mössbauer spectroscopy strongly suggests Fe-based reduction.56 Accord-

■

RESULTS AND DISCUSSION A. Molecular and Electronic Structure of [MoFe3S4] Cubanes. Before we discuss substrate reduction reaction mechanisms, it is useful to discuss the electronic structure of [MoFe3S4] cubanes. The system studied is a [MoFe3S4]3+ cubane with chloride ligands on Fe and Cl4cat (tetrachlorC


Article

Inorganic Chemistry

Figure 3. Reaction profile at the BP86 level of theory (free energies in kcal/mol) for hydrazine reduction on the [MoFe3S4] cubane 1 with a closeup of the saddlepoint geometry (H-d) for the rate-determining step.

calculate our redox reactions relative to the calculated redox energy of cobaltocene (at same level of theory) under the same conditions as used in the experiments, i.e., [M] + [CoCp2] → [M]− + [CoCp2]+ (in acetonitrile solution) and this should cancel out some of the systematic errors present in redox potential calculations as discussed by Roy et al.58 and Konezny et al.59 Thus, we expect our calculated redox energies to be more accurate than the comparison of absolute redox potentials suggests. B. Hydrazine Reduction. The reduction of 1 molecule of hydrazine to 2 molecules of ammonia is a 2e−/2H+ reaction. Coucouvanis et al. demonstrated in 199336 the catalytic reduction of hydrazine with [MoFe3S4] cubane clusters. Special care was taken in distinguishing catalytic reductions from disproportionation reactions (occurring an order of magnitude slower). It was demonstrated that a single cubane binding the substrate, was the likely active catalyst as characterized complexes involving hydrazine bridging two [MoFe3S4] cores, showed little ammonia formation. It was suggested that an uncoordinated NH2-group (with a nitrogen lone pair) is necessary for reactivity. Catalysis was convincingly demonstrated with different substrate-catalyst ratios with closeto-full conversion shown in 0.5−2 h and the integrity of the cubane was retained after catalysis. A reaction mechanism was proposed that involved hydrazine binding to Mo (replacing acetonitrile), followed by protonation and subsequent reduction, the release of a first product molecule, NH3 (or possibly NH4+), another reduction and protonation step and release of the second product molecule. Molybdenum was considered the likely substrate-binding site due to a crystal structure showing phenylhydrazine bound to Mo and due to differences in kinetics seen for ligand substitutions on Mo. Later studies by Coucouvanis and co-workers37 showed preference for polycarboxylate ligands such as citrate on Mo for catalytic activity. It was further demonstrated that [Fe4S4Cl4]2− clusters (with [Fe4S4]2+ cores) are completely inactive for hydrazine reduction and that strong binding ligands such as PEt3 and CO inhibit catalysis of [MoFe3S4]

ing to our calculations the electron ends up in the single ferric site, changing the SFerric = 5/2 Fe(III) ion to a SFerrous = 2 Fe(II) ion. The total spin state change to S = 2 upon Fe reduction, can be understood in terms of the S = 9/2 mixedvalence pair now coupling antiferromagnetically to an SFerrous = 2 Fe(II) ion) and the unchanged SMo = 1/2 ion: STot = 9/2 − 2 − 1/2 = 2 (see Figure 2). As discussed later, these 2 redox states of the cubanes are probably the only ones relevant for substrate reduction mechanisms. Three broken-symmetry state spin isomers, however, can be imagined for the ground spin state for each redox state, depending on which Fe ion is flipped (becoming antiferromagnetically coupled to the mixed-valence pair); for cubane 1 in [MoFe3S4]3+, there is only a small energy difference between spin isomers for cubane 1 (0.04 kcal/mol) because of symmetry but this energy difference is larger for different reaction intermediates (but never larger than 1 kcal/ mol). This spin isomerism for the cubane is similar to the broken-symmetry spin isomers discussed for FeMoco in a recent QM/MM study from our group.57 The relative energies of these different spin isomers were checked for different intermediates and can be seen in Table S2. The lowest energy one for each reaction intermediate was chosen and this always corresponded to Fe1 being “spin-down” (see atom labeling in the close-up in Figure 3). Cyclic voltammetry data for this 1-electron process ([MoFe3S4]3+ → [MoFe3S4]2+) exist for a cubane with a trispyrazolyl borate ligand on Mo and chloride ligands on Fe, giving a redox potential of −0.57 V vs SCE.56 At the BP86 level (using the COSMO-SMD solvation model for describing acetonitrile solution) we calculate this potential to be −0.92 V vs SCE, i.e. overestimating by 0.35 V. This potential is slightly closer to experiment at −0.85 V at the hybrid TPSSh level but further away at −1.10 at the hybrid B3LYP level. The BP86 level of theory thus seems to be a reasonable compromise between functional accuracy and computational cost. We note that computing redox potentials to a high accuracy can be very difficult due to errors from approximate solvation models as well as the approximate density functionals. In this study we D


Article

Inorganic Chemistry

Figure 4. Reaction profile at the BP86 level of theory (free energies in kcal/mol) for dimethyldiazene reduction on [MoFe3S4] cubane 1.

the reduction being somewhat more favorable (H-b′) than the protonation (H-b). Following reduction it is downhill to protonate the hydrazine, H-c, (or alternatively reduce after protonation). We also note an alternative scenario: this involves hydrazine being initially protonated to hydrazinium in acetonitrile solution (calculated to be only uphill by +0.7 kcal/mol) and then binding to the cubane via acetonitrile substitution, either in the [MoFe3S4]3+ state (exothermic by −7.0 kcal/mol) or [MoFe3S4]2+ state (exothermic by −12.6 kcal/mol). The uncertainty in the calculation can not dismiss this possibility. We also explored protonation of the oxygen in the catecholate ligand instead of the hydrazine substrate (as in H-c) but this was found to be uphill by 8.3 kcal/mol. After this 1e−/1H+ event it is now favorable to cleave off the first molecule of NH3 from the hydrazine moiety which is exothermic by −25.9 kcal/mol (relative to the starting state) via a saddlepoint that is only 2 kcal/mol higher in energy than the starting point or 13.9 kcal/mol uphill from H-c. Intermediate H-e is a Mo-bound NH2 species that can be protonated to form a Mo-NH3+ species (H-f, this is made possible by partial Mo−N bond breaking) that can then be reduced to form H-g or alternatively, via reduction (H-f′) that is then followed by protonation to form H-g. Substitution of the Mo-bound ammonia molecule for a new molecule of hydrazine releases the second ammonia molecule and reforms the active form of the catalyst H-a. The cubane alternates between redox states [MoFe3S4]3+ and [MoFe3S4]2+ in the reaction profile (Table S1 shows the redox state and spin state of each intermediate). When the H-e or H-f states in the [MoFe3S4]2+ redox state are reduced to form H-f′ or H-g respectively, the NH2/NH3 moiety receives an electron pair, coming from both the added electron and the one electron stored in the cubane (in the form of a Fe(II) ion antiferromagnetically coupled to the mixed-valence pair as shown in Figure 2). The hydrazine reduction mechanism is thus fairly straightforward; it involves Mo as a crucial metal for binding and activating the substrate (Fe will not bind hydrazine here,

cubanes. These results further suggested Mo as the site of catalysis. Other studies on similar [VFe3S4] cubanes from the same group showed similar catalytic activity for hydrazine reduction.60 When the coordination sites on V were blocked by PEt3, catalysis ceased and the possible role of Fe sites was investigated by varying the halide ligands on Fe, showing little change in rate. This again strongly implicates the heterometal as the site of hydrazine binding in these experiments. It was further demonstrated that protonation of the substrate likely precedes reduction as the reduction potential (as determined via electrochemical measurements) of the [VFe3S4] cubane in CH3CN solution (−1.2 V vs Ag/AgCl) is more negative than the corresponding potential of cobaltocene in CH3CN (−0.93 V vs Ag/AgCl) and reacting protonated hydrazine (as a BF4− salt) with the cubane, lowers the reduction wave of the cubane to −0.89 V.60 It should be noted that for the [MoFe3S4] analogue the sequence of reduction and protonation steps is less clear as the reduction potential is approximately −0.9 V vs Ag/AgCl,60 i.e., slightly more positive than cobaltocene. Figure 3 shows the computed reaction profile for the most favorable hydrazine reduction mechanism found for the [MoFe3S4] cubane. The calculations were performed at the BP86 level of theory using the COSMO-SMD solvation model to describe acetonitrile solution. Calculations with another functional TPSSh are presented in the SI that reveal only small changes compared to the BP86 results. Multiple mechanistic possibilities were explored but it became clear early on that substitution of acetonitrile for hydrazine was the most favorable first step, i.e., hydrazine binding to Mo, giving intermediate H-a. This is in agreement with experimental data that shows phenylhydrazine easily replaces acetonitrile as a ligand,36,61 and binds to Mo. We explored hydrazine binding to Fe but this was found to be unfavorable by several kcal/mol (both in the [MoFe3S4]3+ state and the [MoFe3S4]2+ state). After hydrazine binding to Mo we can either reduce the cluster to [MoFe3S4]2+ or protonate the hydrazine moiety directly, both of which is found to be only slightly uphill with E


Article

Inorganic Chemistry

Figure 5. Reaction profile at the BP86 level of theory (free energies in kcal/mol) for hydrazine reduction on [MoFe3S4] cubane vs an artificial Mo− Fe dimer compound.

likely due to the lack of a labile ligand) thus confirming the experimental proposal. As the reduction steps ([MoFe3S4]3+ → [MoFe3S4]2+) involve an Fe-centered redox event, and the electron then transferred to the substrate, the Fe ions are a crucial component of the cluster. Without the presence of the Fe ions, an unfavorable Mo(III)→ Mo(II) reduction would be required instead. There is thus a cooperative advantage to the heterometallic cluster: Mo binds the substrate while the [Fe 3S 4] part performs the redox chemistry with low reorganization energies (as is common for iron−sulfur clusters in biological systems). The partial covalency of the cluster (including partial Mo−Fe bonding interactions20) then likely enables easy electron transfer to the substrate. Replacing the catecholate and acetonitrile ligands on Mo by a tridentate citrate ligand was shown experimentally to result in a [MoFe3S4] cluster that exhibited faster hydrazine reduction.37 This is despite the citrate-ligated cluster requiring a change in coordination to bind hydrazine. Although faster proton-transfer via the carboxylate groups of the citrate is easily imagined as a reason for the faster kinetics (and was suggested37), our reaction mechanism reveals the N−N bond cleavage rather than a protonation step to be the ratedetermining step for hydrazine reduction. It seems likely that the citrate ligand thus offers a way of better stabilizing the transition state via hydrogen bonding, thus enabling faster kinetics and it could additionally participate in proton-shuttling processes as well. The reaction mechanisms involving the cluster with a citrate-ligand will be explored in a future study. C. Dimethyldiazene Reduction. The reduction of dimethyldiazene to 2 equiv of methylamine is a 4e−/4H+ reaction that requires activation of a NN double bond in contrast to the single N−N bond of hydrazine. In their experiments, Coucouvanis and co-workers found that cubane 1 catalyzes the reduction of cis-dimethyldiazene to methylamine. Interestingly, the enzyme catalyzes the same substrate to a mixture of methylamine, ammonia, and methane. Phosphine inhibition studies again suggested Mo as the site of substrate

binding and the product methylamine was found to bind to Mo.62 The calculated reaction mechanism in Figure 4 reveals similarities to the hydrazine mechanism with initial substrate binding (D-a), followed by a reduction step to yield D-b′ (more favorable than protonation) which is then followed by protonation of the substrate to yield D-c (on the distal nitrogen). We also explored protonation of the oxygen in the catecholate ligand instead of the cis-dimethyldiazene substrate (as in D-c) but this was found to be uphill by 6.0 kcal/mol. A second reduction is then followed by a protonation step to the Mo-bound nitrogen to yield intermediate D-e, i.e., Mo-bound cis-dimethylhydrazine. The reaction can now proceed in a similar fashion to the hydrazine pathway with a third round of reduction and protonation (to distal nitrogen) that gives intermediate D-g. From this intermediate it is now favorable to cleave the N−N bond (analogously to the hydrazine pathway in Figure 3) to give the first methylamine product molecule. This proceeds via saddlepoint D-h giving favorable intermediate D-i, a Mo-bound MeNH moiety. A final round of reduction and protonation can then yield Mo-bound methylamine which could be replaced by the next molecule of substrate to complete the catalytic cycle. The formation of Mo-bound cis-dimethylhydrazine along the pathway, D-e, (“symmetric” substrate reduction), is consistent with the experiments performed using cis-dimethylhydrazine as a substrate to yield methylamine. The cubane system was found by Coucouvanis and co-workers to not be capable of reducing cis-dimethyldiazene to methane and ammonia, unlike the enzyme. As discussed in their study, this suggests either a different mechanism for the enzyme (concomitant reduction of the NN and C−N bonds) or (if methylamine is first formed) some ability of the enzyme to weaken the C−N bond without protonation (since no lonepair is available when methylamine is bound to a metal ion). It is unclear at this stage, how this could take place. F


Article

Inorganic Chemistry F. Why the Cubanes Are Good Catalysts. Having now established the reaction mechanisms for hydrazine and cisdimethyldiazene substrate reductions with these cubane systems it is interesting to gain a better understanding of why they work as well as they do as catalysts. The key components appear to be (i) a Mo ion with available coordination site for substrate binding that usually remains in the same oxidation state, (ii) metal ions (Fe) that are favorably reduced and can store an electron until needed, (iii) a flexible ligand environment that allows for displacing a weakly bound ligand (e.g., acetonitrile). As the reduction events were always localized on the Fe(III) ion, we were curious whether a simpler Mo−Fe dimer system could be constructed that would be capable of similar chemistry as the [MoFe3S4] cubane. The catalytic reduction of hydrazine was chosen as an example. Complex 1 was thus simplified by deleting two of the Fe atoms, keeping two terminal SH groups on the Fe ion and one terminal SH group on Mo. Two bridging sulfide groups were kept between Mo and Fe. The final system was calculated as [Mo(III)Fe(III)S2(SH)3Cl4catAcn]3−. The reaction profile for hydrazine reduction was calculated analogously to complex 1 and is shown in Figure 5 alongside the reaction profile for 1. The reaction and protonation steps ended up being similar in the dimer and cubane. However, the energies of the most important steps are completely different. After initial binding of hydrazine to the Mo−Fe dimer complex (with similar energies), the hydrazine protonation is found to be rather more favorable at the dimer than the cubane, whereas the reduction, either before or after protonation, is revealed to be very unfavorable. The protonated reaction intermediate at −17.3 kcal/mol is thus so favorable and the reduction steps so unfavorable that the saddlepoint for N−N bond-breaking of hydrazine is ∼31 kcal/mol above the first protonated intermediate. Thus, this would result in a thermodynamic sink and the Mo−Fe dimer would not be a catalyst for hydrazine reduction at room temperature. The primary reason appears to be the unfavorable reduction events on the Mo−Fe dimer. The cubane appears to be a good catalyst compared to the dimer as the redox potential of the Fe(III)→ Fe(II) reduction in the cubane is shifted toward more favorable values, possibly via partial delocalization of the added electron, something not possible in the dimer.

vanadium nitrogenase revealed the presence of a light atom ligand in a proposed reaction intermediate (likely NH or OH)64 between the analogous Fe ions in FeVco (calculations from our group suggest OH to be more likely than NH for this ligand65). These results suggest that the bridging sulfides are labile, which may open up a binding site for substrates, something not possible for the Fe sites in the [MoFe3S4] cubanes. The calculated mechanisms revealed in this study demonstrate, however, how a similar metal−sulfur cluster to FeMoco (though smaller) can participate in 2- and 4-electron redox catalysis of substrates common to both catalysts. Even though the substrate binding site at FeMoco may turn out to be different (Fe is generally considered the more likely scenario), it would be surprising if some of the mechanistic information revealed here for [MoFe3S4] cubanes, would turn out to be very different for FeMoco. We expect for example that cubane and FeMoco may share the feature of having reduction events and substrate binding at different sites. As an example, our calculated mechanism for the hydrazine reaction revealed that the cubane require only a single-electron reduction of the cluster for cleavage of the N−N bond, interestingly not at the metal ion binding the substrate. It seems likely that this aspect is shared by FeMoco (that features even more metal ions). FeMoco is known experimentally to not bind any substrate in its resting state (E0) and hydrazine is usually assumed to bind to either the 1-electron reduced (E1) or the 2-electron reduced state (E2). Based on the mechanism proposed here for a related cubane it seems likely that only a single e−/H+ event is required for cleavage of the N−N bond. If hydrazine is able to bind to the E1 state of FeMoco, it seems reasonable to assume that the N−N bond can then be cleaved at the same redox level (like the [MoFe3S4] cubane). This then offers an explanation for why an intact hydrazine intermediate has not been trapped and characterized but rather hydrazine-derived −NH2 and −NH3 intermediates.30,31 Finally, based on the interesting Mo-based chemistry of the [MoFe3S4] cubanes , Coucouvanis hypothesized that while the N2 substrate might bind at an Fe site, the reduced N2-form (perhaps at the diazene or hydrazine level) might migrate over to the molybdenum site where the substrate is reduced to ammonia.21,23 As little is still known about the mechanistic details of N2 reduction on FeMoco, this possibility has not been ruled out and may even explain the known importance of Mo and the Mo-bound homocitrate for enzymatic nitrogen reduction.

■

CONCLUSIONS This study has uncovered mechanistic details on the interesting reactivity and catalysis that was found for the molybdenum− iron−sulfur model cubanes in the lab of Coucouvanis in the nineties. Because of the strong similarity of the molecular and electronic structure of the cubanes compared to FeMoco and the similarity in reducing substrates it would be surprising if some of this chemistry is not related to the chemistry of the enzyme. We note, however, that this is not evidence of FeMoco binding substrates at the molybdenum site (trial calculations unsurprisingly reveal that the [MoFe3S4] core can not sufficiently activate N2) though Mo should not be excluded as a possible binding site. Also of note is the double-cubane structure of FeMoco, and the presence of the interstitial carbide and the bridging sulfides (not present in the cubane) will certainly result in different reactivity of FeMoco compared to [MoFe3S4] cubanes. In fact recent crystal structures have demonstrated CO binding to the Fe2 and Fe6 atoms (replacing sulfide) of FeMoco63 and a recent crystal structure of

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.inorgchem.8b02669. Additional reaction profile for dimethyldiazene reduction mechanism showing a distal mechanism; calculated reaction profiles with the TPSSh functional; table with charge, spin state, and spin populations for all intermediates; energy differences and spin populations for different broken-symmetry solutions of some intermediates; Cartesian coordinates for all optimized structures (PDF) (ZIP) G


Article

Inorganic Chemistry

■

(16) Coucouvanis, D. Fe-M-S complexes derived from MS42- anions (M = molybdenum, tungsten) and their possible relevance as analogs for structural features in the molybdenum site of nitrogenase. Acc. Chem. Res. 1981, 14, 201−209. (17) Coucouvanis, D. Use of preassembled iron/sulfur and iron/ molybdenum/sulfur clusters in the stepwise synthesis of potential analogs for the Fe/Mo/S site in nitrogenase. Acc. Chem. Res. 1991, 24, 1−8. (18) Malinak, S. M.; Coucouvanis, D. The Chemistry of Synthetic Fe–Mo–S Clusters and their Relevance to the Structure and Function of the Fe–Mo–S Center in Nitrogenase. Prog. Inorg. Chem. 2007, 49, 599−662. (19) Rees, J. A.; Bjornsson, R.; Kowalska, J. K.; Lima, F. A.; Schlesier, J.; Sippel, D.; Weyhermüller, T.; Einsle, O.; Kovacs, J. A.; DeBeer, S. Comparative electronic structures of nitrogenase FeMoco and FeVco. Dalton. Trans. 2017, 46, 2445−2455. (20) Bjornsson, R.; Lima, F. A.; Spatzal, T.; Weyhermueller, T.; Glatzel, P.; Einsle, O.; Neese, F.; DeBeer, S. Identification of a spincoupled Mo (III) in the Nitrogenase Iron-Molybdenum Cofactor. Chem. Sci. 2014, 5, 3096−3103. (21) Coucouvanis, D.; Demadis, K. D.; Malinak, S. M.; Mosier, P. E.; Tyson, M. A.; Laughlin, L. J. Catalytic Multielectron Reduction of Hydrazine to Ammonia and Acetylene to Ethylene with Clusters That Contain the MFe3S4 Cores (M= Mo, V): Relevance to the Function of Nitrogenase. ACS. Symposium. Series. 1996, 653, 117−134. (22) Coucouvanis, D. Functional analogs for the reduction of certain nitrogenase substrates. Are multiple sites within the Fe/Mo/S active center involved in the 6e–reduction of N2? J. Biol. Inorg. Chem. 1996, 1, 594−600. (23) Coucouvanis, D.; Demadis, K. D.; Malinak, S. M.; Mosier, P. E.; Tyson, M. A.; Laughlin, L. J. Catalytic and stoichiometric multielectron reduction of hydrazine to ammonia and acetylene to ethylene with clusters that contain the MFe3S4 cores (M Mo, V). Relevance to the function of nitrogenase. J. Mol. Catal. A: Chem. 1996, 107, 123− 135. (24) Davis, L. C. Hydrazine as a substrate and inhibitor of Azotobacter vinelandii nitrogenase. Arch. Biochem. Biophys. 1980, 204, 270−6. (25) Danyal, K.; Inglet, B. S.; Vincent, K. A.; Barney, B. M.; Hoffman, B. M.; Armstrong, F. A.; Dean, D. R.; Seefeldt, L. C. Uncoupling nitrogenase: catalytic reduction of hydrazine to ammonia by a MoFe protein in the absence of Fe protein-ATP. J. Am. Chem. Soc. 2010, 132, 13197−9. (26) Thorneley, R. N. F.; Lowe, D. J. In Molybdenum Enzymes; Spiro, T. G., Ed.; Wiley-Interscience: New York, 1985; p 221. (27) Dilworth, M. J.; Eady, R. R. Hydrazine is a product of dinitrogen reduction by the vanadium-nitrogenase from Azotobacter chroococcum. Biochem. J. 1991, 277, 465−468. (28) Barney, B. M.; Yang, T. C.; Igarashi, R. Y.; Dos Santos, P. C.; Laryukhin, M.; Lee, H. I.; Hoffman, B. M.; Dean, D. R.; Seefeldt, L. C. Intermediates trapped during nitrogenase reduction of N triple bond N, CH3-N = NH, and H2N-NH2. J. Am. Chem. Soc. 2005, 127, 14960−1. (29) Barney, B. M.; Laryukhin, M.; Igarashi, R. Y.; Lee, H. I.; Dos Santos, P. C.; Yang, T. C.; Hoffman, B. M.; Dean, D. R.; Seefeldt, L. C. Trapping a hydrazine reduction intermediate on the nitrogenase active site. Biochemistry 2005, 44, 8030−7. (30) Lukoyanov, D.; Dikanov, S. A.; Yang, Z. Y.; Barney, B. M.; Samoilova, R. I.; Narasimhulu, K. V.; Dean, D. R.; Seefeldt, L. C.; Hoffman, B. M. ENDOR/HYSCORE studies of the common intermediate trapped during nitrogenase reduction of N2H2, CH3N2H, and N2H4 support an alternating reaction pathway for N2 reduction. J. Am. Chem. Soc. 2011, 133, 11655−64. (31) Lukoyanov, D.; Yang, Z. Y.; Barney, B. M.; Dean, D. R.; Seefeldt, L. C.; Hoffman, B. M. Unification of reaction pathway and kinetic scheme for N2 reduction catalyzed by nitrogenase. Proc. Natl. Acad. Sci. U. S. A. 2012, 109, 5583−7. (32) Hoffman, B. M.; Lukoyanov, D.; Dean, D. R.; Seefeldt, L. C. Nitrogenase: A Draft Mechanism. Acc. Chem. Res. 2013, 46, 587−595.

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Ragnar Bjornsson: 0000-0003-2167-8374 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS R.B. acknowledges support from the Icelandic Research Fund, Grants 141218051 and 162880051, and the University of Iceland Research Fund.

■

REFERENCES

(1) Burgess, B. K.; Lowe, D. J. Mechanism of Molybdenum Nitrogenase. Chem. Rev. 1996, 96, 2983−3012. (2) Hoffman, B. M.; Lukoyanov, D.; Yang, Z. Y.; Dean, D. R.; Seefeldt, L. C. Mechanism of Nitrogen Fixation by Nitrogenase: The Next Stage. Chem. Rev. 2014, 114, 4041−4062. (3) Kim, J.; Rees, D. C. Structural models for the metal centers in the nitrogenase molybdenum-iron protein. Science 1992, 257, 1677− 1682. (4) Einsle, O.; Tezcan, F. A.; Andrade, S. L. A.; Schmid, B.; Yoshida, M.; Howard, J. B.; Rees, D. C. Nitrogenase MoFe-Protein at 1.16 Å Resolution: A Central Ligand in the FeMo-Cofactor. Science 2002, 297, 1696−1700. (5) Spatzal, T.; Aksoyoglu, M.; Zhang, L.; Andrade, S. L. A.; Schleicher, E.; Weber, S.; Rees, D. C.; Einsle, O. Evidence for Interstitial Carbon in Nitrogenase FeMo Cofactor. Science 2011, 334, 940. (6) Lancaster, K. M.; Roemelt, M.; Ettenhuber, P.; Hu, Y.; Ribbe, M. W.; Neese, F.; Bergmann, U.; DeBeer, S. X-ray Emission Spectroscopy Evidences a Central Carbon in the Nitrogenase Iron-Molybdenum Cofactor. Science 2011, 334, 974−977. (7) Sippel, D.; Einsle, O. The structure of vanadium nitrogenase reveals an unusual bridging ligand. Nat. Chem. Biol. 2017, 13, 956− 960. (8) Rees, J. A.; Bjornsson, R.; Schlesier, J.; Sippel, D.; Einsle, O.; DeBeer, S. The Fe-V Cofactor of Vanadium Nitrogenase Contains an Interstitial Carbon Atom. Angew. Chem., Int. Ed. 2015, 54, 13249−52. (9) Harris, D. F.; Lukoyanov, D. A.; Shaw, S.; Compton, P.; Tokmina-Lukaszewska, M.; Bothner, B.; Kelleher, N.; Dean, D. R.; Hoffman, B. M.; Seefeldt, L. C. Mechanism of N2 Reduction Catalyzed by Fe-Nitrogenase Involves Reductive Elimination of H2. Biochemistry 2018, 57, 701−710. (10) Seefeldt, L. C.; Dance, I. G.; Dean, D. R. Substrate interactions with nitrogenase: Fe versus Mo. Biochemistry 2004, 43, 1401−1409. (11) Igarashi, R. Y.; Laryukhin, M.; Dos Santos, P. C.; Lee, H. I.; Dean, D. R.; Seefeldt, L. C.; Hoffman, B. M. Trapping H- bound to the nitrogenase FeMo-cofactor active site during H2 evolution: characterization by ENDOR spectroscopy. J. Am. Chem. Soc. 2005, 127, 6231−41. (12) Bjornsson, R.; Neese, F.; Schrock, R. R.; Einsle, O.; DeBeer, S. The discovery of Mo(III) in FeMoco: reuniting enzyme and model chemistry. J. Biol. Inorg. Chem. 2015, 20, 447−60. (13) Lee, S. C.; Holm, R. H. The clusters of nitrogenase: synthetic methodology in the construction of weak-field clusters. Chem. Rev. 2004, 104, 1135−1158. (14) Wolff, T. E.; Berg, J. M.; Warrick, C.; Hodgson, K. O.; Holm, R. H.; Frankel, R. B. The molybdenum-iron-sulfur cluster complex [Mo2Fe6S9 (SC2H5) 8] 3-. A synthetic approach to the molybdenum site in nitrogenase. J. Am. Chem. Soc. 1978, 100, 4630−4632. (15) Lee, S. C.; Lo, W.; Holm, R. H. Developments in the biomimetic chemistry of cubane-type and higher nuclearity iron-sulfur clusters. Chem. Rev. 2014, 114, 3579−600. H


Article

Inorganic Chemistry (33) Barney, B. M.; Igarashi, R. Y.; Dos Santos, P. C.; Dean, D. R.; Seefeldt, L. C. Substrate interaction at an iron-sulfur face of the FeMo-cofactor during nitrogenase catalysis. J. Biol. Chem. 2004, 279, 53621−4. (34) McKenna, C. E.; Simeonov, A. M.; Eran, H.; BravoLeerabhandh, M. Reduction of cyclic and acyclic diazene derivates by Azotobacter vinelandii nitrogenase: diazirine and trans-dimethyldiazene. Biochemistry 1996, 35, 4502−14. (35) Barney, B. M.; McClead, J.; Lukoyanov, D.; Laryukhin, M.; Yang, T. C.; Dean, D. R.; Hoffman, B. M.; Seefeldt, L. C. Diazene (HN = NH) is a substrate for nitrogenase: insights into the pathway of N2 reduction. Biochemistry 2007, 46, 6784−94. (36) Coucouvanis, D.; Mosier, P. E.; Demadis, K. D.; Patton, S.; Malinak, S. M.; Kim, C. G.; Tyson, M. A. The catalytic reduction of hydrazine to ammonia by the MoFe3S4 cubanes and implications regarding the function of nitrogenase. Evidence for direct involvement of the molybdenum atom in substrate reduction. J. Am. Chem. Soc. 1993, 115, 12193−12194. (37) Demadis, K. D.; Malinak, S. M.; Coucouvanis, D. Catalytic Reduction of Hydrazine to Ammonia with MoFe3S4−Polycarboxylate Clusters. Possible Relevance Regarding the Function of the Molybdenum-Coordinated Homocitrate in Nitrogenase. Inorg. Chem. 1996, 35, 4038−4046. (38) Neese, F. The ORCA program system. WIREs. Comput. Mol. Sci. 2012, 2, 73−78. (39) Grimme, S.; Antony, J.; Ehrlich, S.; Krieg, H. A consistent and accurate ab initio parametrization of density functional dispersion correction (DFT-D) for the 94 elements H-Pu. J. Chem. Phys. 2010, 132, 154104. (40) Grimme, S.; Ehrlich, S.; Goerigk, L. Effect of the damping function in dispersion corrected density functional theory. J. Comput. Chem. 2011, 32, 1456−1465. (41) Weigend, F.; Ahlrichs, R. Balanced basis sets of split valence, triple zeta valence and quadruple zeta valence quality for H to Rn: Design and assessment of accuracy. Phys. Chem. Chem. Phys. 2005, 7, 3297−3305. (42) Klamt, A.; Schüürmann, G. COSMO: a new approach to dielectric screening in solvents with explicit expressions for the screening energy and its gradient. J. Chem. Soc., Perkin Trans. 2 1993, 2, 799−805. (43) Cao, L.; Caldararu, O.; Ryde, U. Protonation and Reduction of the FeMo Cluster in Nitrogenase Studied by Quantum Mechanics/ Molecular Mechanics (QM/MM) Calculations. J. Chem. Theory Comput. 2018, 14, 6653. (44) Raugei, S.; Seefeldt, L. C.; Hoffman, B. M. Critical computational analysis illuminates the reductive-elimination mechanism that activates nitrogenase for N2 reduction. Proc. Natl. Acad. Sci. U. S. A. 2018, 115, E10521. (45) Szilagyi, R. K.; Winslow, M. A. On the accuracy of density functional theory for iron-sulfur clusters. J. Comput. Chem. 2006, 27, 1385−1397. (46) Harris, T. V.; Szilagyi, R. K. Iron−sulfur bond covalency from electronic structure calculations for classical iron−sulfur clusters. J. Comput. Chem. 2014, 35, 540−552. (47) Patchkovskii, S.; Ziegler, T. Improving ″difficult″ reaction barriers with self-interaction corrected density functional theory. J. Chem. Phys. 2002, 116, 7806. (48) Grüning, M.; Gritsenko, O. V.; Baerends, E. J. Improved Description of Chemical Barriers with Generalized Gradient Approximations (GGAs) and Meta-GGAs. J. Phys. Chem. A 2004, 108, 4459−4469. (49) Sherwood, P.; de Vries, A. H.; Guest, M. F.; Schreckenbach, G.; Catlow, C. R. A.; French, S. A.; Sokol, A. A.; Bromley, S. T.; Thiel, W.; Turner, A. J.; Billeter, S.; Terstegen, F.; Thiel, S.; Kendrick, J.; Rogers, S. C.; Casci, J.; Watson, M.; King, F.; Karlsen, E.; Sjøvoll, M.; Fahmi, A.; Schäfer, A.; Lennartz, C. QUASI: A general purpose implementation of the QM/MM approach and its application to problems in catalysis. J. Mol. Struct.: THEOCHEM 2003, 632, 1−28.

(50) Metz, S.; Kästner, J.; Sokol, A. A.; Keal, T. W.; Sherwood, P. ChemShell modular software package for QM/MM simulations. WIREs. Comput. Mol. Sci. 2014, 4, 101−110. (51) Kästner, J.; Carr, J. M.; Keal, T. W.; Thiel, W.; Wander, A.; Sherwood, P. DL-FIND: an open-source geometry optimizer for atomistic simulations. J. Phys. Chem. A 2009, 113, 11856−11865. (52) Mascharak, P. K.; Papaefthymiou, G. C.; Armstrong, W. H.; Foner, S.; Frankel, R. B.; Holm, R. H. Electronic properties of singleand double-MoFe3S4 cubane-type clusters. Inorg. Chem. 1983, 22, 2851−2858. (53) Cook, M.; Karplus, M. Electronic structure of the molybdenum-iron-sulfur cluster MoFe3S4 (SH) 63-ion. J. Am. Chem. Soc. 1985, 107, 257−259. (54) Cook, M.; Karplus, M. Electronic structure of the MoFe3S4(SH)3−6 ion: A broken-symmetry metal−sulfur cluster. J. Chem. Phys. 1985, 83, 6344. (55) Armstrong, W. H.; Mascharak, P. K.; Holm, R. H. Demonstration of the existence of single cubane-type molybdenum iron sulfide (MoFe3S4) clusters with S= 3/2 ground states: preparation, structure, and properties. Inorg. Chem. 1982, 21, 1699− 1701. (56) Fomitchev, D. V.; McLauchlan, C. C.; Holm, R. H. Heterometal Cubane-Type MFe 3S 4Clusters (M = Mo, V) Trigonally Symmetrized with Hydrotris(pyrazolyl)borate(1−) and Tris(pyrazolyl)methanesulfonate(1−) Capping Ligands. Inorg. Chem. 2002, 41, 958−966. (57) Benediktsson, B.; Bjornsson, R. QM/MM Study of the Nitrogenase MoFe Protein Resting State: Broken-Symmetry States, Protonation States, and QM Region Convergence in the FeMoco Active Site. Inorg. Chem. 2017, 56, 13417−13429. (58) Roy, L. E.; Jakubikova, E.; Guthrie, M. G.; Batista, E. R. Calculation of One-Electron Redox Potentials Revisited. Is It Possible to Calculate Accurate Potentials with Density Functional Methods? J. Phys. Chem. A 2009, 113, 6745−6750. (59) Konezny, S. J.; Doherty, M. D.; Luca, O. R.; Crabtree, R. H.; Soloveichik, G. L.; Batista, V. S. Reduction of Systematic Uncertainty in DFT Redox Potentials of Transition-Metal Complexes. J. Phys. Chem. C 2012, 116, 6349−6356. (60) Malinak, S. M.; Demadis, K. D.; Coucouvanis, D. Catalytic Reduction of Hydrazine to Ammonia by the VFe3S4 Cubanes. Further Evidence for the Direct Involvement of the Heterometal in the Reduction of Nitrogenase Substrates and Possible Relevance to the Vanadium Nitrogenases. J. Am. Chem. Soc. 1995, 117, 3126−3133. (61) Palermo, R. E.; Singh, R.; Bashkin, J. K.; Holm, R. H. Molybdenum atom ligand substitution reactions of molybdenumiron-sulfur (MoFe3S4) cubane-type clusters: synthesis and structures of clusters containing molybdenum-bound pseudosubstrates of nitrogenase. J. Am. Chem. Soc. 1984, 106, 2600−2612. (62) Malinak, S. M.; Simeonov, A. M.; Mosier, P. E.; McKenna, C. E.; Coucouvanis, D. Catalytic Reduction of cis-Dimethyldiazene by the [MoFe3S4]3+ Clusters. The Four-Electron Reduction of a NN Bond by a Nitrogenase-Relevant Cluster and Implications for the Function of Nitrogenase. J. Am. Chem. Soc. 1997, 119, 1662−1667. (63) Spatzal, T.; Perez, K. A.; Einsle, O.; Howard, J. B.; Rees, D. C. Ligand binding to the FeMo-cofactor: Structures of CO-bound and reactivated nitrogenase. Science 2014, 345, 1620−1623. (64) Sippel, D.; Rohde, M.; Netzer, J.; Trncik, C.; Gies, J.; Grunau, K.; Djurdjevic, I.; Decamps, L.; Andrade, S. L. A.; Einsle, O. A bound reaction intermediate sheds light on the mechanism of nitrogenase. Science 2018, 359, 1484−1489. (65) Benediktsson, B.; Thorhallsson, A. Th.; Bjornsson, R. QM/MM calculations reveal a bridging hydroxo group in a vanadium nitrogenase crystal structure. Chem. Commun. 2018, 54, 7310−7313.

I


Computational Mechanistic Study of [MoFe3S4] Cubanes for Catalytic

Recommend Documents