Interplay between Theory and Experiment for Ammonia Synthesis

Apr 22, 2016 - Institute for Materials Chemistry and Engineering, Kyushu University, ... He received the Chemical Society of Japan Award for Distingui...
0 downloads 0 Views 2MB Size
Article pubs.acs.org/accounts

Interplay between Theory and Experiment for Ammonia Synthesis Catalyzed by Transition Metal Complexes Published as part of the Accounts of Chemical Research special issue “Computational Catalysis for Organic Synthesis”. Hiromasa Tanaka,† Yoshiaki Nishibayashi,*,‡ and Kazunari Yoshizawa*,† †

Institute for Materials Chemistry and Engineering, Kyushu University, Nishi-ku, Fukuoka 819-0395, Japan Department of Systems Innovation, School of Engineering, The University of Tokyo, Yayoi, Bunkyo-ku, Tokyo 113-8656, Japan



CONSPECTUS: Nitrogen fixation is an essential chemical process both biologically and industrially. Since the discovery of the first transition-metal−dinitrogen complex in 1965, a great deal of effort has been devoted to the development of artificial nitrogen fixation systems that work under mild reaction conditions. However, the transformation of chemically inert dinitrogen using homogeneous catalysts is still challenging because of the difficulty in breaking the strong triple bond of dinitrogen, and a very limited number of transition metal complexes have exhibited the catalytic activity for the direct transformation of dinitrogen into ammonia with low turnover numbers. To develop more effective nitrogen fixation systems, it is necessary to retrieve as much information as possible from the limited successful examples. Computational chemistry will provide valuable insights in the understanding of the reaction mechanisms involving unstable intermediates that are hard to isolate or characterize. We have been applying it for clarifying detailed mechanisms of dinitrogen activation and functionalization by transition metal complexes as well as for designing new catalysts for more effective nitrogen fixation. This Account summarizes recent progress in the elucidation of catalytic mechanisms of nitrogen fixation by using mono- and dinuclear molybdenum complexes, as well as cubane-type metal−sulfido clusters from a theoretical point of view. First, we briefly introduce experimental and theoretical contributions to the elucidation of the reaction mechanism of nitrogen fixation catalyzed by a mononuclear Mo−triamidoamine complex. Special attention is paid to our recent studies on Mo-catalyzed nitrogen fixation using dinitrogen-bridged dimolybdenum complexes. A possible catalytic mechanism is proposed based on theoretical and experimental investigations. The catalytic mechanism involves the formation of a monuclear molybdenum−nitride (MoN) intermediate, as well as the regeneration of a dimolybdenum intermediate with the Mo−NN−Mo moiety. Comparison of the reactivity of di- and monomolybdenum complexes suggests that the dimolybdenum structure is essential for the catalytic activity. Synergy between the two Mo cores connected with a bridging N2 ligand is observed in the protonation of coordinated N2. Intermetallic electron transfer through the bridging N2 ligand reductively activates the coordinated N2 to be protonated. On the basis of the proposed catalytic mechanism, we used DFT calculations for rational design of dimolybdenum complexes serving as more effective catalysts for nitrogen fixation. Newly prepared dimolybdenum complexes with modified PNP-type pincer ligands exhibit greater catalytic activity than the original one.



INTRODUCTION

N triple bond of dinitrogen to synthesize ammonia under ambient reaction conditions with a homogeneous catalyst of high turnover number. In contrast, biological nitrogen fixation as a part of the global nitrogen cycle is attained by the enzyme nitrogenase in certain bacteria at ambient temperature and pressure. Experimental studies of the Mo-containing nitrogenase revealed that its active site to reduce dinitrogen involves a double cubane-type iron− molybdenum cluster MoFe7S9C (FeMo cofactor; Scheme 1).2,3 Using eight pairs of proton and electron, FeMo cofactor

Nitrogen fixation, the reduction of atmospheric dinitrogen to ammonia, is one of the most important catalytic reactions in biology and chemistry. Since dinitrogen is chemically inert due to its extremely strong triple bond, both artificial and biological nitrogen fixation can proceed in the presence of metal-based catalysts. The Haber−Bosch process transforms N2 and H2 gases into ammonia on the surface of iron-based heterogeneous catalysts under harsh reaction conditions, where a huge amount of energy is consumed for the H2 production from fossil fuels such as natural gas.1 The development of an alternative to the Haber−Bosch process without the use of dihydrogen is still challenging because of the difficulty in breaking the strong N © XXXX American Chemical Society

Received: January 19, 2016

A

DOI: 10.1021/acs.accounts.6b00033 Acc. Chem. Res. XXXX, XXX, XXX−XXX

Article

Accounts of Chemical Research

Scheme 3. Nitrogen fixation Catalyzed by a DinitrogenBridged Dimolybdenum Complex

Scheme 1. Biological Nitrogen Fixation by the MoContaining Nitrogenase

converts dinitrogen to ammonia with the mandatory evolution of dihydrogen. Although the mechanism of biological nitrogen fixation has not completely been understood, a possible reaction pathway is proposed for N2 transformation starting from a “Janus” E4 intermediate, in which four protons and four electrons are accumulated in the FeMo cofactor and two of the hydrogens form [Fe−H−Fe] moieties.3 Since the discovery of [Ru(N2)(NH3)5]2+ in 1965,4 a great number of well-defined transition-metal dinitrogen complexes have been prepared and the reactivity of coordinated N2 ligand has been extensively investigated in a quest to discover a novel nitrogen fixation system that works under mild reaction conditions.5 However, there are only a few examples of the direct transformation of dinitrogen into ammonia catalyzed by transition-metal−dinitrogen complexes in the presence of proton and electron sources.6−10 In 2003 Yandulov and Schrock reported the first catalytic transformation of dinitrogen into ammonia under ambient conditions using Mo−triamidoamine complex 1 as a catalyst, where 8 equiv of ammonia was produced based on the catalyst, as shown in Scheme 2.8

system,24,25 which is a result of our synergetic relationship between theory and experiment in the chemistry of dinitrogen activation and transformation.26 More recently Peters and co-workers reported the first Fecatalyzed direct transformation of dinitrogen into ammonia at −78 °C.10 In this system, an anionic iron−dinitrogen complex bearing tris(phosphine)borane produced 7 equiv of ammonia based on the catalyst. While a cationic Fe−NNH2 intermediate is spectroscopically characterized,27 theoretical approaches will provide a better insight into the Fe-catalyzed system, relevant to the biological nitrogen fixation. Experimental and theoretical works on an alternative way to transform dinitrogen via photochemical activation of dinitrogen in dinitrogen-bridged bimetallic complexes should be described here.28−33 Floriani and co-workers reported the first photoinduced splitting of the bridging dinitrogen ligand in [Mo(Mes)3]2(μ-N2).28 Kirchner and co-workers theoretically proposed a photochemical activation scheme of a bridging dinitrogen ligand in bimetallic complexes using Sellmann-type FeII and RuII complexes as models.29 The π−π* excitation of the bridging dinitrogen ligand weakens its triple bond to yield a bent, diazene-like structure of the M−N−N−M moiety, which is prepared for the formation of diazene. Kunkely and Vogler reported that photoirradiation of a dinitrogen-bridged OsII− OsIII complex induces the reductive splitting of the bridging dinitrogen through MLCT excitation.30 Blank and co-workers performed pump−probe spectroscopy to investigate the photoexcitation dynamics of the dintrogen splitting in the photolysis of [Mo(N[t-Bu]Ar)](μ-N2).31 Furche, Evans, and co-workers have demonstrated photochemical activation of dinitrogen using rare-earth (M = Y, Lu, Dy) complexes bearing cyclopentadienyl ligands.32 DFT calculations were applied for rationalizing a low-energy charge transfer from one of the cyclopentadienyl ligands to the metal center, which leads to the binding and activation of dinitrogen by two metal units. More recently, we have found the cleavage and formation of dinitrogen in a single system assisted by molybdenum complexes bearing ferrocenyldiphosphine.33 In this system, a dimolybdenum μ-N2 complex was split into two molybdenum− nitride complexes by photoirradiation, and the dimolybdenum complex was regenerated from the nitride complexes by oxidation.

Scheme 2. Nitrogen Fixation Catalyzed by a Mo− Triamidoamine Complex

Schrock and co-workers proposed a catalytic mechanism (the Yandulov−Schrock cycle) based on the isolation and observation of reactive intermediates.11,12 The validity of the proposed mechanism is supported by intensive theoretical studies on the energetics of the entire catalytic cycle.13−23 Nishibayashi and co-workers found the second example of Mo-catalyzed nitrogen fixation using a dinitrogen-bridged dimolybdenum complex bearing a PNP-type pincer ligand (2) under ambient reaction conditions, where 23 equiv of ammonia was produced based on the catalyst (Scheme 3).9 Detailed descriptions in this Account are devoted to the recent work for understanding this Mo-catalyzed nitrogen fixation



NITROGEN FIXATION CATALYZED BY Mo−TRIAMIDOAMINE COMPLEXES Yandulov and Schrock reported Mo-catalyzed nitrogen fixation using [HIPTN3N]Mo(N2) 1 under ambient reaction conB

DOI: 10.1021/acs.accounts.6b00033 Acc. Chem. Res. XXXX, XXX, XXX−XXX

Article

Accounts of Chemical Research

was experimentally confirmed by Schrock and co-workers.12 Tuczek and co-workers have recently conducted a large-scale calculation using the full [HIPTN 3 N]Mo catalyst for quantitative comparison with the available experimental data.20 The use of the full Mo catalyst resulted in almost the same energetics as obtained earlier for the small model,14 and the authors described that a larger model does not necessarily lead to an improvement in the calculated energetics. Leitner and Hölscher discussed the potential of transition metal complexes (M = Mo, Ru, Os) bearing a HIPT-substituted triamidoamine ligand as catalysts for reducing dinitrogen using dihydrogen as the reducing reagent.22 They have also been pursuing the possibility of the transformation of dinitrogen into ammonia by dihydrogen catalyzed by transition metal complexes bearing various pincer ligands.34 Quite recently, Reiher and co-workers developed a heuristics-guided protocol for the automatic exploration of chemical reaction spaces and applied it for finding possible elementary reactions in the Yandulov−Schrock cycle.23 They optimized about 10000 intermediates and more than 2000 transition states using a small model of 1. The automated screening of chemical reaction spaces has been growing in importance for a full understanding of the catalytic mechanism as well as deactivation mechanisms determining turnover number of the catalyst.

ditions in the presence of 2,6-lutidinium tetrakis[3,5-bis(trifluoromethyl)phenylborate] ([LutH]BArF4) as a proton source and decamethylchromocene (CrCp*2) as a reducing reagent (Scheme 2).8 They isolated and characterized reactive intermediates in the catalytic reaction to propose a plausible catalytic cycle presented in Figure 1.8,11,12 In this cycle, the

Figure 1. Yandulov−Schrock cycle for the transformation of dinitrogen into ammonia catalyzed by 1. [Mo] represents [HIPTN3N]Mo.



coordinated N2 is converted to a nitride [Mo(N)] complex and an ammonia molecule by the stepwise addition of protons and electrons. Their experimental studies prompted computational approaches to clarify the mechanism of nitrogen fixation in the Mo−triamidoamine system.13−23 Cao and co-workers first reported DFT calculations using a simplified model of 1, [PhN3N]Mo(N2).13 They adopted NH4+ and an iron−sulfur cluster as a pair of proton/electron sources. Studt and Tuczek14 and Magistrato and co-workers15 investigated the entire energy profile of the catalytic cycle using simplified models, [HN3N]Mo(N2) and [PhN3N]Mo(N2), respectively. Reiher and coworkers pointed out that the oversimplification of the HIPTN3N ligand can cause substantial errors in the calculated energetics.16,17 They proposed an alternative reaction pathway for the formation of a diazenide intermediate, where protonation of an amide N atom of the HIPTN3N ligand precedes protonation of the N2 ligand.16,18 Later, this pathway

NITROGEN FIXATION CATALYZED BY CUBANE-TYPE METAL−SULFIDO COMPLEXES The chemistry of cubane-type metal−sulfido clusters ligating nitrogenous substrates has been receiving much attention as structural and functional models of FeMo cofactor.35 Mizobe and co-workers reported the first isolation of a cubane-type metal−sulfido cluster ligating N2: [{Ru(N2) (tmeda)}(Cp*Ir)3(μ3-S)4] (3) in Figure 2a.36 A characteristic NN stretching at 2019 cm−1 is considerably red-shifted from that of free dinitrogen (2359 cm−1), indicating effective N2 activation in 3. Their study encouraged us to computationally investigate the potential of 3 as a catalyst for nitrogen fixation.37 DFT calculations demonstrated reductive activation of the NN bond in a model of 3, [{Ru(tmeda)}(N2) (CpIr)3(μ3S)4] ([Ru(N2)]). Figure 2b presents a possible catalytic cycle of nitrogen fixation by [Ru]. The calculated cycle is based on the

Figure 2. (a) The structure of [{Ru(N2)(tmeda)}(Cp*Ir)3(μ3-S)4] 3. (b) Calculated catalytic cycle of nitrogen fixation by [{Ru(N2)(tmeda)}(CpIr)3(μ3-S)4] ([Ru(N2)]). Energy changes in individual reaction steps are presented in kcal/mol, where protonation (red) and reduction (blue) steps are shown in different colors. C

DOI: 10.1021/acs.accounts.6b00033 Acc. Chem. Res. XXXX, XXX, XXX−XXX

Article

Accounts of Chemical Research Yandulov−Schrock cycle, where LutH+ and CoCp*2 were adopted as proton and electron sources. Energy changes (ΔE) in the protonation steps were evaluated based on the equation [X] + LutH+ → [XH]+ + Lut, where [X] is a neutral intermediate. The energy profiles of the reduction steps were calculated based on the equation [XH]+ + [CoCp*2] → [XH] + [CoCp*2]+, where [XH]+ is a protonated intermediate. As shown in Figure 2b, almost all the reaction steps proceed in an exothermic way although the first protonation is endothermic by 17.8 kcal/mol. The calculated energy profile indicates that [Ru] can serve as a catalyst for nitrogen fixation if appropriate proton/electron sources are chosen. Contrary to the promising DFT result, no experimental evidence has been found for the protonation of the N2 ligand in 3. Mizobe and co-workers observed only catalytic evolution of H2 followed by elimination of the N2 ligand.38 This result suggests that the N2 ligand in 3 is not sufficiently activated for functionalization. To pursue the possibility of the MIr3S4 core as a catalyst, we predicted the reactivity of dinitrogen complexes of MIr3S4 clusters [M(N2)] with LutH+, where M = V, Cr, Mn, Fe, Co, Ni, Cu, Ru, Mo, and W.38 Generally, the degree of N2 activation is judged from elongation of the N−N bond and red-shift of the NN stretching relative to free dinitrogen. The amount of negative charges on coordinated N2 are also a useful criterion for judging the degree of N2 activation39 We found that MoIr3S4 and WIr3S4 cores particularly exhibit high N2-activating ability. The N−N distance and NN stretching of [Mo(N2)] ([W(N2)]) are 1.150 (1.161) Å and 1952 (1889) cm−1, respectively, indicating N2-activating ability superior to [Ru(N2)] (1.128 Å and 2118 cm−1). The NPA charges40 on the N2 ligand are −0.33 and −0.49 in [Mo(N2)] and [W(N2)], respectively, both of which are significantly more negative than that in [Ru(N2)] (−0.10). We next assessed the reactivity of [M(N2)] (M = Ru, Mo, W) with LutH+. Optimized structures and energy profiles for the protonation of the N2 ligand are presented in Figure 3. Protonation of the N2 ligand in [Ru(N2)] would not proceed due to a high activation energy of 39.5 kcal/mol (Figure 3a), which agrees with the experimental result. On the other hand, the N2 ligand in [Mo(N2)] and [W(N2)] is easily protonated (Figure 3b). Formation of the protonated complex requires very low activation energies for the Mo and W complexes (6.4 and 1.6 kcal/mol). All the results suggest that the MoIr3S4 and WIr3S4 clusters are best suited for the activation and protonation of the coordinated N2. After this prediction, Mizobe and co-workers synthesized a metal−sulfido cluster having the MoIr3S4 core. As a preliminary result, the Mo site binds CO, which is isoelectronic with N2, whereby the CO ligand exhibited a νCO band in an extremely low frequency region, 1725 cm−1 for [{Mo(CO)(dppe)}(Cp*Ir)3(μ3-S)4] (dppe = Ph2PCH2CH2PPh2), in contrast to 1925 cm−1 for [{Ru(CO)(dppe)}(Cp*Ir)3(μ3-S)4].

Figure 3. Optimized structures and energy profiles for the proton transfer from LutH+ to N2 bound to (a) the RuIr3S4 cluster and (b) the MIr3S4 clusters (M = Mo and W (in parentheses)). Relative energies and interatomic distances are given in kcal/mol and Å, respectively.

equiv of NH3 were produced based on the catalyst. We retrieved some pieces of information from the experimental results9,24 to contemplate a possible catalytic cycle: (i) A COsubstituted dinitrogen-bridged dimolybdenum complex, [{Mo(CO)2(PNP)}2(μ-N2)], produced less than a stoichiometric amount of ammonia. (ii) A mononuclear molybdenum complex, [Mo(N2)2(PNP)(PMe2Ph)], has no catalytic activity. (iii) The catalytic activity of 2 strongly depends on the counteranion of LutH+. The amount of produced ammonia was drastically decreased when [LutH]Cl and [LutH]BArF4 were used as proton sources. (iv) Mo−nitride complexes, [Mo( N)Cl(PNP)] and [Mo(N)Cl(PNP)]OTf showed a similar catalytic activity to 2. Considering these findings, we developed a hypothesis in the DFT-based proposal: (a) Terminal N2 ligands are necessary for the catalytic reaction. (b) Dinuclear Mo−N2 complexes are involved in the catalytic cycle. (c) Triflate (OTf−) that is the counteranion of LutH+ plays a role. (d) A dinuclear Mo complex is separated into mononuclear ones to afford a Mo−nitride complex, [Mo(N)(OTf)(PNP)]. We also assumed (e) the stepwise protonation and reduction of coodinated N2. A catalytic mechanism deduced from theoretical and experimental studies is summarized in Figure 4. The mechanism involves separation of a dinuclear Mo complex after sequential protonation/reduction of a terminal N2 ligand as well as regeneration of 2 linked with the exchange of ammonia for dinitrogen. Batista and co-workers independently proposed a remarkable role of dinuclear Mo complexes in the catalytic activity of 2.41 Figure 5 presents detailed mechanims for the transformation of dinitrogen into ammonia starting from 2. Activation energies of the proton transfer from LuH+ were evaluated for all protonation steps. Energy profiles of the reduction steps by CoCp2 were calculated based on the equation [XH]+ + [CoCp2] → [XH] + [CoCp2]+, where [XH]+ is a protonated



NITROGEN FIXATION CATALYZED BY DINITROGEN-BRIDGED DIMOLYBDENUM COMPLEXES Nishibayashi and co-workers reported the Mo-catalyzed nitrogen fixation using dinitrogen-bridged dimolybdenum complex 2, as shown in Scheme 3.9 Dinitrogen was converted into ammonia at room temperature in the presence of [LutH]OTf (OTf = CF3SO3) as a proton source and cobaltocene (CoCp2) as a reducing reagent, where up to 23 D

DOI: 10.1021/acs.accounts.6b00033 Acc. Chem. Res. XXXX, XXX, XXX−XXX

Article

Accounts of Chemical Research

Figure 4. Proposed catalytic cycle of the transformation of dinitrogen into ammonia by dimolybdenum complex 2. The numbering of intermediates are given in Figure 5.

Figure 5. A possible reaction pathway starting from 2. Energy changes (activation energies) for individual reaction steps are presented in kcal/mol.

E

DOI: 10.1021/acs.accounts.6b00033 Acc. Chem. Res. XXXX, XXX, XXX−XXX

Article

Accounts of Chemical Research intermediate. Energy changes and activation energies (in parentheses) were calculated at the B3LYP*42 level of theory. The catalytic cycle starts with the protonation of a terminal dinitrogen ligand in 2 to form [Mo(N2)(NNH)−NN− Mo(N2)2]+, A, where Mo = Mo(PNP). The N2 ligand trans to the NNH group in A is then replaced by OTf− (A → B → C; Figure 5a). Protonation and reduction of C afford a hydrazide(2−) complex [Mo(OTf)(NNH 2 )−NN− Mo(N2)2], D. Reduction of [Mo(OTf)(NNH3)−NN− Mo(N2)2]+, E, results in the formation of ammonia and [Mo(OTf)(N)−NN−Mo(N2)2], F. Complex F is separated into two mononuclear complexes, a key intermediate [Mo(OTf)(N)], G, and [Mo(N2)3], H (Figure 5b). Figure 5c describes sequential protonation/reduction of G leading to ammine complex L. Coordination of dinitrogen to imide complex I gives a six-coordinate complex [Mo(OTf)(N2)(NH)], J, and then the imide ligand is reduced to the ammine ligand to afford complex L. The catalytic cycle is completed by the regeneration of 2 linked with the exchange of the NH3 ligand for an incoming N2 molecule (Figure 5d). Reduction of L results in the elimination of OTf−. A fivecoordinate complex [Mo(N2)(NH3)], M, couples with H generated from F to afford a dinuclear complex N. Finally, the NH3 ligand in N is replaced by dinitrogen to regenerate 2. Here we focus on two key steps in Figure 5a,b, (i) the protonation of a terminal N2 ligand in 2 and (ii) the formation of mononuclear molybdenum−nitride complex G. Dinitrogen complex 2 has two possible reaction sites for the protonation, the terminal and bridging N2 ligands. The protonation of the bridging N2 ligand requires a high activation energy (Ea = 40.7 kcal/mol) due to steric hindrance from tertbutyl groups in the pincer ligands. On the other hand, the protonation of a terminal N2 ligand is likely to occur (Ea = 8.4 kcal/mol) although this reaction is endothermic by 6.5 kcal/ mol. While proton detachment from the NNH group in A easily occurs (A → 2; Ea = 1.9 kcal/mol), the protonation of 2 prompts the elimination of the N2 ligand trans to the NNH group, where the reaction is 2.7 kcal/mol exothermic with an activation energy of 4.4 kcal/mol. After the N2 elimination, OTf− occupies the vacant coordination site in B to cancel the electronic charge (B + OTf− → C; ΔE = −15.6 kcal/mol). Totally, the first protonation process (2 → C) is exothermic by 11.8 kcal/mol. In the proposed mechanism in Figure 5, the formation of [Mo(OTf)(N)], G, is regarded as a key reaction step. Once the nitride intermediate F is formed, the bond energy between the Mo(N) atom and the bridging N2 ligand is decreased to only 4.4 kcal/mol. The small Mo−N bond energy would rationalize the separation of F into the corresponding mononuclear complexes G and H. Experimentally, fragments assignable to [Mo(OTf)(N)−NN−Mo(N 2 )] and [Mo(OTf)(N)] were observed by mass spectrometry from the stoichiometric reaction of 2 with [LutH]OTf. Here we would like to emphasize that the mononuclear Mo− N2 complex, [Mo(N2)3], H, does not react with LutH+, and hence the dinuclear structure must be regenerated to start the next catalytic cycle. It may be strange because the NPA charges on a terminal N2 ligand, which is a good criterion for judging the degree of N2 activation, are almost identical in the di- and mononuclear Mo complexes (−0.12 for 2 and −0.11 for H). Why is a dimolybdenum structure connected with dinitrogen required for the protonation of N2 ligands? Figure 6a shows the distribution of the NPA charges in A. The sum of the NPA

Figure 6. (a) NPA charge distributions in the protonated complex A. The values present the total atomic charges assigned to units A and B. (b) A schematic drawing of the intermetallic electron transfer between two Mo cores at the protonation step. (c) Spatial distribution of the HOMO of 2.

charges assigned to each Mo unit is calculated to be +0.66 for unit A and +0.34 for unit B, indicating that unit B donates 0.34 electron to unit A during the protonation of a terminal N2 ligand in unit A. This suggests that one Mo unit in 2 supports the other to accept a proton and stabilize the generated diazenide ligand by the intermetallic electron transfer. From a viewpoint of the NPA charges, a terminal N2 ligand to be protonated in 2 is not “preactivated”, but it can be activated with the aid of the other Mo unit when necessary (Figure 6b). Synergy of the Mo centers through the bridging N2 ligand can be observed in the HOMO of 2 (Figure 6c). The HOMO of 2 is comprised of d-orbitals of the two Mo atoms and a π-orbital of the bridging N2 ligand. The bonding π orbital interacting F

DOI: 10.1021/acs.accounts.6b00033 Acc. Chem. Res. XXXX, XXX, XXX−XXX

Article

Accounts of Chemical Research with the metal d-orbitals is distorted due to the twisted connection of the two Mo units (the N2−Mo−Mo−N2 dihedral angle of −54.4°). The intermetallic electron transfer allows 2 to accept a proton at the first step of the nitrogen fixation. The proposed catalytic mechanism prompted us to design more reactive dinitrogen-bridged dimolybdenum complexes as catalysts for nitrogen fixation.22 As shown in Figure 5a, the first protonation of a terminal N2 ligand is energetically most unfavorable. We envisaged enhancing the electron-donating ability of the Mo centers to accelerate the first protonation process. Introduction of an electron-donating group to the pyridine ring of 2 is expected to intensify the back-donating ability of the metals to the N2 ligands. We evaluated the impact of introduction of electron-donating groups to the 4-position of the pyridine ring in the pincer ligand. Figure 7 describes energy profiles of the first

Scheme 4. Nitrogen Fixation Catalyzed by Dimolybdenum Complexes Bearing a Variety of Substituted PNP Pincer Ligands

molybdenum−triamidoamine complex, the dinitrogen-bridged dimolybdenum complexes, and the cubane-type MIr 3 S 4 clusters. In particular, DFT calculations in close liaison with experiments lead to the proposal of detailed reaction pathways for nitrogen fixation catalyzed by the dimolybdenum complex. The calculated mechanism tells us a crucial role of the dinuclear Mo−NN−Mo structure in the protonation of coordinated N2. Intermetallic electron transfer via the bridging N2 ligand stabilizes the generated diazenide (Mo−NNH) intermediate. Dimolybdenum complexes newly designed based on the theoretical findings exhibit higher catalytic activity than the original complex. We now have confidence that the mechanistic understanding and rational ligand design provided by theoretical calculations will further accelerate the development of metal-complex-catalyzed nitrogen fixation systems. Quite recently we have found a new Mo-catalyzed nitrogen fixation system using molybdenum-nitride complexes bearing a triphosphine (PPP) ligand, [Mo(N)(PPP)Cl] and [Mo(N)(PPP)Cl]BArF4 as catalysts.43 The latter has the highest catalytic activity for nitrogen fixation, where 63 equiv of NH3 were produced based on the Mo atom. Interestingly, no dinuclear complexes such as [Mo(N2)2(PPP)]2(μ-N2) have been isolated and characterized yet. We now have great interest in whether the formation of dinuclear complexes may be indispensable in the reaction pathway of the transformation dinitrogen into ammonia.

Figure 7. Energy profiles of the first protonation process of 2a, 2b, and 2c. Energies relative to 2−LutH+ are presented in kcal/mol.

protonation process calculated for 2a (R = H), 2b (R = Me), and 2c (R = OMe). The activation energies for the protonation of 2a (8.4 kcal/mol) are decreased to 7.4 kcal/mol for 2b and 7.0 kcal/mol for 2c. Methyl- and methoxy-substituted PNP ligands can accelerate the protonation of 2. The NPA charge on a terminal N2 ligand is slightly increased from −0.125 in 2a to −0.127 in 2b and −0.133 in 2c. The increased negative charge on N2 indicates reductive activation of the coordinated N2 through an enhanced back-donation from Mo. We therefore expected that 2c would serve as an effective catalyst. Nishibayashi and co-workers prepared a variety of substituted dimolybdenum complexes (2b−2f), in which substituents such as Me, MeO, Ph, Me3Si, and tBu groups were introduced to the 4-position of the pyridine ring of the PNP-pincer ligand.23 The catalytic transformation of dinitrogen into ammonia were then examined using 2b−2f as catalysts (Scheme 4). In excellent accordance with the computational design, 2c produced the largest amount of NH3 (34 equiv based on 2c). The use of a larger amount of CoCp2 (360 equiv) and [LutH]OTf (480 equiv) gave 52 equiv of NH3.



AUTHOR INFORMATION

Corresponding Authors

*E-mail: kazunari@ms.ifoc.kyushu-u.ac.jp (K.Y.) *E-mail: ynishiba@sys.t.u-tokyo.ac.jp (Y. N.). Funding

K.Y. and H.T. acknowledge Grants-in-Aid for Scientific Research (Nos. 22245028, 24109014, 24550190, and 26888008) from the Japan Society for the Promotion of Science (JSPS) and the Ministry of Education, Culture, Sports, Science and Technology of Japan (MEXT) and the MEXT Projects of “Integrated Research on Chemical Synthesis” and “Elements Strategy Initiative to Form Core Research Center”. Y.N. acknowledges Grants-in-Aid for Scientific Research (Nos. 26288044, 26105708, 15K13687, and 15H05798) from JSPS and MEXT. The present project is supported by CREST, Japan Science and Technology Agency (JST).



CONCLUDING REMARKS We have summarized the synergetic relationships between computation and experiment in elucidating the mechanisms of nitrogen fixation catalyzed by transition metal complexes, the

Notes

The authors declare no competing financial interest. G

DOI: 10.1021/acs.accounts.6b00033 Acc. Chem. Res. XXXX, XXX, XXX−XXX

Article

Accounts of Chemical Research Biographies

dinitrogen complexes. Chem. Rev. 1995, 95, 1115−1133. (c) MacKay, B. A.; Fryzuk, M. D. Dinitrogen coordination chemistry: On the biomimetic borderlands. Chem. Rev. 2004, 104, 385−401. (7) (a) Nishibayashi, Y. Recent progress in transition-metal-catalyzed reduction of molecular dinitrogen under ambient reaction conditions. Inorg. Chem. 2015, 54, 9234−9247. (b) Khoenkhoen, N.; de Bruin, B.; Reek, J. N. H.; Dzik, W. I. Reactivity of dinitrogen bound to mid- and late-transition-metal centers. Eur. J. Inorg. Chem. 2015, 2015, 567−598. (8) Yandulov, D. V.; Schrock, R. R. Catalytic reduction of dinitrogen to ammonia at a single molybdenum center. Science 2003, 301, 76−78. (9) Arashiba, K.; Miyake, Y.; Nishibayashi, Y. A molybdenum complex bearing PNP-type pincer ligands leads to the catalytic reduction of dinitrogen into ammonia. Nat. Chem. 2011, 3, 120−125. (10) Anderson, J. S.; Rittle, J.; Peters, J. C. Catalytic conversion of nitrogen to ammonia by an iron model complex. Nature 2013, 501, 84−88. (11) Schrock, R. R. Catalytic reduction of dinitrogen to ammonia by molybdenum: Theory versus experiment. Angew. Chem., Int. Ed. 2008, 47, 5512−5522. (12) (a) Kinney, R. A.; McNaughton, R. L.; Chin, J.; Schrock, R. R.; Hoffman, B. M. Protonation of the dinitrogen-reduction catalyst [HIPTN3N]MoIII investigated by ENDOR spectroscopy. Inorg. Chem. 2011, 50, 418−420. (b) Munisamy, T.; Schrock, R. R. An electrochemical investigation of intermediates and processes involved in the catalytic reduction of dinitrogen by [HIPTN3N]Mo (HIPTN3N = (3,5-(2,4,6-i-Pr3C6H2)2C6H3NCH2CH2)3N). Dalton Trans. 2012, 41, 130−137. (13) Cao, Z.; Zhou, Z.; Wan, H.; Zhang, Q. Enzymatic and catalytic reduction of dinitrogen into ammonia: Density functional theory characterization of alternative molybdenum active sites. Int. J. Quantum Chem. 2005, 103, 344−353. (14) Studt, F.; Tuczek, F. Energetics and mechanism of a roomtemperature catalytic process for ammonia synthesis (Schrock cycle): Comparison with biological nitrogen fixation. Angew. Chem., Int. Ed. 2005, 44, 5639−5642. (15) Magistrato, A.; Robertazzi, A.; Carloni, P. Nitrogen fixation by a molybdenum catalyst mimicking the function of nitrogenase enzyme: A critical evaluation of DFT and solvent effects. J. Chem. Theory Comput. 2007, 3, 1708−1720. (16) Reiher, M.; Le Guennic, B.; Kirchner, B. Theoretical study of catalytic dinitrogen reduction under mild conditions. Inorg. Chem. 2005, 44, 9640−9642. (17) Le Guennic, B.; Kirchner, B.; Reiher, M. Nitrogen fixation under mild ambient conditions: Part I−The initial dissociation/association step at molybdenum triamidoamine complexes. Chem. - Eur. J. 2005, 11, 7448−7460. (18) Schenk, S.; Le Guennic, B.; Kirchner, B.; Reiher, M. Firstprinciples investigation of the Schrock mechanism of dinitrogen reduction employing the full HIPTN3N ligand. Inorg. Chem. 2008, 47, 3634−3650. (19) Schenk, S.; Kirchner, B.; Reiher, M. A stable six-coordinate intermediate in ammonia-dinitrogen exchange at Schrock’s molybdenum catalyst. Chem. - Eur. J. 2009, 15, 5073−5082. (20) Thimm, W.; Gradert, C.; Broda, H.; Wennmohs, F.; Neese, F.; Tuczek, F. Free reaction enthalpy profile of the Schrock cycle derived from density functional theory calculations on the full [MoHIPTN3N] catalyst. Inorg. Chem. 2015, 54, 9248−9255. (21) Schenk, S.; Reiher, M. Ligands for dinitrogen fixation at Schrock-type catalysts. Inorg. Chem. 2009, 48, 1638−1648. (22) Hölscher, M.; Leitner, W. DFT investigation of the potential of [H-M{(NHCH2CH2)3X}] catalysts (M = Mo, Ru, Os; X = N, P) for the reduction of N2 to NH3 by H2. Eur. J. Inorg. Chem. 2006, 2006, 4407−4417. (23) Bergeler, M.; Simm, G. N.; Proppe, J.; Reiher, M. Heuristicsguided exploration of reaction mechanisms. J. Chem. Theory Comput. 2015, 11, 5712−5722. (24) Tanaka, H.; Arashiba, K.; Kuriyama, S.; Sasada, A.; Nakajima, K.; Yoshizawa, K.; Nishibayashi, Y. Unique behaviour of dinitrogen-

Hiromasa Tanaka received his Ph.D. degree from Tohoku University in 2001 under the supervision of Prof. Hiroshi Kudo. After two postdoctoral years in the group of Prof. R. E. Silverans and Prof. P. Lievens at KU Leuven (Belgium), he worked for RIKEN, Ritsumeikan University, Kyushu University, and Kyoto University as a researcher. In 2015, he returned to Kyushu University and was promoted to research associate professor in the group of Prof. Yoshizawa. Yoshiaki Nishibayashi received his Ph.D. in 1995 from Kyoto University under the supervision of Professor Sakae Uemura. He became an assistant professor at the University of Tokyo in 1995 and moved to Kyoto University in 2000. In 2005, he became an associate professor at the University of Tokyo as PI. Since 2016, he has been a full professor at the University of Tokyo as PI. He received the Chemical Society of Japan Award for Distinguished Young Chemists in 2001, the JSPS Prize in 2012, the Green & Sustainable Chemistry Honorable Award in 2012, and the SSOCJ Company Award for Novel Reaction & Method in 2015. His current research interests are focused on organic and organometallic chemistry. Kazunari Yoshizawa received his Bachelor, Master, and Ph.D. degrees at Kyoto University under the direction of Kenichi Fukui and Tokio Yamabe. After spending one year at Cornell University as a postdoctoral associate with Roald Hoffmann, he joined the faculty of Kyoto University, where he became an assistant and then associate professor. In 2001, he moved as a full professor to Kyushu University, where his research interests are extended to enzymatic and catalytic reactions and electronic properties of molecules and solids. He received the Chemical Society of Japan Award for Creative Work in 2011.

■ ■

DEDICATION Dedicated to the late professor Yasushi Mizobe of the University of Tokyo. REFERENCES

(1) Liu, H. Ammonia Synthesis Catalysts; Chemical Industry Press & World Scientific: Singapore, 2013. (2) (a) Einsle, O.; Tezcan, F. A.; Andrade, S. L.; Schmid, B.; Yoshida, M.; Howard, J. B.; Rees, D. C. Nitrogenase MoFe-protein at 1.15 Å resolution: a central ligand in the FeMo-cofactor. Science 2002, 297, 1696−1700. (b) 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. (c) 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 ironmolybdenum cofactor. Science 2011, 334, 974−977. (3) (a) Burgess, B. K.; Lowe, D. J. Mechanism of molybdenum nitrogenase. Chem. Rev. 1996, 96, 2983−3011. (b) 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. (4) Allen, D.; Senoff, C. V. Nitrogenpentaammineruthenium(II) complexes. Chem. Commun. 1965, 24, 621. (5) (a) Hinrichsen, S.; Broda, H.; Gradert, C.; Söncksen, L.; Tuczek, F. Recent developments in synthetic nitrogen fixation. Annu. Rep. Prog. Chem., Sect. A: Inorg. Chem. 2012, 108, 17−47. (b) Nishibayashi, Y. Molybdenum-catalyzed reduction of molecular dinitrogen under mild reaction conditions. Dalton Trans. 2012, 41, 7447−7453. (d) Tanabe, Y.; Nishibayashi, Y. Developing more sustainable processes for ammonia synthesis. Coord. Chem. Rev. 2013, 257, 2551−2564. (6) (a) Chatt, J.; Dilworth, J. R.; Richards, R. L. Recent advances in the chemistry of nitrogen fixation. Chem. Rev. 1978, 78, 589−625. (b) Hidai, M.; Mizobe, Y. Recent advances in the chemistry of H

DOI: 10.1021/acs.accounts.6b00033 Acc. Chem. Res. XXXX, XXX, XXX−XXX

Article

Accounts of Chemical Research

cooperativity: A computational study. Eur. J. Inorg. Chem. 2014, 2014, 6126−6133. (e) Hölscher, M.; Leitner, W. Reduction of metal coordinated N2 to NH3 with H2 by heterolytic hydrogen cleavage induced by external Lewis bases − a DFT study. Z. Anorg. Allg. Chem. 2015, 641, 72−77. (35) (a) Hidai, M.; Mizobe, Y. Research inspired by the chemistry of nitrogenase − Novel metal complexes and their reactivity toward dinitrogen, nitriles, and alkynes. Can. J. Chem. 2005, 83, 358−374. (b) Henderson, R. A. Mechanistic studies on synthetic Fe−S-based clusters and their relevance to the action of nitrogenases. Chem. Rev. 2005, 105, 2365−2438. (36) Mori, H.; Seino, H.; Hidai, M.; Mizobe, Y. Isolation of a cubanetype metal sulfido cluster with a molecular nitrogen ligand. Angew. Chem., Int. Ed. 2007, 46, 5431−5434. (37) Tanaka, H.; Mori, H.; Seino, H.; Hidai, M.; Mizobe, Y.; Yoshizawa, K. DFT study on chemical N2 fixation by using a cubanetype RuIr3S4 cluster: Energy profile for binding and reduction of N2 to ammonia via Ru−N−NHx (x = 1−3) intermediates with unique structures. J. Am. Chem. Soc. 2008, 130, 9037−9047. (38) Tanaka, H.; Ohsako, F.; Seino, H.; Mizobe, Y.; Yoshizawa, K. Theoretical study on activation and protonation of dinitrogen on cubane-type MIr3S4 clusters (M = V, Cr, Mn, Fe, Co, Ni, Cu, Mo, Ru, and W). Inorg. Chem. 2010, 49, 2464−2470. (39) (a) Deeth, R. J.; Field, C. N. A computational study of metal− dinitrogen co-ordination. J. Chem. Soc., Dalton Trans. 1994, 1943− 1948. (b) Studt, F.; Tuczek, F. Theoretical, spectroscopic, and mechanistic studies on transition-metal dinitrogen complexes: Implications to reactivity and relevance to the nitrogenase problem. J. Comput. Chem. 2006, 27, 1278−1291. (40) Reed, A. E.; Curtiss, L. A.; Weinhold, F. Intermolecular interactions from a natural bond orbital, donor-acceptor viewpoint. Chem. Rev. 1988, 88, 899−926. (41) Tian, Y.-H.; Pierpont, A. W.; Batista, E. R. How does Nishibayashi’s molybdenum complex catalyze dinitrogen reduction to ammonia? Inorg. Chem. 2014, 53, 4177−4183. (42) (a) Reiher, M.; Salomon, O.; Hess, B. A. Reparametrization of hybrid functionals based on energy differences of states of different multiplicity. Theor. Chem. Acc. 2001, 107, 48−55. (b) Reiher, M. Theoretical study of the [Fe(phen)2(NCS)2] spin-crossover complex with reparametrized density functionals. Inorg. Chem. 2002, 41, 6928− 6935. (43) Arashiba, K.; Kinoshita, E.; Kuriyama, S.; Eizawa, A.; Nakajima, K.; Tanaka, H.; Yoshizawa, K.; Nishibayashi, Y. Catalytic reduction of dinitrogen to ammonia by use of molybdenum-nitride complexes bearing a tridentate triphosphine as catalysts. J. Am. Chem. Soc. 2015, 137, 5666−5669.

bridged dimolybdenum complexes bearing pincer ligand towards catalytic formation of ammonia. Nat. Commun. 2014, 5, 3737. (25) Kuriyama, S.; Arashiba, K.; Nakajima, K.; Tanaka, H.; Kamaru, N.; Yoshizawa, K.; Nishibayashi, Y. Catalytic formation of ammonia from Molecular dinitrogen by use of dinitrogen-bridged dimolybdenum−dinitrogen complexes bearing PNP-pincer ligands: Remarkable effect of substituent at PNP-pincer ligand. J. Am. Chem. Soc. 2014, 136, 9719−9731. (26) (a) Tanaka, H.; Sasada, A.; Kouno, T.; Yuki, M.; Miyake, Y.; Nakanishi, H.; Nishibayashi, Y.; Yoshizawa, K. Molybdenum-catalyzed transformation of molecular dinitrogen into silylamine: Experimental and DFT study on the remarkable role of ferrocenyldiphosphine ligands. J. Am. Chem. Soc. 2011, 133, 3498−3506. (b) Yuki, M.; Tanaka, H.; Sasaki, K.; Miyake, Y.; Yoshizawa, K.; Nishibayashi, Y. Iron-catalysed transformation of molecular dinitrogen into silylamine under ambient conditions. Nat. Commun. 2012, 3, 1254. (c) Imayoshi, R.; Tanaka, H.; Matsuo, Y.; Yuki, M.; Nakajima, K.; Yoshizawa, K.; Nishibayashi, Y. Cobalt-catalyzed transformation of molecular dinitrogen into silylamine under ambient reaction conditions. Chem. - Eur. J. 2015, 21, 8905−8909. (d) Kuriyama, S.; Arashiba, K.; Nakajima, K.; Tanaka, H.; Yoshizawa, K.; Nishibayashi, Y. Nitrogen fixation catalyzed by ferrocene-substituted dinitrogen-bridged dimolybdenum−dinitrogen complexes: unique behavior of ferrocene moiety as redox active site. Chem. Sci. 2015, 6, 3940−3951. (27) Anderson, J. S.; Cutsail, G. E., III; Rittle, J.; Connor, B. A.; Gunderson, W. A.; Zhang, L.; Hoffman, B. M.; Peters, J. Characterization of an FeN−NH2 intermediate relevant to catalytic N2 reduction to NH3. J. Am. Chem. Soc. 2015, 137, 7803−7809. (28) Solari, E.; Da Silva, C.; Iacono, B.; Hesschenbrouck, J.; Rizzoli, C.; Scopelliti, R.; Floriani, C. Photochemical activation of the NN bond in a dimolybdenum−dinitrogen complex: Formation of a molybdenum nitride. Angew. Chem., Int. Ed. 2001, 40, 3907−3909. (29) Reiher, M.; Kirchner, B.; Hutter, J.; Sellmann, D.; Hess, B. A. A photochemical activation scheme of inert dinitrogen by dinuclear RuII and FeII complexes. Chem. - Eur. J. 2004, 10, 4443−4453. (30) Kunkely, H.; Vogler, A. Photolysis of aqueous [(NH3)5Os(μN2)Os(NH3)5]5+: Cleavage of dinitrogen by an intramolecular photoredox reaction. Angew. Chem., Int. Ed. 2010, 49, 1591−1593. (31) Huss, A. S.; Curley, J. J.; Cummins, C. C.; Blank, D. A. Relaxation and dissociation following photoexcitation of the (μN2)[Mo(N[t-Bu]Ar3)2 dinitrogen cleavage intermediate. J. Phys. Chem. B 2013, 117, 1429−1436. (32) (a) Fieser, M. E.; Bates, J. E.; Ziller, J. W.; Furche, F.; Evans, W. J. Dinitrogen reduction via photochemical activation of heteroleptic tris(cyclopentadienyl) rare-earth complexes. J. Am. Chem. Soc. 2013, 135, 3804−3807. (b) Fieser, M. E.; Johnson, C. W.; Bates, J. E.; Ziller, J. W.; Furche, F.; Evans, W. J. Dinitrogen reduction, sulfur reduction and isoprene polymerization via photochemical activation of trivalent bis(cyclopentadienyl) rare-earth-metal allyl complexes. Organometallics 2015, 34, 4387−4393. (33) (a) Miyazaki, T.; Tanaka, H.; Tanabe, Y.; Yuki, M.; Nakajima, K.; Yoshizawa, K.; Nishibayashi, Y. Cleavage and formation of molecular dinitrogen in a single system assisted by molybdenum complexes bearing ferrocenyldiphosphine. Angew. Chem., Int. Ed. 2014, 53, 11488−11492. (b) Rebreyend, C.; de Bruin, B. Photolytic N2 splitting: Road to sustainable NH3 production? Angew. Chem., Int. Ed. 2015, 54, 42−44. (34) (a) Hölscher, M.; Prechtl, M. H. G.; Leitner, W. Can [M(H)2(H2) (PCP) pincer complexes (M = Fe, Ru, Os; X = N, O, S) serve as catalyst lead structures for NH3 synthesis from N2 and H2? Chem. - Eur. J. 2007, 13, 6636−6643. (b) Hölscher, M.; Leitner, W. Lewis acid assisted stabilization of side-on bonded N2 in [Ru(NCN)]pincer complexes−Computational catalyst design directed at NH3 synthesis from N2 and H2. Chem. - Eur. J. 2010, 16, 14266−14271. (c) Hölscher, M.; Leitner, W. Heterolytic outer-sphere cleavage of H2 for the reduction of N2 in the coordination sphere of transition metals−A DFT study. Angew. Chem., Int. Ed. 2012, 51, 8225−8229. (d) Hölscher, M.; Leitner, W. Low activation barriers in N2 reduction with H2 at ruthenium pincer complexes induced by ligand I

DOI: 10.1021/acs.accounts.6b00033 Acc. Chem. Res. XXXX, XXX, XXX−XXX