Methane C–H Activation via 3d Metal Methoxide ... - ACS Publications

Sep 22, 2017 - four-centered, kite-shaped transition state, σ-bond metathesis, or oxidative hydrogen migration has been found for methane activation ...
1 downloads 0 Views 3MB Size
Download PDF
Article pubs.acs.org/IC

Methane C−H Activation via 3d Metal Methoxide Complexes with Potentially Redox-Noninnocent Pincer Ligands: A Density Functional Theory Study Ahmad Najafian and Thomas R. Cundari* Department of Chemistry, Center of Advanced Scientific Computing and Modeling (CASCaM), University of North Texas, 1155 Union Circle, No. 305070, Denton, Texas 76203-5017, United States

Downloaded via UNIV OF SOUTH DAKOTA on June 30, 2018 at 15:29:53 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

S Supporting Information *

ABSTRACT: This paper reports a density functional theory study of 3d transition-metal methoxide complexes with potentially redox-noninnocent pincer supporting ligands for methane C−H bond activation to form methanol (LnM-OMe + CH4 → LnM−Me + CH3OH). The three types of tridentate pincer ligands [terpyridine (NNN), bis(2-pyridyl)phenylC,N,N′ (NCN), and 2,6-bis(2-phenyl)pyridine-N,C,C′ (CNC)] and different first-row transition metals (M = Ti, V, Cr, Mn, Fe, Co, Ni, and Cu) are used to elucidate the reaction mechanism as well as the effect of the metal identity on the thermodynamics and kinetics of a methane activation reaction. Spin-density analysis indicates that some of these systems, the NNN and NCN ligands, have redox-noninnocent character. A four-centered, kite-shaped transition state, σ-bond metathesis, or oxidative hydrogen migration has been found for methane activation for the complexes studied. Calculations suggest that the d electron count is a more significant factor than the metal formal charge in controlling the thermodynamics and kinetics of C−H activation and late 3d metal methoxides, with high d counts preferred. Notably, early-to-middle metals tend toward oxidative hydrogen migration and late metals undergo a pathway that is more akin to σ-bond metathesis, suggesting that metal methoxide complexes that favor σ-bond metathesis pathways for methane activation will yield lower barriers for C−H activation.

■

added transition state (TS),20 and σ-complex-assisted metathesis.21 In general, two major reaction sequences are needed to complete a catalytic cycle for alkane functionalization: C−O bond formation and C−H bond activation (Scheme 1). Previously, Cundari et al. conducted intermolecular insertion of an oxygen atom into a metal−methyl bond to produce a

INTRODUCTION Catalytic functionalization of methane, the major component of natural gas, to an easily transportable and highly desirable liquid such as methanol in a single step has attracted considerable attention from both academic and industrial chemists.1−10 Methane activation faces substantial difficulties such as the inertness of the aliphatic C−H bond, very high C−H bond dissociation enthalpies, and the fact that methane is an exceedingly poor base and acid.11 Other challenges include stopping methane at its first level of oxidation (methanol) and thus catalyst selectivity.9 The economic importance of these processes has encouraged many researchers to focus on the experimental and theoretical aspects of selective catalyst strategies by well-defined transition-metal complexes.12−15 On the basis of the complex’s active site, C−H activation reactions may be subdivided into three major pathways: metal-centered (e.g., oxidative addition), ligand-centered (e.g., hydrogen-atom abstraction), and metal−ligand-centered (e.g., [2 + 2] addition).16 In another type of classification, C−H bond activation mechanisms have been described by σ-bond metathesis (σBM) and oxidative addition/reductive elimination (OA/RE) mechanisms and alternative mechanisms that lie between these two extremes such as oxidative hydrogen migration (OHM),17,18 metal-assisted σBM,19 an oxidatively © 2017 American Chemical Society

Scheme 1. Catalytic Cycle for Alkane Functionalization

Received: July 7, 2017 Published: September 22, 2017 12282

DOI: 10.1021/acs.inorgchem.7b01736 Inorg. Chem. 2017, 56, 12282−12290

Article

Inorganic Chemistry methoxide intermediate.22−25 The most notable conclusion from these studies is that there are three major reaction pathways: (1) a nonredox, one-step insertion of an oxygen atom into a M−C bond by an organometallic Baeyer−Villiger (OMBV) pathway, (2) a two-step oxidation pathway in which an initial oxygen-atom transfer occurs to form an oxo intermediate, followed by a methyl migration to the oxo ligand to yield a methoxide product,26 and (3) reductive functionalization, which is in essence an organometallic SN2 reaction.27 Garrett et al. carried out a study of M−C bond functionalization using β-diketiminate methyl complexes spanning the 3d series from the s block (Ca) to the p block (Ge), demonstrating that the OMBV mechanism for oxy insertion is preferred only when the dn electron count renders a chemically infeasible formal oxidation state, e.g., a ZnIV-O. In the work of Garrett et al., the metal d-orbital count was more important than the impact of the ligand. 26 In a study of the reductive functionalization mechanism by Fallah and Cundari, 3d metal [M(diamine)2(CH3)(Cl)] complexes were used to study a SN2 mechanism for the nucleophilic attack of hydroxide as a route to C−O bond formation; it was concluded that the ligand was of greater significance than the metal in moderating the reaction coordinate.27 Regardless of the mechanism, once a metal methoxide intermediate has been made, one must then generate methanol and regenerate the LnM−Me species to complete the catalytic cycle, which is the focus of the present research. Oxgaard and co-workers reported the selective C−H activation of benzene with an oxygen-donor iridium methoxo complex and demonstrated the generation of a phenyl complex while producing methanol in high yield.28 To the best of our knowledge, a systematic study of C−H activation of methane to produce methanol via 3d metal methoxide complexes has not yet been reported. Most of the less expensive and more abundant first-row transition metals undergo one-electron transformations, whereas many organometallic reactions require two-electron, e.g., oxidative addition or reductive, elimination steps to avoid radical intermediates. Recently, some researchers have incorporated redox-active ligands into the metal coordination sphere of base metals to mimic the electronic properties and reactivity of precious metals.29−32 The close energy levels of the ligand- and metal-based frontier orbitals, as well as the spatial overlap among these orbitals, can facilitate delocalization across both the metal and ligand.33 In these systems, the redox-noninnocent (RNI) ligand can act as reservoir and participate at different steps of the catalytic cycle by receiving and/or supplying electrons.34 We focus our attention in the present study on 3d metal catalyst models with terpyridine (tpy) and two other tpy-like supporting ligands. These pincer ligands are σ donors and π acceptors and thus tend to stabilize metals in lower oxidation states. Furthermore, the rigidity of tridentate pincers may confer high thermal stability to complexes. Terpyridine complexes have been used for a number of important organic and inorganic reactions.35−37 For example, Pahls and coworkers investigated the effect of different electronic environments of tpy-[Rh] complexes for reductive functionalization via both experimental and computational methods.38 Furthermore, previous studies by Jones et al. show that terpyridine has redoxactive properties.39 Also, Vicic et al. have done calculations on nickel terpyridine complexes and showed a large amount of radical character at the carbon atoms ortho and para to the

central nitrogen of the terpyridine ring, indicating ligand redox noninnocence.40 Herein, density functional theory (DFT) calculations of the conversion of methane to methanol through a four-centered TS by first-row transition-metal methoxide complexes with three potentially redox-active pincer ligands are computed to elucidate important trends in this reaction (Scheme 2). An Scheme 2. General Reaction Scheme for the Conversion of Methane to Methanol through a Four-Centered TS Mediated by a Metal Methoxide Complex

improved understanding of this reaction could pave the way for the design of catalysts for the selective partial oxidation of light alkanes using earth-abundant 3d metals. In this research, the following questions will be addressed: (i) What are the pertinent pathways of the reaction? (ii) Which factors control the thermodynamics and kinetics of the reaction? (iii) What is the impact of the metal identity? (iv) What is the influence of the supporting ligand? (v) Do the modeled supported ligands act in a RNI manner?

■

COMPUTATIONAL METHODS

DFT calculations were performed using the Gaussian 09 software package.41 The energies and geometries of the reactants, TSs, and products were calculated with the hybrid functional, combining Becke’s three-parameter exchange functional (B3) with the Lee− Yang−Parr (LYP) functional, denoted as B3LYP, in conjunction with the 6-31G(d) basis set.42−47 This level of theory was chosen based on previous work on oxy insertion.27,48 Quoted energies are free energies in kilocalories per mole and were carried out at 1 atm and 298.15 K. All DFT calculations were optimized in a dimethyl sulfoxide (DMSO; ε = 46.826) solvent using the SMD continuum solvent model.49 All possible spin states including low, intermediate, and high spins were calculated for each metal complex in their reactants, TSs, and products. The lowest free-energy spin multiplicities were selected to calculate thermodynamics (ΔGrxn) and free-energy barriers (ΔG⧧) for the studied reactions. All geometries were optimized and evaluated for the correct number of imaginary frequencies through vibrational frequency calculations using an analytic Hessian. The ground states (GSs) are characterized by the possession of no imaginary frequencies. TSs have one imaginary frequency through harmonic frequency calculations.

■

RESULTS AND DISCUSSION 1. Methoxide Reactant Complex Models. In order to have a more fundamental understanding of the methane-tomethanol conversion process mediated by methoxide complexes, a series of analogous four-coordinate methoxide complexes were modeled using first-row transition metals with different d counts and formal charges. Typical optimized structures of (I) (tpy)M-OMe, (II) (NCN)M-OMe, and (III) (CNC)M-OMe reactants are depicted in Figure 1. A set of metals where M = Ti, V, Cr, Mn, Fe, Co, Ni, and Cu is used with formal d-orbital electron counts (assuming that the supporting ligands are in typical redox states) from 1 to 10 to

12283

DOI: 10.1021/acs.inorgchem.7b01736 Inorg. Chem. 2017, 56, 12282−12290

Article

Inorganic Chemistry

Figure 2. Prototypical DFT-optimized four-centered TS for the pincer-M-OMe reacting with methane.

Figure 1. Representation of the optimized geometries for the three types of modeled methoxide complex reactants: (I) (tpy)M-OMe (X, Y, Z = N); (II) (NCN)M-OMe (X, Z = N; Y = C); (III) (CNC)MOMe (Y = N; X, Z = C).

and two new σ bonds are formed without a change in the metal formal oxidation state.21 Metathesis mechanisms are known for both early and late transition metals for C−H bond activation53 and for addition across both single and multiple metal−ligand bonds. In this mechanism, both the metal and ligand are involved in cleaving the C−H bond; therefore, it is a metal− ligand-directed mechanism. Metal−ligand-centered methane C−H bond activation can facilitate the possible utilization of earth-abundant 3d metals because there is, unlike oxidative addition, no requirement for a formal two-electron oxidation state change, nor are radical intermediates involved. According to Cundari et al. and Wolczanski and co-workers, computational54 and experimental55 evidence indicates that metal− ligand-centered mechanisms are best for activation of the strongest Csp3−H bonds. The OHM TS has a characteristic between σBM and OA/RE and also happens in a concerted fashion like these two mechanisms. It has a character similar to that of oxidative addition and requires accessibility of the Mn+2 formal oxidation state, but because there is no stable intermediate, it is not a true OA/RE pathway. The TS types can be differentiated by measuring the M−H distance (Scheme 3).

evaluate possible trends in the reactions. Also, using different metals delineates which factors control the thermodynamics and kinetics of the C−H methane activation reaction, one of the main aims of this research. Three pincer supporting ligands were evaluated: terpyridine (NNN or tpy), 2,6-bis(2-pyridyl)phenyl-C,N,N′ (NCN), and 2,6-bis(2-phenyl)pyridine-N,C,C′ (CNC) (Figure 1). These tpy-like supporting ligands were chosen because of the fact that metal complexes of polypyridines such as 2,2′:6′,2″-terpyridine (terpy), bipyridine, and other diimine ligands have been extensively investigated by experimentalists over the past three-quarters of a century.50−52 In reactant methoxide compounds, the pincer supporting ligands along with the methoxide actor ligand yield a squareplanar (or pseudo-square-planar) geometry, which then reacts with methane to yield methanol and methyl complex products (Scheme 2). For most complexes, except Ti and V, the metal coordination geometry is planar or nearly so. On the basis of an initial assumption of redox innocence, terpyridine would be a neutral ligand, NCN would have a 1− charge, and CNC is 2−, which results in the central metal ions being formally 1+, 2+, and 3+, respectively. The overall charges of the metal reactant, TS, and product complexes are neutral to minimize the effect of the solvent on the computed free energies. By changing the metal identity and spectator ligand, the aim is to understand how the d-orbital electron count and metal formal charge may influence the thermodynamics and kinetics of C−H methane activation. 2. Transition States. One of the most important goals for this research is finding an appropriate TS for methane activation by the methoxide complex. Figure 2 illustrates a typical TS and suggests a four-centered TS for C−H activation of methane. When methane activation occurs at the metal, the methoxide ligand will serve as a base while a M−C bond is formed. The TS is kite-shaped because of the near-linear C−H−O angle. The H 1s orbital is nondirectional and capable of simultaneously interacting with CH3, OMe, and/or the metal to lower the reaction barrier. For such a geometry, two TS descriptions seem plausible: σ-bond metathesis (σBM⧧) and oxidative hydrogen migration (OHM⧧). σBM⧧ involves a concerted, one-step reaction by which two σ bonds are broken

Scheme 3. M−H Bond Distance Differentiating σBM from OHM TSs

In both of these pathways, the mechanistic nuance involves whether the hydrogen atom of the C−H bond being activated forms a bond with the metal to create a metal hydride intermediate.56 If the OHM TS presents a fully or nearly fully formed bond between the metal and active hydrogen, it is logical to expect the M−H distance in an OHM TS to be 12284

DOI: 10.1021/acs.inorgchem.7b01736 Inorg. Chem. 2017, 56, 12282−12290

Article

Inorganic Chemistry shorter compared to an analogous σBM TS for which this bond is not fully formed. 3. Terpyridine Complexes. Generally, terpyridine is considered to be a neutral ligand; however, some literature indicates a potential for redox RNI behavior for this ligand.39,40 Scheme 4 depicts two possible electronic structures for a

and TS; the terpyridine ligand in the V reactant and product states is also redox-innocent (RI). Interestingly, in Ti, the supporting ligand in the reactant is tpy0 (i.e., innocent), but in the product, it is tpy− (RRI → TSRI → PRNI). For V, tpy is RNI only in the TS, which means that the reaction coordinate has changed its character; thus, RRI → TSRNI → PRI. Hence, in RNI cases, the metal has reduced the terpyridine to 1− and is itself formally oxidized from 1+ to 2+ (note that very little spin density was calculated on the OMe ligands). On the basis of the spin-density plot and Mulliken spin population, we describe the majority of terpyridine complexes as [(tpy−)MII-OMe] in the reactant or [(tpy−)MIICH3] in the product. Therefore, one should consider the metal in the 2+ oxidation state for most terpyridine complexes, because one electron prefers to go to the more energetically accessible π* orbital of terpyridine rather than a metal-centered orbital. For example, in the Cr case, in the reactant methoxide complex, terpyridine has approximately one electron unpaired, thus resulting in a 4[(tpy)−CrII(OMe)−] description, which would make CrII a high-spin d4 ion in this quartet complex. In the chromium methyl product, the description is the same: 4 [(tpy)−(HS-d4-CrII)(Me)−], which implies that the spin on the terpyridine radical anion is antiferromagnetically coupled to the four-electron spin on HS-d4-CrII. Spin states are conserved in terpyridine-ligated reactants and products for most metals, except Mn and Ti. The results are reported in Table 1. For Mn, the most stable multiplicity in the

Scheme 4. Depiction of the RI (blue) and RNI (green) States of a (tpy)M-OMe Complexa

a

Note that for most systems the spin density is spread through the pyridine rings.

reactant terpyridine complex. To gain insight into these systems, spin-density plots were calculated (Figure 3).

Table 1. Calculated Lowest-Energy Spin State, Thermodynamics (ΔGrxn, kcal/mol), and Kinetics (ΔG⧧, kcal/mol) of the (tpy)M-OMe + CH4 → (tpy)M-Me + CH3OH Reaction Ti V Cr Mn Fe Co Ni Cu

R

TS

P

ΔG⧧

ΔGrxn

doublet triplet quartet septet sextet triplet doublet singlet

doublet triplet quartet quintet sextet quintet doublet singlet

quartet triplet quartet quintet sextet triplet doublet singlet

53.4 45.0 47.5 40.2 46.1 42.5 45.5 43.2

48.2 38.0 34.0 26.3 27.9 27.0 26.3 30.6

reactant was septet, which is an unusual spin state for a formally MnI ion (vide infra) and was another factor that initially indicated that terpyridine can be RNI in the complexes of interest. 4. NCN-M and CNC-M Pincer Motifs. The spin is conserved throughout the reaction coordinate for both NCN and CNC supporting ligands (see Tables 2 and3). For example, for the (CNC)Ni-OMe complex, C−H activation of methane is endergonic, ΔGrxn = 15.6 kcal/mol, and has a barrier free energy of ΔG⧧ = 44.1 kcal/mol; the former quantity is the lowest ΔGrxn among all complexes studied here (Scheme 5). To evaluate whether NCN and CNC are RI or not, spindensity plots were prepared and demonstrated that for most complexes the spin density is mainly concentrated on the 3d metal in these two ligands (Figure 4). Hence, it is concluded that NCN and CNC are RI ligands for these complexes and the metal in these complexes are formally 2+ and 3+, respectively. In addition, the results are supported by Mulliken spin analysis, which shows nearly all of the unpaired electron densities on the metal for these two families of complexes excluding the

Figure 3. Spin densities (isovalue = 0.01) of (a) RNI 7[(tpy)MnOMe] and (b) RNI 4[(tpy)Ti-CH3].

Significant spin density is present on the terpyridine, suggesting redox noninnocence for terpyridine for most reactants, TSs, and products. To elucidate this further, the Mulliken spindensity population was analyzed, and it was further confirmed that terpyridine is a RNI ligand for the majority of terpyridine complexes in this research, except for titanium in the reactant 12285

DOI: 10.1021/acs.inorgchem.7b01736 Inorg. Chem. 2017, 56, 12282−12290

Article

Inorganic Chemistry Table 2. Calculated Lowest-Energy Spin State, Thermodynamics (ΔGrxn, kcal/mol), and Kinetics (ΔG⧧, kcal/mol) of the (NCN)M-OMe + CH4 → (NCN)M-Me + CH3OH Reaction Ti V Cr Mn Fe Co Ni Cu

R

TS

P

ΔG⧧

ΔGrxn

triplet quartet quintet sextet quintet doublet singlet doublet

triplet quartet quintet sextet quintet doublet singlet doublet

triplet quartet quintet sextet quintet doublet singlet doublet

54.0 41.4 44.2 42.1 48.7 42.4 53.2 45.7

41.8 33.8 34.5 28.5 31.3 29.8 30.6 25.0

Table 3. Calculated Lowest-Energy Spin State, Thermodynamics (ΔGrxn, kcal/mol), and Kinetics (ΔG⧧, kcal/mol) of the (CNC)M-OMe + CH4 → (CNC)M-Me + CH3OH Reaction Ti V Cr Mn Fe Co Ni

R

TS

P

ΔG⧧

ΔGrxn

doublet triplet quartet quintet quartet triplet doublet

doublet triplet quartet quintet quartet triplet doublet

doublet triplet quartet quintet quartet triplet doublet

54.1 51.8 45.4 53.4 51.8 48.4 44.1

45.1 47.2 41.2 25.1 25.1 27.5 15.6

Scheme 5. B3LYP/6-31G(d) Computed Free-Energy (kcal/ mol) Pathway of Methane Activation by a CNC-Ni-OMe Complexa

Figure 4. Spin density (isovalue = 0.01) of (a) RNI 3[(NCN)TiOMe] and (b) RI 4[(CNC)Fe-CH3].

between full RNI and RI behavior. For instance, for doublet (NCN)Cu-OMe and (NCN)Cu-Me Mulliken spin density on metal is ∼0.6 electron, indicating mixed RNI and RI character for this system. In the TSs, hydrogen oscillates between the methyl group and the oxygen atom of methoxide, typically corresponding to an imaginary frequency around 700i−1400i cm−1 in different TSs. As expected, in TSs, the bond lengths of the M−methyl, C−H, and O−H are longer in comparison with these bonds in a typical GS. Consider 2(CNC)Ni-OMe as a typical example; the bond distances are C−H = 1.10 Å (GS), C···H = 1.45 Å (TS); Ni-CH3 = 1.77 Å (GS), Ni···CH3 = 2.23 Å (TS); and O− Hmethanol = 0.97 Å (GS), O···.H = 1.19 Å (TS), indicating that the stretching in the bond changes from a single covalent bond in the reactant to a dative bond in the product, which is methanol. Other complexes are similar (see Table S-2 in the Supporting Information). A comparison of the C−H and O−H bond distances in the TS indicates that C−H cleavage of methane by methoxide complexes has a late TS because the O−H bond is shorter that the C−H bond in these TSs. Thus, these bonds are closer to the OH of the methanol product than the C−H bond length of the methane reactant, implying that the TS is more product-like than reactant-like. All optimized complexes have roughly similar bond lengths. The results

a

Other supporting ligands and metal ions are qualitatively similar. Prefixes indicate multiplicity. R = reactant, TS = transition state, and P = product.

NCN[Ti] system in its reactant, TS, and the product states and for NCN[V] in only the reactant state. It is worth mentioning that some reactants and products are ambiguous and lie 12286

DOI: 10.1021/acs.inorgchem.7b01736 Inorg. Chem. 2017, 56, 12282−12290

Article

Inorganic Chemistry

Figure 5a indicates that for metal formal charge there is no obvious trend with ΔGrxn, and so it was concluded that the effect of the formal charge on ΔGrxn is minimal. Note that we counted the (tpy)M-OMe complex as formally M+, although changing those with RNI tpy ligands to 2+ does not alter the foregoing conclusion. On the other hand, the d count, dn, seems to impact more significantly the thermodynamics than the formal charge of the 3d metal ion. In Figure 5b, the average ΔGrxn for the same d-orbital-count metals is plotted. The present research implies that higher d-orbital-electron-count metal methoxides are more thermodynamically favorable for methane activation. On the basis of the ΔGrxn calculations, middle and late metal alkoxide complexes are clearly more attractive for future study. B. Kinetics of Methane Activation. Calculation of the barrier energies for all modeled complex reactions indicates that all activation energies are large, ranging from 40.2 to 54.1 kcal/ mol (data plotted in Figure 6 and tabulated in the Supporting

indicate a late TS, which is consistent with an endothermic reaction by Hammond’s postulate.

■

SUMMARY, CONCLUSIONS, AND PROSPECTUS In this paper, a comprehensive DFT study of first-row transition-metal methoxide complexes with potentially redoxactive supporting ligands for methane C−H bond activation has been presented. The three types of pincer ligands and 3d metals spanning from Ti to Cu seek to identify novel catalysts. A fourcentered, kite-shaped TS for methane activation was found for the different complexes studied. Several significant observations have emerged regarding metal−ligand-mediated C−H activation and are summarized as follows. A. Thermodynamics of Methane Activation. Moving from left to right in the 3d series indicates that for most supporting ligands, early metals such as Ti, V, and Cr have ΔGrxn that are higher compared to middle and late 3d metals like Mn, Fe, Co, Ni and Cu. Given the anticipated greater strength of M−O versus M−C bonds, all methane activation reactions are endergonic from 48.2 kcal/mol for tpy[Ti] to 15.6 kcal/mol for CNC[Ni], which is reasonable because the C−O bond formation part of the methane oxidation catalytic cycle is notably exergonic.26,27 Figure 5 illustrates the calculated ΔGrxn (Scheme 2; LnM-OMe + CH4 → LnM-CH3 + MeOH) reaction as a function of the (a) metal formal charge and (b) d-orbital occupation (dn) to assess which factor affects the reaction more.

Figure 6. Plot of (a) average activation free energies (kcal/mol) for methane activation by LnM-OMe versus metal formal charge and (b) average activation free energies (kcal/mol) versus d count. The supporting ligands are assumed to possess their most common formal charge: tpy0, NCN−, or CNC2−.

Information). By a comparison of ΔG⧧ for the different metal complexes through the plots provided in Figure 6, it is concluded that the effect of the metal ion charge is less important than the d count and that high d-orbital-electroncount metals have lower ΔG⧧ for methane activation by 3d metal methoxide complexes. This conclusion reflects the trend in dn versus ΔGrxn discussed above, which happens to be inconsistent with the Hammond postulate.

Figure 5. Plot of the (a) average reaction free energies (kcal/mol) of methane activation versus 3d metal formal charge and (b) average reaction free energies (kcal/mol) of methane activation versus d count. The supporting ligands are assumed to possess their most common formal charge: tpy0, NCN−, or CNC2−. 12287

DOI: 10.1021/acs.inorgchem.7b01736 Inorg. Chem. 2017, 56, 12282−12290

Article

Inorganic Chemistry

M−H distance for Fe and Mn is between the OHM and σBM extremes. In the case of (CNC)M-OMe, early metals Ti and Cr activate via an OHM pathway, the V complex is through an intermediate and three middle metals (Mn, Fe, and Co), and Ni activates via an σBM pathway using the M−H bond lengthening as a guide. D. Other Factors Evaluated. The computed absolute value of the Mulliken spin density (in electrons) on the oxygen atom of the methoxide ligand shows no obvious correlation between ΔGrxn or ΔG⧧ and the spin density (Figure S-1 in the Supporting Information). Another obvious trend is that, upon moving from Ti to Cu, the electronegativity increases while the energy barrier decreases (Figure S-2 in the Supporting Information). In part, the latter trend mirrors the dn versus ΔG⧧ correlation discussed above because the more electronegative metals are generally found to the right in the 3d series. Analysis of the active site bond lengths (M−C, C−H, O−H, M−H, or M−O) did not reveal any obvious correlation between these bond distances and ΔG⧧ (see Table S-2 in the Supporting Information). Taken together, the present research indicates that the d electron count is a more significant factor than the metal formal charge in controlling the thermodynamics and kinetics of C−H activation by 3d metal methoxide complexes. This is a key reaction in realizing complete catalytic cycles for selective methane functionalization. Late 3d metal methoxides with high d counts are computed to be more favorable thermodynamically and kinetically. The calculated TSs are four-centered, and generally early-to-middle metals prefer OHM while late metals undergo a pathway that is more akin to σBM using the M−H distance as a determinant. The σBM pathway typically implies a good degree of proton character in the TS for the C−H bond being activated, perhaps implying that most 3d metal methoxides are simply not basic enough to activate the very weak base methane (pKa ∼ 50). Therefore, we hypothesize on the basis of this research that metal methoxide complexes that favor σBM pathways for methane activation will yield lower barriers for C−H activation. Because the kinetic barriers in this computational study are very high, future studies will focus on modification of the supporting ligands to late metal methoxides that may reduce barrier free energies. There is computational and experimental support of this possibility in studies of C−O bond formation by late metal complexes with substituted terpyridine ligands, so it appears reasonable to expect that similar reductions in ΔG⧧ may be obtained via supporting ligand manipulation.38,57

C. Nature of the TS: σBM or OHM? To assess which complexes prefer a specific TS, M−H bond lengths were measured in a representative GS (LnM-H) and compared to the same bond lengths in the calculated TSs. From these, the percent bond-length changes were calculated (see Figure 7 and

■

Figure 7. M−H distance (Å; left y axis) plots in the GS (blue line) and TS (red line) and percentage change (right, y axis; green line) of the (a) [(tpy)M], (b) [(NCN)M], and (c) [(CNC)M] complexes.

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.inorgchem.7b01736. Cartesian coordinates of all calculated species (TXT) Calculated thermodynamic and kinetic data for the methane activation reaction and optimized bond distances in TSs for all complexes modeled, plots of barrier energies versus oxygen spin density and metal electronegativity, and metal−hydrogen bond lengths (PDF)

Table S-3 in the Supporting Information). For NNN-[M] complexes, the percentage change in the TSs for late 3d metals, i.e., Co, Ni, and Cu, is almost 2 times longer (∼20%) versus most middle and early metals (∼10% M−H lengthening), indicating that late metals undergo σBM and the others are via a OHM mechanism or an intermediate pathway closer to OHM. Most NCN-[M] complexes are calculated to activate methane through an OHM pathway except Ni and Cu. Ni and Cu bear the signature of σBM, but the percent change in the 12288

DOI: 10.1021/acs.inorgchem.7b01736 Inorg. Chem. 2017, 56, 12282−12290

Article

Inorganic Chemistry

■

(14) Palkovits, R.; Antonietti, M.; Kuhn, P.; Thomas, A.; Schüth, F. Solid Catalysts for the Selective Low-Temperature Oxidation of Methane to Methanol. Angew. Chem., Int. Ed. 2009, 48, 6909−6912. (15) Wasserman, E. P.; Moore, C. B.; Bergman, R. G. Gas-Phase Rates of Alkane C-H Oxidative Addition to a Transient CpRh(CO) Complex. Science 1992, 255, 315. (16) Moulder, C. A.; Cundari, T. R. 5d Metal(IV) Imide Complexes. The Impact (or Lack Thereof) of d-Orbital Occupation on Methane Activation and Functionalization. Inorg. Chem. 2017, 56, 1823−1829. (17) Oxgaard, J.; Muller, R. P.; Goddard, W. A.; Periana, R. A. Mechanism of Homogeneous Ir(III) Catalyzed Regioselective Arylation of Olefins. J. Am. Chem. Soc. 2004, 126, 352−363. (18) Oxgaard, J.; Periana, R. A.; Goddard, W. A. Mechanistic Analysis of Hydroarylation Catalysts. J. Am. Chem. Soc. 2004, 126, 11658− 11665. (19) Webster, C. E.; Fan, Y.; Hall, M. B.; Kunz, D.; Hartwig, J. F. Experimental and Computational Evidence for a Boron-Assisted, σBond Metathesis Pathway for Alkane Borylation. J. Am. Chem. Soc. 2003, 125, 858−859. (20) Ng, S. M.; Lam, W. H.; Mak, C. C.; Tsang, C. W.; Jia, G.; Lin, Z.; Lau, C. P. C−H Bond Activation by a Hydrotris(pyrazolyl)borato Ruthenium Hydride Complex. Organometallics 2003, 22, 641−651. (21) Perutz, R. N.; Sabo-Etienne, S. The σ-CAM Mechanism: σ Complexes as the Basis of σ-Bond Metathesis at Late-Transition-Metal Centers. Angew. Chem., Int. Ed. 2007, 46, 2578−2592. (22) Figg, T. M.; Cundari, T. R. Mechanistic Study of Oxy Insertion into Nickel−Carbon Bonds with Nitrous Oxide. Organometallics 2012, 31, 4998−5004. (23) Figg, T. M.; Cundari, T. R.; Gunnoe, T. B. Non-redox OxyInsertion via Organometallic Baeyer−Villiger Transformations: A Computational Hammett Study of Platinum(II) Complexes. Organometallics 2011, 30, 3779−3785. (24) Figg, T. M.; Webb, J. R.; Cundari, T. R.; Gunnoe, T. B. Carbon−Oxygen Bond Formation via Organometallic Baeyer−Villiger Transformations: A Computational Study on the Impact of Metal Identity. J. Am. Chem. Soc. 2012, 134, 2332−2339. (25) Pouy, M. J.; Milczek, E. M.; Figg, T. M.; Otten, B. M.; Prince, B. M.; Gunnoe, T. B.; Cundari, T. R.; Groves, J. T. Flavin-Catalyzed Insertion of Oxygen into Rhenium−Methyl Bonds. J. Am. Chem. Soc. 2012, 134, 12920−12923. (26) Garrett, E. C.; Figg, T. M.; Cundari, T. R. Impact of d-Orbital Occupation on Metal−Carbon Bond Functionalization. Inorg. Chem. 2014, 53, 7789−7798. (27) Fallah, H.; Cundari, T. R. Reductive functionalization of 3d metal−methyl complexes: The greater importance of ligand than metal. Comput. Theor. Chem. 2015, 1069, 86−95. (28) Tenn, W. J.; Young, K. J. H.; Bhalla, G.; Oxgaard, J.; Goddard, W. A.; Periana, R. A. CH Activation with an O-Donor Iridium− Methoxo Complex. J. Am. Chem. Soc. 2005, 127, 14172−14173. (29) Cowley, R. E.; Bill, E.; Neese, F.; Brennessel, W. W.; Holland, P. L. Iron(II) Complexes with Redox-Active Tetrazene (RNNNNR) Ligands. Inorg. Chem. 2009, 48, 4828−4836. (30) Lippert, C. A.; Hardcastle, K. I.; Soper, J. D. Harnessing RedoxActive Ligands for Low-Barrier Radical Addition at Oxorhenium Complexes. Inorg. Chem. 2011, 50, 9864−9878. (31) Nawn, G.; Waldie, K. M.; Oakley, S. R.; Peters, B. D.; Mandel, D.; Patrick, B. O.; McDonald, R.; Hicks, R. G. Redox-Active Bridging Ligands Based on Indigo Diimine (“Nindigo”) Derivatives. Inorg. Chem. 2011, 50, 9826−9837. (32) Ray, K.; Petrenko, T.; Wieghardt, K.; Neese, F. Joint spectroscopic and theoretical investigations of transition metal complexes involving non-innocent ligands. Dalton Trans 2007, 1552−66. (33) Lever, A. B. P. Electronic characteristics of an extensive series of ruthenium complexes with the non-innocent o-benzoquinonediimine ligand: A pedagogical approach. Coord. Chem. Rev. 2010, 254, 1397− 1405.

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Ahmad Najafian: 0000-0003-3802-8538 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was supported by the National Science Foundation (NSF) under Grant CHE-1464943. The authors also acknowledge the NSF for their support of the UNT Chemistry CASCaM high-performance computing facility through Grant CHE-1531468.

■

REFERENCES

(1) Conley, B. L.; Ganesh, S. K.; Gonzales, J. M.; Tenn, W. J.; Young, K. J. H.; Oxgaard, J.; Goddard, W. A.; Periana, R. A. Facile Functionalization of a Metal Carbon Bond by O-Atom Transfer. J. Am. Chem. Soc. 2006, 128, 9018−9019. (2) Conley, B. L.; Tenn, W. J.; Young, K. J. H.; Ganesh, S. K.; Meier, S. K.; Ziatdinov, V. R.; Mironov, O.; Oxgaard, J.; Gonzales, J.; Goddard, W. A.; Periana, R. A. Design and study of homogeneous catalysts for the selective, low temperature oxidation of hydrocarbons. J. Mol. Catal. A: Chem. 2006, 251, 8−23. (3) Crabtree, R. H. Alkane C-H activation and functionalization with homogeneous transition metal catalysts: a century of progress-a new millennium in prospect. J. Chem. Soc., Dalton Trans. 2001, 2437−2450. (4) Fu, R.; Nielsen, R. J.; Goddard, W. A.; Fortman, G. C.; Gunnoe, T. B. DFT Virtual Screening Identifies Rhodium−Amidinate Complexes As Potential Homogeneous Catalysts for Methane-toMethanol Oxidation. ACS Catal. 2014, 4, 4455−4465. (5) Jones, C. J.; Taube, D.; Ziatdinov, V. R.; Periana, R. A.; Nielsen, R. J.; Oxgaard, J.; Goddard, W. A. Selective Oxidation of Methane to Methanol Catalyzed, with C□H Activation, by Homogeneous, Cationic Gold. Angew. Chem., Int. Ed. 2004, 43, 4626−4629. (6) Labinger, J. A. Selective alkane oxidation: hot and cold approaches to a hot problem. J. Mol. Catal. A: Chem. 2004, 220, 27−35. (7) Periana, R. A.; Bhalla, G.; Tenn, W. J.; Young, K. J. H.; Liu, X. Y.; Mironov, O.; Jones, C. J.; Ziatdinov, V. R. Perspectives on some challenges and approaches for developing the next generation of selective, low temperature, oxidation catalysts for alkane hydroxylation based on the CH activation reaction. J. Mol. Catal. A: Chem. 2004, 220, 7−25. (8) Periana, R. A.; Taube, D. J.; Gamble, S.; Taube, H.; Satoh, T.; Fujii, H. Platinum Catalysts for the High-Yield Oxidation of Methane to a Methanol Derivative. Science 1998, 280, 560−564. (9) Shilov, A. E.; Shul’pin, G. B. Activation of C−H Bonds by Metal Complexes. Chem. Rev. 1997, 97, 2879−2932. (10) Shiota, Y.; Yoshizawa, K. Methane-to-Methanol Conversion by First-Row Transition-Metal Oxide Ions: ScO+, TiO+, VO+, CrO+, MnO+, FeO+, CoO+, NiO+, and CuO+. J. Am. Chem. Soc. 2000, 122, 12317−12326. (11) Olah, G. A.; Rasul, G. From Kekulé’s Tetravalent Methane to Five-, Six-, and Seven-Coordinate Protonated Methanes. Acc. Chem. Res. 1997, 30, 245−250. (12) McMullin, C. L.; Pierpont, A. W.; Cundari, T. R. Complete methane-to-methanol catalytic cycle: A DFT study of oxygen atom transfer from N2O to late-row (MNi, Cu, Zn) β-diketiminate CH activation catalysts. Polyhedron 2013, 52, 945−956. (13) Mei, J.; Carsch, K. M.; Freitag, C. R.; Gunnoe, T. B.; Cundari, T. R. Variable Pathways for Oxygen Atom Insertion into Metal−Carbon Bonds: The Case of Cp*W(O)2(CH2SiMe3). J. Am. Chem. Soc. 2013, 135, 424−435. 12289

DOI: 10.1021/acs.inorgchem.7b01736 Inorg. Chem. 2017, 56, 12282−12290

Article

Inorganic Chemistry (34) Dzik, W. I.; van der Vlugt, J. I.; Reek, J. N. H.; de Bruin, B. Ligands that Store and Release Electrons during Catalysis. Angew. Chem., Int. Ed. 2011, 50, 3356−3358. (35) Gong, H.; Gagné, M. R. Diastereoselective Ni-Catalyzed Negishi Cross-Coupling Approach to Saturated, Fully Oxygenated C-Alkyl and C-Aryl Glycosides. J. Am. Chem. Soc. 2008, 130, 12177−12183. (36) Gong, H.; Sinisi, R.; Gagné, M. R. A Room Temperature Negishi Cross-Coupling Approach to C-Alkyl Glycosides. J. Am. Chem. Soc. 2007, 129, 1908−1909. (37) Lin, X.; Phillips, D. L. Density Functional Theory Studies of Negishi Alkyl−Alkyl Cross-Coupling Reactions Catalyzed by a Methylterpyridyl-Ni(I) Complex. J. Org. Chem. 2008, 73, 3680−3688. (38) O’Reilly, M. E.; Pahls, D. R.; Webb, J. R.; Boaz, N. C.; Majumdar, S.; Hoff, C. D.; Groves, J. T.; Cundari, T. R.; Gunnoe, T. B. Reductive functionalization of a rhodium(iii)-methyl bond by electronic modification of the supporting ligand. Dalton Trans. 2014, 43, 8273−8281. (39) Jones, G. D.; Martin, J. L.; McFarland, C.; Allen, O. R.; Hall, R. E.; Haley, A. D.; Brandon, R. J.; Konovalova, T.; Desrochers, P. J.; Pulay, P.; Vicic, D. A. Ligand Redox Effects in the Synthesis, Electronic Structure, and Reactivity of an Alkyl−Alkyl Cross-Coupling Catalyst. J. Am. Chem. Soc. 2006, 128, 13175−13183. (40) Ciszewski, J. T.; Mikhaylov, D. Y.; Holin, K. V.; Kadirov, M. K.; Budnikova, Y. H.; Sinyashin, O.; Vicic, D. A. Redox Trends in Terpyridine Nickel Complexes. Inorg. Chem. 2011, 50, 8630−8635. (41) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H. P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery, J. A.; Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin, K. N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.; Raghavachari, K.; Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega, N.; Millam, J. M.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Martin, R. L.; Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas; Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian 09, revision D.01; Gaussian Inc.: Wallingford, CT, 2009. (42) Becke, A. D. Becke’s three parameter hybrid method using the LYP correlation functional. J. Chem. Phys. 1993, 98, 5648−5652. (43) Francl, M. M.; Pietro, W. J.; Hehre, W. J.; Binkley, J. S.; Gordon, M. S.; DeFrees, D. J.; Pople, J. A. Self-consistent molecular orbital methods. XXIII. A polarization-type basis set for second-row elements. J. Chem. Phys. 1982, 77, 3654−3665. (44) Lee, C.; Yang, W.; Parr, R. G. Development of the Colle-Salvetti correlation-energy formula into a functional of the electron density. Phys. Rev. B: Condens. Matter Mater. Phys. 1988, 37, 785. (45) Rassolov, V. A.; Pople, J. A.; Ratner, M. A.; Windus, T. L. 631G* basis set for atoms K through Zn. J. Chem. Phys. 1998, 109, 1223−1229. (46) Stephens, P.; Devlin, F.; Chabalowski, C.; Frisch, M. J. Ab initio calculation of vibrational absorption and circular dichroism spectra using density functional force fields. J. Phys. Chem. 1994, 98, 11623− 11627. (47) Vosko, S. H.; Wilk, L.; Nusair, M. Accurate spin-dependent electron liquid correlation energies for local spin density calculations: a critical analysis. Can. J. Phys. 1980, 58, 1200−1211. (48) Fallah, H.; Horng, F.; Cundari, T. R. Theoretical Study of Two Possible Side Reactions for Reductive Functionalization of 3d Metal− Methyl Complexes by Hydroxide Ion: Deprotonation and Metal− Methyl Bond Dissociation. Organometallics 2016, 35, 950−958. (49) Marenich, A. V.; Cramer, C. J.; Truhlar, D. G. Universal solvation model based on solute electron density and on a continuum model of the solvent defined by the bulk dielectric constant and atomic surface tensions. J. Phys. Chem. B 2009, 113, 6378−6396.

(50) Braterman, P. S.; Song, J. I.; Peacock, R. D. Electronic absorption spectra of the iron (II) complexes of 2, 2′-bipyridine, 2, 2′bipyrimidine, 1, 10-phenanthroline, and 2, 2′: 6′, 2″-terpyridine and their reduction products. Inorg. Chem. 1992, 31, 555−559. (51) de Bruin, B.; Bill, E.; Bothe, E.; Weyhermüller, T.; Wieghardt, K. Molecular and Electronic Structures of Bis(pyridine-2,6-diimine)metal Complexes [ML2](PF6)n (n = 0, 1, 2, 3; M = Mn, Fe, Co, Ni, Cu, Zn). Inorg. Chem. 2000, 39, 2936−2947. (52) Hayashi, Y.; Kita, S.; Brunschwig, B. S.; Fujita, E. Involvement of a Binuclear Species with the Re− C (O) O− Re Moiety in CO2 Reduction Catalyzed by Tricarbonyl Rhenium (I) Complexes with Diimine Ligands: Strikingly Slow Formation of the Re− Re and Re− C (O) O− Re Species from Re (dmb)(CO) 3S (dmb= 4, 4 ‘-Dimethyl-2, 2 ‘-bipyridine, S= Solvent). J. Am. Chem. Soc. 2003, 125, 11976−11987. (53) Janowicz, A. H.; Bergman, R. G. Carbon-hydrogen activation in completely saturated hydrocarbons: direct observation of M+ RH. fwdarw. M (R)(H). J. Am. Chem. Soc. 1982, 104, 352−354. (54) Olatunji-Ojo, O. A.; Cundari, T. R. C−H Activation by Multiply Bonded Complexes with Potentially Noninnocent Ligands: A Computational Study. Inorg. Chem. 2013, 52, 8106−8113. (55) Cummins, C. C.; Baxter, S. M.; Wolczanski, P. T. Methane and benzene activation via transient (tert-Bu3SiNH)2Zr:NSi-tert-Bu3. J. Am. Chem. Soc. 1988, 110, 8731−8733. (56) Prince, B. M.; Cundari, T. R. C−H Bond Activation of Methane by PtII−N-Heterocyclic Carbene Complexes. The Importance of Having the Ligands in the Right Place at the Right Time. Organometallics 2012, 31, 1042−1048. (57) Pahls, D. R.; Groves, J. T.; Gunnoe, T. B.; Cundari, T. R. Theoretical Study of Reductive Functionalization of Methyl Ligands of Group 9 Complexes Supported by Two Bipyridyl Ligands: A Key Step in Catalytic Hydrocarbon Functionalization. Organometallics 2014, 33, 1936−1944.

12290

DOI: 10.1021/acs.inorgchem.7b01736 Inorg. Chem. 2017, 56, 12282−12290