Search for: All records

Bimetallic cleavage of dinitrogen has emerged as a highly promising approach to synthesis using N2, particularly its conversion to NH3. It is generally considered that thermal bimetallic cleavage proceeds only through MNNM units with a delocalized 𝜋10 electronic configuration. We report herein a N2-bridged complex with a 𝜋8 configuration, [(PArNP)MoI]2(μ-1:1-N2) (2; PArNP = Ozerov’s anionic PNP pincer ligand). As expected, 2 displays a high barrier to thermal N2 cleavage which occurs only slowly at 110 °C (k = 1.65 x 10-4 s-1). However, at room temperature 2 catalyzes the conversion of N2 to NH3 by Cp*2Co and collidinium triflate. Experiments in the absence of reductant reveal that cleavage is catalyzed by Brønsted acids. DFT analysis indicates that this proceeds via protonation of the μ-N2 ligand, to give a diazenido bridge; N-N cleavage of this bridge is spin- and symmetry-allowed with a low calculated barrier (G‡ = 20 kcal/mol). The mononuclear product of cleavage of 2, (PArNP)Mo(N)I (1-(N)I), was characterized crystallographically and by EPR spectroscopy. 1-(N)I has a half-filled non-bonding d orbital; as a result, hydrogen-atom transfer or proton-coupled electron transfer to yield the corresponding imide is calculated to be much more thermodynamically favorable than analogous additions to the closed-shell nitrides derived from 𝜋10 complexes. This finding is calculated to be general for 𝜋8 versus 𝜋10 cleavage products, with implications for the design of molecular catalysts for N2 conversion to NH3. 

The thioether–diphosphine pincer-ligated molybdenum complex (PSP)MoCl3 (1-Cl3, PSP = 4,5-bis(diisopropylphosphino)-2,7-di-tert-butyl-9,9-dimethyl-9H-thioxanthene) has been synthesized as a catalyst-precursor for N2 reduction catalysis with a focus on an integrated experimental/computational mechanistic investigation. The (PSP)Mo unit is isoelectronic with the (PNP)Mo (PNP = 2,6-bis(di-t-butylphosphinomethyl)pyridine) fragment found in the family of catalysts for the reduction of N2 to NH3 first reported by Nishibayashi and co-workers. Electrochemical studies reveal that 1-Cl3 is significantly more easily reduced than (PNP)MoCl3 (with a potential ca. 0.4 eV less negative). The reaction of 1-Cl3 with two reducing equivalents, under N2 atmosphere and in the presence of iodide, affords the nitride complex (PSP)Mo(N)(I). This observation suggests that the N2-bridged complex [(PSP)Mo(I)]2(N2) is formed and undergoes rapid cleavage. DFT calculations predict the splitting barrier of this complex to be low, in accord with calculations of (PNP)Mo and a related (PPP)Mo complex reported by Merakeb et al. Conversion of the nitride ligand to NH3 has been investigated in depth experimentally and computationally. Considering sequential addition of H atoms to the nitride through proton coupled electron-transfer or H-atom transfer, formation of the first N–H bond is thermodynamically relatively unfavorable. Experiment and theory, however, reveal that an N–H bond is readily formed by protonation of (PSP)Mo(N)(I) with lutidinium chloride, which is strongly promoted by coordination of Cl− to Mo. Other anions, e.g. triflate, can also act in this capacity although less effectively. These protonations, coupled with anion coordination, yield MoIV imide complexes, thereby circumventing the difficult formation of the first N–H bond corresponding to a low BDFE and formation of the respective MoIII imide complexes. The remaining two N–H bonds required to produce ammonia are formed thermodynamically much more favorably than the first. Computations suggest that formation of the MoIV imide is followed by a second protonation, then a rapid and favorable one-electron reduction, followed by a third protonation to afford coordinated ammonia. This comprehensive analysis of the elementary steps of ammonia synthesis provides guidance for future catalyst design. 

Iridium dibromide complexes of the phenyldiimine ligand 2,6-bis(1-((2,6-dimethylphenyl)imino)ethyl)phenyl, trans-(XyPhDI)IrBr2L, have been synthesized, and relative Ir-L BDFEs have been experimentally determined for a wide range of corresponding adducts of ligands L. An estimate of the absolute enthalpy of Ir-L binding has been obtained from dynamic NMR measurements. The results of DFT calculations are in very good agreement with the relative and absolute experimental values. Computational studies were extended to the formation of adducts of (XyPhDI)IrH2 and (XyPhDI)IrI, as well as other (pincer)IrI fragments, (Phebox)IrI and (PCP)IrI, to enable a comparison of electronic and steric effects with these archetypal pincer ligands. Attempts to reduce (XyPhDI)IrBr2(MeCN) to a hydride or an IrI complex yielded a dinuclear CN-bridged complex with a methyl ligand on the cyanide-C-bound Ir center (characterized by scXRD), indicating that C-CN bond cleavage took place at that Ir center. DFT calculations indicate that the C-CN bond cleavage occurs at one Ir center with strong assistance by coordination of the CN nitrogen to the other Ir center. 

The thioether-diphosphine pincer-ligated molybdenum complex, (PSP)MoCl3 (1-Cl3, PSP = 4,5-bis(diisopropylphosphino)-2,7-di-tert-butyl-9,9-dimethyl-9H-thioxanthene) has been synthesized as a catalyst-precursor for N2 reduction catalysis, with a focus on an integrated experimental/computational mechanistic investigation. The (PSP)Mo unit is isoelectronic with the (PNP)Mo (PNP = 2,6-bis(di-t-butylphosphinomethyl)pyridine) fragment found in the family of catalysts for the reduction of N2 to NH3 first reported in 2011 by Nishibayashi and co-workers. Under an atmosphere of N2 the reaction of 1-Cl3 with three reducing equivalents yields the dinuclear penta-dinitrogen Mo complex [(PSP)Mo(N2)2](-N2), 2. Electrochemical studies reveal that 1-Cl3 is significantly more easily reduced than (PNP)MoCl3 (with a potential ca. 0.4 eV less negative). The bridging-nitrogen complex 2 shows no indication of undergoing N2 cleavage to Mo nitride complexes. The reaction of 1-Cl3 with only two reducing equivalents, however, under N2 atmosphere and in the presence of iodide, affords the product of N2 cleavage, the nitride complex (PSP)Mo(N)(I). DFT calculations implicate another N2-bridged complex, [(PSP)Mo(I)]2(N2), as a viable intermediate in facile N2 cleavage to yield (PSP)Mo(N)(I). Conversion of the nitride ligand to NH3 has been studied. If considering sequential addition of H atoms to the nitride, formation of the first N-H bond is by far the thermodynamically least favorable of the three N-H bond formation steps. The first N-H bond was formed by reaction of (PSP)Mo(N)(I) with [LutH]Cl, where coordination of Cl– to Mo plays an essential role. Computations suggest that a second protonation, followed by a rapid and very favorable one-electron reduction, and then a third protonation, furnishes ammonia. In agreement with calculations, ammonia can be generated using either mild H-atom transfer reagents or mild reductants/acids. This comprehensive analysis of the elementary steps of ammonia synthesis and the role of the central pincer donor and halide association provides guidance for future catalyst designs. 

Iridium dibromide complexes of the phenyldiimine ligand 2,6-bis(1-((2,6-dimethylphenyl)imino)ethyl)phenyl, trans-(XyPhDI)IrBr2L, have been synthesized, and relative Ir-L BDFEs have been experimentally determined for a wide range of corresponding adducts of ligands L. An estimate of the absolute enthalpy of Ir-L binding has been obtained from dynamic NMR measurements. The results of DFT calculations are in very good agreement with the relative and absolute experimental values. Computational studies were extended to the formation of adducts of (XyPhDI)IrH2 and (XyPhDI)Ir(I), as well as other (pincer)Ir(I) fragments, (Phebox)Ir(I) and (PCP)Ir(I), to enable a comparison of electronic and steric effects with these archetypal pincer ligands. Attempts to reduce (XyPhDI)IrBr2(MeCN) to a hydride or an Ir(I) complex yielded a dinuclear CN-bridged complex with a methyl ligand on the cyanide-C-bound Ir center (characterized by scXRD), indicating that C-CN bond cleavage took place at that Ir center. DFT calculations indicate that the C-CN bond cleavage occurs at one Ir center with strong assistance by coordination of the CN nitrogen to the other Ir center. 

Metal ligand cooperativity (MLC) has revealed a plethora of unusual reactivity in catalysis in the last couple of decades. Since Milstein's report of aromatization-dearomatization of the pincer backbone of pyridine-based-pincer complexes, ruthenium has played a partic ularly important role in the develo pment of M LC. We have recently reported a (H- P3 )Ir complex which is the fastest known catalyst for alkane-transfer dehydrogenation. The active species results from P- to-Ir migration of H in this system. We further explored the possib ility of MLC in an analogous Ru system. Surprisingly, when metalating the same H-P3 ligand with a RuCl2 precursor we only isolated a (Cl-P3 )Ru(H)Cl complex where H had migrated to Ru from P, and Cl to P from Ru ("P- H/M-X exchange"). We have demonstrated that the thermodynamically favored direction of such exchanges depends strongly on the ancillary ligands, with particular driving force for formation of 5-coordinate (pincer)MHCl complexes (M = d6 metal center) . However, for 6- coordinate Ru complexes (H- pincer)MXYL, the electronic nature of L appears to determine if P-H/M-X exchange occurs. Strongly pi-accepting ligands promote P-X/M-H exchange with the reaction observed for L = CO, xylylisonitrile and N O+ , but not for L = N2 , C H3 CN, or PMe3 . While exchange at 5- coordinate (16e- ) Ru centers appears to proceed through initial P-to-Ru migration of X or H, to give a phosphide interme diate, in the case of 6- coordinate (18e- ) Ru centers exchange is believed to proceed through phosphoranyl intermediates. DFT and intrinsic bond orbital anal. has been used to better understand this reactivity.