Catalytic Cycle for N–CN Bond Cleavage by Molybdenum Silyl

Apr 16, 2013 - ... Cited-by Linking service. For a more comprehensive list of citations to this article, users are encouraged to perform a search inSc...
0 downloads 0 Views 3MB Size
Download PDF
Article pubs.acs.org/Organometallics

Catalytic Cycle for N−CN Bond Cleavage by Molybdenum Silyl Catalyst: A DFT Study AbdelRahman A. Dahy,†,‡ Nobuaki Koga,*,‡ and Hiroshi Nakazawa§ †

Department of Chemistry, Faculty of Science, Assiut University, Assiut 71516, Egypt Graduate School of Information Science, Nagoya University, Nagoya 464-8601, Japan § Department of Chemistry, Graduate School of Science, Osaka City University, Sumiyoshi-ku, Osaka 558-8585, Japan ‡

S Supporting Information *

ABSTRACT: Using hybrid density functional theory calculations with the B3LYP functional, the reaction mechanisms for cleavage of R2N−CN (R = H, Me) bonds in the presence of unsaturated molybdenum(II) silyl catalyst, Cp(CO)2MoSiMe3 (Cp = η5-C5H5), were studied. The catalytic cycle takes place in two stages; the first involves cleavage of the R2N−CN bond. The favorable sequence of reactions for this stage is as follows: (i) coordination of a nitrile through the lone pair of electrons on the nitrile nitrogen atom (NCN) to give an end-on complex; (ii) isomerization of the end-on complex to a side-on complex; (iii) migration of the silyl group to NCN to form a stable Mo−C−NCN three-membered-ring intermediate with an Mo−NCN dative bond; (iv) dissociation of NCN from Mo and coordination of an amino N atom (NNR2) to Mo, leading to an Mo− C−NNR2 three-membered-ring intermediate; and (v) cleavage of the R2N−C bond to form a silylisocyanide complex. The second stage involves the regeneration of the active catalyst through two σ-metathesis steps. In the first, Cp(CO)2MoNR2 reacts with HSiMe3 to give Cp(CO)2MoH and R2NSiMe3, and in the second, σ-metathesis of Cp(CO)2MoH with HSiMe3 regenerates Cp(CO)2MoSiMe3. Step (iv) in the first stage possesses the largest activation energy and is the rate-determining step. The activation energies for this step for the reactions of H2NCN and Me2NCN were calculated to be 36.4 and 38.3 kcal/mol, respectively, based on potential energies with zero-point energy correction. After dissociation of the silylisocyanide ligand from the silylisocyanide complex, it will be isomerized to silylcyanide, as in previous studies. The catalytic cycle for the cleavage of R2N−CN bond is compared with that of MeO−CN bond. The effects of the metal atoms are also discussed. cleaved in the SiMI reaction.4d,9,10 The positively charged silyl group migrates to the negatively charged nitrile N atom (NCN) in the side-on complex B to form the M−C−NCN threemembered-ring complex C. C rearranges to form M−C−E three-membered-ring complex D in the cases of E = NH2, NMe2, and MeO. In D, the strain in the three-membered ring activates the C−E bond. Nakazawa et al. recently realized a catalytic reaction for O− CN bond cleavage using Mo complexes.9 A mechanism for the catalytic cycle was proposed based on experimental and theoretical results. The calculated energy profile for this catalytic cycle is shown in Figure 1. In the model reaction of Cp(CO)2MoSiMe3 (Cp = η5-C5H5), 1, with methylcyanate MeO−CN, the O−CN bond is cleaved by the SiMI reaction to form silylisocyanide complex 6a, and then the active catalyst 1 is regenerated by two successive σ-metathesis reactions. The first reaction is that of methoxy complex 7a with HSiMe3 to give hydride complex 9 and methyl silyl ether. In the second reaction, 9 reacts with a second HSiMe3 to regenerate 1. The

1. INTRODUCTION In the last thirty years, a number of experimental and theoretical studies focusing on cleavage of C−H,1−3 C−C,4 C−N,5 and C−O6 bonds in the presence of transition metal complexes have been reported. This is because cleavage of these inert bonds can produce new species that are useful in organic synthesis. One such unreactive bond is the R2N−CN bond in cyanamides, which is strengthened through conjugation, R2N−CN ↔ R2N+CN−. The first experimental cleavage of this bond was realized using the von Braun reaction, but the reaction required harsh conditions.7 In a recent study of the reaction of cyanamides with triethylsilane in the presence of Fe or Mo complexes, Nakazawa et al. found that Et3SiCN is formed, revealing that the R2N−CN bond is cleaved.8 They envisioned a catalytic cycle using these complexes, but found that catalytic reactions were only realized using Mo complexes under photochemical and thermal conditions, allowing them to propose a mechanism that includes the silyl-migration-induced (SiMI) reaction in Scheme 1.8 We have clarified in the study of E−CN (E = Me, NH2, NMe2, and MeO) bond cleavage mediated by Fe or Mo silyl complexes how an E−CN σ-bond is © XXXX American Chemical Society

Received: March 4, 2013

A

dx.doi.org/10.1021/om400179y | Organometallics XXXX, XXX, XXX−XXX

Organometallics

Article

Scheme 1. E−CN Bond Cleavage Induced by Silyl Migration8−10

Figure 1. Energy profile for the catalytic cycle for NC−O bond cleavage of methyl cyanate by Cp(CO)2MoSiMe3, 1.9 Depicted values are potential energies corrected by zero-point energies and Gibbs free energies (in italics) in kcal/mol relative to 1 + methyl cyanate + 2HSiMe3. 8a′ is a weak complex of 7a and HSiMe3. the same method as in previous studies.9,10 The reactions of cyanamides were carried out in toluene in the experiments, and the solvation effects are small, as confirmed in the calculations of the reactions with Fe complexes.10 Accordingly, we did not take solvation effects into account in the present study. Normal coordinate analysis was performed for all stationary points to characterize transition states (TSs) and equilibrium structures. Therefore, the energy minimum structures reported in this paper show positive eigenvalues of the Hessian matrix, whereas TSs have one negative eigenvalue.17 Intrinsic reaction coordinate calculations near the TS region followed by geometry optimization of both reactants and products were performed for all of the reactions to confirm the connectivity of the TS.18 Unscaled vibrational frequencies were used to calculate zero-point energy (ZPE) corrections to the total energy. The Gibbs free energies were also calculated at a temperature of 298.15 K and pressure of 1.00 atm. Unless otherwise noted, we discuss the energy profile obtained using ZPE-corrected potential energies. The program MOLEKEL was used to draw all of the optimized structures.19 In the next section, we discuss the differences in the energy profiles between cyanamides and cyanates using the bond dissociation energies (BDEs) calculated with the above methods. To validate the discussion using those BDEs, we calculated the BDEs using coupled cluster methods with single and double excitations and a perturbative treatment of triple excitations (CCSD(T)).20 Though the differences to some extent are observed, the B3LYP calculations gave the qualitatively correct trend in the BDEs. It is enough in this study to investigate qualitatively the effects of the substituents. Therefore, in the text, for consistency, we used the BDEs evaluated at the B3LYP level of calculations. The results of the BDEs calculated with the CCSD(T) methods are summarized in the Supporting Information.

rate-determining step is the dissociation of NCN from the Mo atom in 4a. In this paper, we theoretically investigated the catalytic cycle for cleavage of the N−CN bond in cyanamides catalyzed by Mo silyl complexes. One aim of the present study is to elucidate (i) whether the reaction mechanism for the complete catalytic cycle in the cases of cyanamides with a molybdenum silyl catalyst is similar to that for O−CN bond cleavage in Figure 1. The stability of intermediates such as 4a and the activation energy of the rate-determining step are also investigated. This study also aims to elucidate (ii) how the NH2, NMe2, and MeO groups attached to a cyanide group affect the reaction profiles and (iii) differences in the reaction mechanisms for the formation of silylisocyanide complexes using Mo and Fe complexes.

2. COMPUTATIONAL METHODS Molecular geometries were optimized at the density functional theory (DFT) level of theory using the B3LYP hybrid functional. The Gaussian 03 and Gaussian 09 suites of program were used in the calculations.11 The B3LYP method takes electron correlation effects into account because it consists of Becke’s three-parameter hybrid exchange functional and the nonlocal correlation functional of Lee, Yang, and Parr.12,13 For Mo, the basis set LANL2DZ was used.14 In this basis set, the 28 innermost electrons of Mo are replaced by the effective core potential of Hay and Wadt.14 For other atoms, the 631G(d) basis set was selected.15 In all the calculations we used five pure d functions. This method gave reasonable results in the study of O−CN bond cleavage by a Mo silyl catalyst, with which we compare the results obtained here.9 The calculations of the Fe complexes mentioned in Section 1 were performed using the same method, and comparison with the results obtained using the 6-311G(d,p) basis set showed that the basis set effect is small.10,16 Therefore, here we used

3. RESULTS AND DISCUSSION To investigate the catalytic cycle for cleavage of the N−CN bond in cyanamides catalyzed by Mo silyl complexes, we B

dx.doi.org/10.1021/om400179y | Organometallics XXXX, XXX, XXX−XXX

Organometallics

Article

Scheme 2. Proposed Catalytic Cycle Studied in This Work

Scheme 3. Other Conceivable Reaction Mechanisms for (I) Cleavage of the R2N−CN Bond and (II) the Regeneration of the Active Catalyst Cp(CO)2MoSiMe3

by Mo silyl catalysts.9 R2N−CN (R = H, Me) were used as models of cyanamides in Scheme 2. Because replacement of Et3Si in experiments with Me3Si in calculations does not cause

theoretically studied the catalytic cycle shown in Scheme 2 for the reaction of 1 with R2NCN (R = H, Me) and HSiMe3, which is a cycle similar to that for cleavage of O−CN bonds mediated C

dx.doi.org/10.1021/om400179y | Organometallics XXXX, XXX, XXX−XXX

Organometallics

Article

Figure 2. Optimized structures in the pathway for the formation of silylisocyanide complex Cp(CO)2Mo(NH2)(CNSiMe3), 6b, from 1 and H2NCN. All bond lengths are in Å.

bond and the resulting product will depend on the orientation of R2N−CN. While we theoretically studied these pathways for the Fe silyl complex to exclude them,10 here we compare these possibilities with the pathway in Scheme 2. In stage (2), after the dissociation of the silylisocyanide ligand from 6x giving Cp(CO)2MoNR2, 7x, there are two conceivable pathways for regeneration of the active catalyst 1 when 7x reacts with Me3SiH. The first possibility (shown in Scheme 2) involves oxidative addition of HSiMe3 to 16electron complex 7x, resulting in formation of complex 8x. Reductive elimination of R2NSiMe3 from 8x produces complex 9. Oxidative addition of HSiMe3 to 9 gives complex 10, and reductive elimination of H2 in 10 regenerates active catalyst 1. This oxidative addition/reductive elimination sequence can be alternatively explained by considering σ-bond metathesis. It is proposed that σ-bond metathesis takes place through complexes 8x′ and 8x″ instead of 8x, and 10′ and 10″ instead of 10. In the second possible pathway for stage (2) (Scheme 3(II)), oxidative addition of HSiMe3 to complex 7x gives complex 13x followed by reductive elimination of RNH2 to regenerate the active catalyst 1. The alternative σ-bond metathesis pathway passes through complexes 13x′ and 13x″ instead of 13x. Here, we will compare these possibilities with the pathway in Scheme 2. To allow clear comparison of these results with the reactions of Fe complexes, we separately present the results for stage (1),

any obvious change in the results of calculations, HSiMe3 was used instead of HSiEt3 in calculations following previous theoretical studies.4d,9,10 The catalytic cycle in Scheme 2 consists of two stages. In stage (1), R2N−CN coordinates to the Mo atom in 1 via the lone pair of electrons on the NCN atom to give end-on σcomplex 2x (x = b and c for R = H and Me, respectively), which isomerizes to side-on π-complex 3x. In the next step, the silyl group migrates from Mo to NCN to form Mo−C−NCN three-membered-ring complex 4x. Cleavage of the Mo−NCN bond and formation of the Mo−NNR2 bond follow to produce Mo−C−NNR2 three-membered-ring complex 5x. Then, the R2N−CN bond is cleaved, leading to silylisocyanide complex 6x. The silylisocyanide Me3SiNC formed in 6x isomerizes to more stable silylcyanide Me3SiCN. The previous studies of E− CN (E = Me, H2N, Me2N, and MeO) bond activation in the presence of Fe silyl complex and MeO−CN bond activation in the presence of Mo silyl catalyst have shown that, after dissociation from the metal atom, silylisocyanide isomerizes to silylcyanide.4d,9,10 In the present study, this isomerization inside and outside the coordination sphere of the Mo catalyst will be compared with that in a free molecule. As shown in Scheme 3(I), there are other possibilities for stage (1) involving cleavage of the R2N−CN bond, starting with the coordination of R2NCN via the lone pair of electrons on the amino N atom (NNR2) to the unsaturated Mo atom in active catalyst 1. The activation energy for the cleavage of this D

dx.doi.org/10.1021/om400179y | Organometallics XXXX, XXX, XXX−XXX

Organometallics

Article

Figure 3. Energy profiles for the catalytic cycle for NC−N bond cleavage of (a) H2NCN and (b) Me2NCN by Cp(CO)2Mo(SiMe3), 1. Depicted values are potential energies corrected by zero-point energies and Gibbs free energies (in italics) in kcal/mol relative to (a) 1 + H2NCN + 2HSiMe3 and (b) 1+ Me2NCN + 2HSiMe3. In (a), 8b′ is a weak complex of 7b with HSiMe3, and in (b), 8c′ is a weak complex of 7c with HSiMe3.

coordination in 2b1 to weaker π-coordination in 3b is endothermic and requires an activation energy of 10.7 kcal/ mol, as shown in Figure 3(a). The change in the length of the CN triple bond during the reaction is similar to that in the reactions of the Fe complex; while upon end-on coordination it does not change, the back-donation to the π* orbital in 3b elongates this bond by 0.053 Å (the localized orbitals of the Fe and Mo side-on complexes that demonstrate the back-donation are shown in Figure S7). In the next step, the migration of the silyl group to NCN transforms 3b to η 2 -amidino complex 4b via TS 3b−4b accompanied by transformation of the CN triple bond to a double bond. The NH2N−C−N−Si and NH2N−C−N−Mo dihedral angles in 4b are 25.8° and 179.1°, respectively. Although the former angle is slightly deviated from planarity, both angles suggest the conjugation of the amino lone pair electrons with the CN bond, which makes the H2N−C fragment close to coplanar with an H−NH2N−CCN−H dihedral angle of −171.9°. This N-silylated η2-amidino complex 4b with a coordinated NCN lone pair can be considered a Mo(II) 18electron species. The activation energy for the above step is small, just 0.8 kcal/mol. The low activation barrier for this type of silyl migration is ascribed to the large exothermicity of this step (−38.6 kcal/mol in the present reaction), hypervalent character of the Si atom, and significant charges on the NCN and Si

formation of silylisocyanide complex 6, and those for stage (2), the steps needed to regenerate active complex 1. 3.1. R2N−CN Bond Cleavage and Formation of Silylisocyanide Complex 6x. 3.1.1. H2N−CN. The steps for the formation of silylisocyanide from cyanamides using a Mo complex as a catalyst are similar to those for the Fe complex. A detailed comparison will be made later in Section 3.5. The optimized structures of the intermediates and TSs and the associated energy profiles for the reaction pathway of H2N−CN bond cleavage to form the silylisocyanide complex are shown in Figures 2 and 3(a), respectively. The model active catalyst 1 is an unsaturated Mo(II) 16-electron complex and its vacant site is occupied by the agostic CH bond of one of the methyl groups in the SiMe3 ligand (Figure 2),9 similar to the active Fe complex Cp(CO)FeSiMe3.4d,10 The C−NH2N bond distance of 1.348 Å in the free H2N−CN molecule is shorter than the standard C−N single-bond distance (for instance, 1.455 Å for NMe3 at the present level of calculation),10 because of conjugation of the lone pair electrons with the CN π* orbital. However, this conjugation is not strong enough to make the H2N plane and C−N bond coplanar; the angle between them is 138.6°. The reaction first passes through complex 2b1 with end-on σ-coordination via the lone pair of electrons on the NCN atom, and then 2b1 isomerizes to 3b in the side-on π-coordination mode of H2N−CN through TS2b1−3b. Transformation of σE

dx.doi.org/10.1021/om400179y | Organometallics XXXX, XXX, XXX−XXX

Organometallics

Article

atoms.4d,9,10 The Mulliken charges on the Si and NCN atoms of 3b are +0.875 and −0.376, respectively. Therefore, the positively charged Si atom may be strongly attracted by negatively charged NCN to form a strong bond. The Si−NCN bond in 4b is much stronger than the Mo−Si bond in 3b that is broken during the reaction (D0(Mo−Si) = 28.2 kcal/mol in 3b and D0(Si−NCN) = 87.6 kcal/mol in H2NCHNSiMe3). This bond alternation results in the large exothermicity. As in other SiMI reactions, the C−NH2N bond has not yet been activated in the η2-amidino complex 4b;9,10 the C−NH2N bond distance in 4b is 1.338 Å, which is 0.010 Å shorter than that in free H2NCN. In the next step, 4b is transformed into another η2-amidino complex, 5b1, via TS4b‑5b1 (Figure 2). During this reaction, the dative Mo−NCN bond is broken, and an Mo−NH2N bond is formed. However, this bond alternation is not synchronous; the Mo−NCN bond is broken before TS4b‑5b1 is reached, and rotation around the C−NH2N bond begins after passing TS4b‑5b1 so that the lone pair on the NH2N atom coordinates to the Mo atom. At TS4b‑5b1, while Mo−NCN bond cleavage has proceeded by 64.3% (=(2.888−2.269)/(3.232−2.269) × 100), Mo−NH2N bond formation is only 18.9% (=(3.107−3.307)/(2.251−3.307) × 100) complete. The activation energy of 36.4 kcal/mol for this step is large because it involves cleavage of the strong Mo− NCN bond with a distance of 2.269 Å (see Section 3.4.2). A similarly large activation energy (29.9 kcal/mol) was theoretically obtained for the same reaction with the Fe complex, which supports the experimental findings that the Fe analogue of 4b was isolated and that it yielded silylcyanide upon heating.8 The C−NH2N bond in 5b1 (1.488 Å) is 0.150 Å longer than that in 4b, partly because the conjugation between the lone pair and CN double bond is broken. This bond is even longer than a standard C−N single bond, which suggests that C−NH2N bond activation occurs during coordination of the lone pair to the Mo atom, as found in the reaction with the Fe complex. This will be discussed in detail in Section 3.4.3. The last step of C−NH2N bond cleavage is the reaction of 5b1 to 6b via TS5b1−6b. The C−NH2N bond is activated in 5b1 and is easily broken, with a small activation energy of 5.4 kcal/ mol. There is a trans isomer of 5b1, called 5b2, as shown in Figure S4(b) in the Supporting Information, which has a trans NH2N− CN−Si structure and is 3.3 kcal/mol more stable than 5b1 with a cis NH2N−CN−Si structure. However, no reaction pathway was found for NH2N−C bond cleavage from 5b2. The NH2N−C bond distance in 5b1 (1.488 Å) is longer than that in 5b2 (1.472 Å), which suggests that the NH2N−C bond is more activated in 5b1 than in 5b2. This difference is presumably caused by an effect similar to the anomeric effect between the NCN lone pair and C−NH2N σ* orbital in 5b1. This is why bond cleavage passes through 5b1 and not through more stable 5b2. The other possibility for H2N−CN bond cleavage starts with coordination via the lone pair of electrons on the NH2N atom, as shown in Scheme 3(I). The structures of the corresponding complexes, 2b2 and 2b3 (Figure 4), were determined, revealing that 2b2 and 2b3 are less stable than 2b1 by 15.4 and 12.5 kcal/mol, respectively. The coordination of H2NCN via the lone pair of electrons on NH2N is weaker than that via the lone pair of electrons on NCN, similar to the Fe complexes.10 The possible reactions from 2b2 and 2b3 found in the calculations are oxidative addition and σ-metathesis, respectively, both of which require large activation energies of 33.6 and 41.3 kcal/ mol, respectively. Accordingly, the reaction with a smaller

Figure 4. Optimized structures of 2b2 and 2b3 in the pathways for the formation of (a) Cp(CO)2Mo(NH2)(CN)(SiMe3), 11b, and (b) Cp(CO)2Mo(CN)(NH2SiMe3), 12b, from 1 and H2NCN. All bond lengths are in Å. Depicted values are potential energies corrected by zero-point energies and Gibbs free energies (in italics) in kcal/mol relative to 1 + H2NCN.

activation energy starting from more stable 2b1 is more favorable, and the reaction pathways passing through 2b2 and 2b3 can be excluded. The optimized structures and relative energies for the transition states and products of these reactions are shown in Figure S1 and discussed in the Supporting Information. The present calculations for stage (1) in Scheme 2 show that the reaction pathway including the SiMI reaction is favorable for N−CN bond cleavage, and N-silylated η2-amidino complex 4b is a stable intermediate. 3.1.2. Me2NCN. In the study of the reaction of Me2NCN, we focused on the effects of substituents on the mechanism. Accordingly, here we mainly discuss the differences between the reactions of H2NCN and Me2NCN. The energy profile for the entire reaction pathway involving Me2NCN is shown in Figure 3(b). In cleavage of the Me2N−CN bond, the optimized structures for isomerization of the σ-complex to π-complex and the migration of a silyl group from the Mo atom to NCN are similar to those in the case of H2N−CN (Figure 2), as shown in Figure S2(a) in the Supporting Information. The activation energies of these steps with Me2NCN are also close to those with H2NCN (14.6 and 0.0 kcal/mol compared to 10.7 and 0.8 kcal/mol, respectively), as shown in Figure 3(b). The low activation energy for the formation of 4c is ascribed to the same reasons stated for 4b in the case of H2NCN.4d,9,10 Compared to the transformation of 4b to 5b1 via a single transition state (TS 4b‑5b1 ) (Figures 2 and 3(a)), the corresponding transformation in the case of Me2NCN (4c to 5c1) requires two steps (Figure 5). In the first step (4c to 4c′ via TS4c‑4c′), NCN dissociates from the Mo atom. In 4c′, the short distance of 2.123 Å between the Mo atom and one of the methyl H atoms and long C−H bond distance of 1.121 Å are indicators of an agostic interaction between the amino Me group and Mo atom. This interaction can take place because the Mo atom in this complex is unsaturated. In this step, the NMe2N−CN−Si structure also changes from cis to trans, which is induced by the steric repulsion between the methyl groups on the Si and amino N atoms. This first step requires a large activation energy of 38.3 kcal/mol because it involves cleavage of the strong Mo−NCN bond. The second step is coordination of the amino N atom (NMe2N), keeping the trans NMe2N−CN−Si structure (4c′ to 5c1 via TS4c′‑5c1). During the course of this step, the agostic interaction is cleaved by the rotation of the Me2N group around the NMe2N−C bond and the formation of the Mo−NMe2N bond. Because the agostic interaction is weak, the activation energy of this step is F

dx.doi.org/10.1021/om400179y | Organometallics XXXX, XXX, XXX−XXX

Organometallics

Article

Figure 5. Optimized structures for rearrangement of 4c to 6c1 in the pathway for the formation of silylisocyanide complex 6c1 from 1 and Me2NCN. All bond lengths are in Å.

Me2NCN passing through an intermediate with an agostic interaction. 3.2. Isomerization of Silylisocyanide to Silylcyanide As a Free Ligand and in the Coordination Sphere of Mo. In experiments, silylcyanide was observed as a product.8,9 Thus, CNSiMe3 formed in the reactions mentioned above must isomerize to more stable NCSiMe3. There are two possible mechanisms for this isomerization. The first one is that CNSiMe3 dissociates from silylisocyanide complexes 6b and 6c1 and then isomerizes to NCSiMe3. The second possibility is that CNSiMe3 isomerizes to NCSiMe3 in the coordination sphere of the Mo complex. The isomerization of free CNSiMe3 has been previously studied, and the activation energy was calculated to be 23.8 kcal/mol.4d The dissociation of CNSiMe3 from 6b and 6c1 requires energies of 13.2 and 7.7 kcal/mol, respectively (Figures S4(a) and S5 in the Supporting Information). The corresponding changes in the Gibbs free energy are −0.1 and −5.7 kcal/ mol, respectively. Accordingly, dissociation occurs readily. Our trial to find the TS for the second possibility gave a TS structure with Mo−NCN distances longer than 4.5 Å, indicating isomerization outside the coordination sphere of the Mo atom, and activation energies obtained are almost the same as those in isomerization of the free ligand (Figures S4(a) and S5). According to these results, isomerization of free CNSiMe3 to NCSiMe3 is more favorable than its isomerization in the coordination sphere of the Mo complex. As a matter of fact, we could determine the TS structure for the isomerization in the coordination sphere in the case of less bulky H2NCN; it has Mo−NCN and Mo−CCN distances of 2.647 and 2.434 Å and requires a much larger activation energy of 53.0 kcal/mol (Figure S4(a)). 3.3. Completion of the Catalytic Cycle and Regeneration of Active Catalyst 1. To complete the catalytic cycle, the active catalyst Cp(CO)2MoSiMe3, 1, must be regenerated from amino complexes 7b and 7c after dissociation of the silylisocyanide ligand. We studied the possible reaction pathways shown in Schemes 2 and 3(II) for both H2NCN and Me2NCN to find that there is no reaction pathway through oxidative addition/reductive elimination. For H2NCN, the

relatively small (3.9 kcal/mol). The structure of TS4c′‑5c1 suggests that, once the agostic interaction is broken, the Me2N lone pair readily coordinates to the Mo atom. Figure 3(b) shows that 4c′ without Mo−NCN or Mo−NMe2N dative bonds is an unstable intermediate, and the activation energy for the transformation of 4c′ to 5c1 is small (3.9 kcal/mol). Accordingly, the first step of 4c to 4c′ involving dissociation of NCN from Mo requires a larger activation energy than the second step. The subsequent cleavage of the C−NMe2N bond in 5c1 to form silylisocyanide complex 6c1 takes place via TS5c1−6c1 with an activation energy of 11.5 kcal/mol, as shown in Figures 5 and 3(b). This activation energy is larger than that for H2N− CN bond cleavage in the present study (5.4 kcal/mol) and those for H2N−CN and Me2N−CN bond cleavage by the Fe silyl complex (1.8 and 2.4 kcal/mol).10 Comparison of 5b1 and 5b2 reveals that the N−CN bond is more activated in the intermediate with a cis NR2N−CN−Si structure (5b1, 5c2). In fact, as shown in Figure S2(b), the N−CN bond of 5c2 with a cis NR2N−CN−Si structure (1.480 Å) is longer than that of 5c1 (1.463 Å). However, 5c2 is 5.1 kcal/mol less stable than 5c1, and TS5c2−6c2 for NMe2N−CN bond cleavage from 5c2 is 2.3 kcal/mol less stable than TS5c1−6c1. Accordingly, the reaction from 5c1 is more favorable than that from 5c2. Note that, as shown in Figure S2(b), the reaction through TS5c2−6c2 produces undesirable silylisocyanide complex 6c2 with a bond between the carbonyl C and NMe2N atoms. The optimized structures and relative energies for the pathway via cis isomer 5c2 are shown in Figure S2(b). We also studied the reaction pathways that start with coordination via the lone pair of electrons on NMe2N, as presented in Scheme 3(I). Calculations showed that, as in the case of H2N−CN, such reaction pathways are unfavorable; the coordination of Me2NCN via the lone pair of electrons on NMe2N is weaker than that via the lone pair of electrons on NCN, and the activation energies are higher, as shown in Figure S3. Overall, the reaction of Me2NCN shown in Figure 3(b) is very similar to that of H2NCN except the step involving isomerization of three-membered-ring intermediate 4 to 5. The difference in isomerization is caused by the reaction of G

dx.doi.org/10.1021/om400179y | Organometallics XXXX, XXX, XXX−XXX

Organometallics

Article

Figure 6. Optimized structures in the pathway for regeneration of active catalyst 1 from complex 7b in Scheme 2. All bond lengths are in Å.

optimized structures included in the σ-metathesis pathway in Scheme 2 and its relative energy profile are shown in Figures 6 and 3(b), respectively. As shown in Figure 6, replacement of the amino group by the silyl group in this pathway is carried out by σ-metathesis of 7b with HSiMe3 to give complex 9 via complexes 8b′ and 8b″ and that of 9 with HSiMe3 to regenerate complex 1 through complexes 10′ and 10″. The first σ-metathesis takes place via TS8b′‑8b″ (Figures 6 and 3(b)). The reactant of this TS is the weak complex between 7b and HSiMe3, 8b′, which is slightly more stable than 7b and HSiMe3 (by 1.1 kcal/mol). The activation energy for this step is 16.9 kcal/mol. Dissociation of complex 8b″ to H2NSiMe3 and 9 requires an energy of 25.5 kcal/mol. The second σ-metathesis then takes place through TS10′‑10″ via the intermediates 10′ and 10″. The reactions after the formation of 9 are the same as those in Figure 1 and have already been reported.9 The complex of HSiMe3 with 9, 10′, is more stable than 9 and HSiMe3 by 18.4 kcal/mol. There is a strong interaction between HSiMe3 and 9 in 10′ through its hydrogen atom, and this results in activation of the H−Si bond in HSiMe3; the interacting H−Si bond in 10′ (1.589 Å) is longer than that in free HSiMe3 (1.496 Å). Therefore, σ-metathesis in this step to give complex 10″ requires a small activation energy of 5.2 kcal/ mol. Removing H2 from 10″ to regenerate complex 1 requires an energy of just 3.6 kcal/mol. After taking entropy into consideration, the activation free energies in this pathway are 20.6 and 6.3 kcal/mol and the free energies of dissociation are 13.0 and −3.4 kcal/mol. The optimized structures in the pathway depicted in Scheme 3(II) and their relative energies are shown in Figure S6(a) and (c), respectively, in the Supporting Information. Again, no reaction pathway through oxidative addition/reductive elimination was found. An activation energy of 33.4 kcal/mol for σmetathesis of 7b with HSiMe3 is much larger than those in the pathway in Scheme 2. Bond formation between the negatively charged H atom of hydrosilane and the nitrogen lone pair is difficult, which is the reason for this high activation energy. Removing NH3 from complex 13b″ requires an energy of 17.0

kcal/mol. The activation free energy is 34.3 kcal/mol, and the free energy of dissociation of NH3 is 6.0 kcal/mol. These results demonstrate that the σ-metathesis pathway in Scheme 2 to complete the catalytic cycle, the regeneration of the active catalyst 1 from 7b and two molecules of HSiMe3, is more favorable than the other. In the case of Me2NCN, there were two σ-metathesis pathways (Schemes 2 and 3(II)) identified for the regeneration of active catalyst 1. The σ-metathesis pathway outlined in Scheme 2 is more favorable, similar to the case of H2NCN. The optimized structures for the two pathways are shown in Figure S6(b) in the Supporting Information, and their energy profiles are shown in Figures 3(b) and S6(c) (Supporting Information). In stage (2) of regenerating the active intermediate, the favorable pathway involves σ-metathesis with two molecules of HSiMe3. The amino group and the hydrogen atom in hydrosilane are removed as a silylamine and hydrogen molecule, respectively. As a whole, the catalytic cycle proposed in Scheme 2 is favorable. The energy profiles depicted in Figure 3 show that dissociation of NCN in 4b and 4c requires the largest activation energy in the whole catalytic cycle, and thus it is the rate-determining step. Intermediates similar to 4b and 4c have been isolated experimentally and identified by X-ray crystallography in the stoichiometric reactions of cyanamides with Fe silyl complexes and those of cyanates with W silyl complexes.8,9 3.4. Effects of Substituents (H2N, Me2N, and MeO) Attached to CN. 3.4.1. Rearrangement of End-on Nitrile Complex 2x1 to Side-on Nitrile Complex 3x. The rearrangement of the end-on nitrile complex to the side-on one is sensitive to the electronic effects of substituents. We have found in the study of the reactions with Fe complexes that a more inductively electron-withdrawing methoxy group enhances back-donation from the Fe atom to the CN π* orbital especially in the side-on complex, which relatively stabilizes the side-on complex.10 This is also observed in the present Mo complexes. The rearrangements of H2N−CN and Me2N−CN complexes are endothermic by 7.3 and 10.1 kcal/mol, H

dx.doi.org/10.1021/om400179y | Organometallics XXXX, XXX, XXX−XXX

Organometallics

Article

interpreted using the dissociation energies of E−CN and E− SiMe3 bonds. The energies for E−CN bonds were presented above in Section 3.4.4, whereas those for E−SiMe3 bonds were calculated to be 105.7, 97.9, and 81.8 kcal/mol for MeO− SiMe3, H2N−SiMe3, and Me2N−SiMe3, respectively. Because the MeO−SiMe3 bond is stronger than the H2N−SiMe3 and Me2N−SiMe3 bonds, the catalytic cycle for MeOCN is more exothermic than that for cyanamides. 3.5. Comparison of Cleavage of N−CN and O−CN Bonds in the Presence of Mo Silyl or Fe Silyl Complexes. The binding energies of nitriles in both the end-on and side-on coordination modes are larger in the Fe complexes than in Mo ones. For instance, while the binding energy of H2NCN in the end-on Fe complex is 24.6 kcal/mol,10 that in the Mo complex is 17.5 kcal/mol. Similarly, the binding energy of silylisocyanide is higher in the Fe complex than in the Mo complex. For example, the binding energy of Me3SiNC to Cp(CO)FeNH2 is 15.8 kcal/mol, and that to Cp(CO)2MoNH2 is 13.2 kcal/mol. These larger binding energies make the relative energies of the intermediates and TSs in the reactions of the Fe complexes more negative. Presumably, stability induced by the larger negative energies could be one of the factors that made the catalytic cycle with the Fe complexes occur less readily than that with Mo complexes. The activation energies for the rate-determining step in the cases of the Fe silyl complexes are smaller than those for the Mo silyl complexes (29.9, 28.0, and 19.1 kcal/mol for Fe compared to 36.4, 38.3, and 26.6 kcal/mol for Mo). Because NCN dissociation is a dominant structure change in the ratedetermining step, this difference in the activation energies suggests that, different from the stronger bonds in the Fe complexes discussed in the last paragraph, the Mo−NCN bond is stronger than the Fe−NCN bond in the M−C−NCN threemembered-ring intermediates. As a matter of fact, at the TSs for the reactions of cyanamides, the weaker Fe−NCN bond is cleaved more quickly than the Mo−NCN bond; the cleavage of the Fe−NCN bond proceeds at the TSs by 76.3% (E = H2N) and 83.8% (E = Me2N), whereas that of the Mo−NCN bond proceeds by 64.3% (E = H2N) and 63.0% (E = Me2N). One notices that the methyl groups in cyanamides affect the difference in the activation energies between Fe and Mo; that for the reaction with E = Me2N (10.3 kcal/mol) is larger than that for the reaction with E = H2N (6.5 kcal/mol). It can be observed that the activation energy for the transformation of 5b1 to 6b (5.4 kcal/mol) is higher than that for the corresponding transformation in the case of the Fe silyl complex (1.8 kcal/mol). This is because the NH2N−C bond in 5b1 is not as activated as in the case of the Fe silyl complex (while the NH2N−C bond in 5b1 is stretched by 0.150 Å compared with that in 4b, the same bond in the Fe silyl complex is stretched by 0.187 Å, as discussed above in Section 3.4.3).

respectively, whereas that of MeO−CN is endothermic by only 4.2 kcal/mol. 3.4.2. Dissociation of NCN in Mo−C−NCN Three-Membered-Ring Complex 4x. In the reactions of the Fe complexes, the activation energy to dissociate the Fe−NCN bond in the Fe−C−NCN three-membered-ring complex in the case of MeOCN is lower compared with the cases of R2NCN (R = H and Me). Similarly, the activation energy of 26.6 kcal/mol for the dissociation of NCN of 4a leading to 5a is smaller than those in the reactions of 4b and 4c (36.4 and 38.3 kcal/mol, respectively). Although in the reactions of cyanamides these rearrangements must be accompanied by the rotation of the amino groups, rotation has not started at TS4b‑5b1 and TS4c‑4c′ and the amino N atom cannot approach the Mo atom. At TS4a−5a for cyanate, the O lone pair can readily interact with the Mo atom, resulting in a shorter Mo−O distance and lower activation energy; the Mo−O distance in TS4a−5a is shorter than that in 4a by 0.505 Å, whereas the Mo−NNR2 distances in TS4b‑5b1 and TS4c‑4c′ are shorter than those in 4b and 4c by only 0.200 and 0.163 Å, respectively. 3.4.3. C−E Bond Cleavage in Mo−C−E Three-MemberedRing Complex 5x. We have clarified in the study of the reactions of Fe complexes that ring strain in the Fe−C−NNR2 and Fe−C−O three-membered rings weakens the C−NNR2 and C−O bonds so these bonds cleave easily. This bond weakening is confirmed by the long distances of these bonds. As discussed above, these bonds are stretched in Mo−C−E threemembered-ring complex 5x. However, the degree of bond stretching in the Mo complexes is smaller than that in the Fe complexes. The NH2N−C bond in 5b1 is longer than that in 4b by 0.150 Å, while the analogue in the case of the Fe silyl complex is longer by 0.187 Å. Similarly, the O−C bond in 5a is stretched by 0.128 Å relative to that in 4a, whereas in the Fe complexes, the O−C bond is stretched by 0.165 Å. This is presumably because Mo−ligand distances are longer than Fe− ligand distances, so the interaction between the Mo atom and C−N and C−O bonds is weaker. 3.4.4. Energy for the Formation of Silylisocyanide Complexes. The energies of the reactions of MeOCN, H2NCN, and Me2NCN leading to silylisocyanide complexes were calculated to be −51.3, −38.6, and −32.0 kcal/mol, respectively. To analyze the difference in the energies of these reactions, we calculated the bond dissociation energies of MeO−CN, H2N−CN, and Me2N−CN bonds in the reactants and those of the Mo−OMe, Mo−NH2, and Mo−NMe2 bonds in the products. The values obtained are 106.7, 116.9, and 106.8 kcal/mol for MeO−CN, H2N−CN, and Me2N−CN bonds, respectively, and 42.2, 39.7, and 23.1 kcal/mol for the Mo−OMe, Mo−H2N, and Mo−Me2N bonds, respectively. It is clear that the reaction of MeOCN in which the weaker bond is broken and the stronger bond is formed is more exothermic. Silylisocyanide complex 6a is more stable than 4a, unlike the reactions of cyanamides, where the silylisocyanide complexes are less stable than 4x. This is because a strong Mo−OMe bond is formed in 6a. 3.4.5. Energy of Catalytic Cycle. The reaction in the catalytic cycle is formally expressed as

4. CONCLUSIONS We studied the pathways for the reactions of the active catalyst Cp(CO)2MoSiMe3 with R2N−CN (R = H, Me) in which the N−CN bond is cleaved. The favorable pathways for the complete catalytic cycles in the cases of H2NCN and Me2NCN are similar to that in the case of MeOCN, and they are as follows: (i) both cyanamides coordinate to the active complex in an end-on coordination mode and then rearrange to the sideon coordination mode to facilitate silyl group migration to NCN; (ii) silyl group migration causes an Mo−C−NCN three-

1

E−CN + 2HSiMe3 → E−SiMe3 + Me3Si−CN + H 2 (E = H 2N, Me2N, MeO)

Accordingly, the difference in the energy of the reaction for the catalytic cycle between cyanamides and cyanate can be I

dx.doi.org/10.1021/om400179y | Organometallics XXXX, XXX, XXX−XXX

Organometallics

■

membered-ring intermediate to form; (iii) dissociation of NCN and coordination of NR2N to the Mo atom give three-membered -ring intermediate Mo−C−NR2N, in which the C−NR2N bond is activated by ring strain; (iv) cleavage of the R2N−CN bond takes place with a small activation energy to give a silylisocyanide complex; (v) after dissociation of silylisocyanide, the active catalyst Cp(CO)2MoSiMe3 is regenerated by two σmetathesis steps with two molecules of HSiMe3, yielding R2NSiMe3 and H2 molecules. The dissociation of NCN in (iii) requires the largest activation energy in the sequence of reactions and thus is the rate-determining step. Comparison of the catalytic cycles for the cleavage of N−CN and O−CN bonds in the presence of Mo silyl catalyst showed that they are qualitatively similar. However, because the Mo−O and Si−O bonds formed during the reactions of cyanate are stronger than Mo−N and Si−N bonds, and the O−CN bond is weaker than the N−CN one, the intermediates in the reaction of cyanate are more stable and the reaction is exothermic. The difference in the Mo−E bond energies is also reflected in the stability of the Mo−C−NCN three-membered-ring intermediates, 4x; these intermediates are the most stable for the cyanamides, whereas the silylisocyanide complex of cyanate, 6a, is more stable than 4a because 6a has a strong Mo−O bond. It was also found that the activation energy for the ratedetermining step of the reaction of MeOCN is smaller than those for H2NCN and Me2NCN (26.6, 36.4, and 38.3 kcal/ mol, respectively). This is because at the TS for this step, the O atom readily interacts with the Mo atom, whereas at the TSs for R2NCN, the amino groups do not rotate and hinder the approach of the lone pair of electrons on the amino N atom to the Mo atom. Comparison of the cleavage of N−CN and O−CN bonds by Fe silyl and Mo silyl complexes revealed that the activation energies for the rate-determining step in the cases of the Mo complexes are larger than those for the Fe complexes. This suggests that the Mo−NCN bonds in the Mo−C−NCN threemembered-ring intermediates, 4b and 4c, are stronger than the Fe−NCN bonds. Also, the activation energies in the N−CN bond cleavage step for the Mo complexes are slightly larger than those for the Fe complexes because of the longer Mo− ligand distances.

■

REFERENCES

(1) (a) Shilov, A. E. Activation of Saturated Hydrocarbons by Transition Metal Complexes; D. Reidel: Boston, MA, 1984. (b) Crabtree, R. H. Chem. Rev. 1985, 85, 245. (c) Ryabov, A. D. Chem. Rev. 1990, 90, 403. (d) Bergman, R. G. J. Organomet. Chem. 1990, 400, 273. (e) Arndtsen, B. A.; Bergman, R. G.; Mobley, T. A.; Peterson, T. H. Acc. Chem. Res. 1995, 28, 154. (f) Gianotti, C.; Green, M. L. H. J. Chem. Soc., Chem Commun. 1972, 1114. (g) Crabtree, R. H.; Miheleic, J. M.; Quirk, J. M. J. Am. Chem. Soc. 1979, 101, 7738. (h) Jones, W. D.; Feher, F. J. J. Am. Chem. Soc. 1982, 104, 4240. (i) Jones, W. D.; Feher, F. J. Organometallics 1983, 2, 562. (j) Hoyano, J. K.; Graham, W. A. J. Am. Chem. Soc. 1982, 104, 3723. (k) Hoyano, J. K.; McAlster, A. D.; Graham, W. A. J. Am. Chem. Soc. 1983, 105, 7190. (l) Janowicz, A. H.; Bergman, R. G. J. Am. Chem. Soc. 1982, 104, 352. (m) Wax, M. J.; Kovac, C. A.; Bergman, R. G. J. Am. Chem. Soc. 1984, 106, 1121. (n) Hackett, M.; Whitesides, G. M. J. Am. Chem. Soc. 1988, 110, 1449. (o) Sakakura, T.; Sodeyama, T.; Sasaki, K.; Wada, K.; Tanaka, M. J. Am. Chem. Soc. 1990, 112, 7221. (p) Yamamoto, Y.; Al-Masum, M.; Asao, N. J. Am. Chem. Soc. 1994, 116, 6019. (2) (a) Saillard, J.-Y.; Hoffman, R. J. Am. Chem. Soc. 1984, 106, 2006. (b) Obara, S.; Kitaura, K.; Morokuma, K. J. Am. Chem. Soc. 1984, 106, 7482. (c) Koga, N.; Morokuma, K. J. Phys. Chem. 1990, 94, 5454. (d) Koga, N.; Morokuma, K. J. Am. Chem. Soc. 1993, 115, 6883. (e) Low, J. J.; Goddard, W. A. J. Am. Chem. Soc. 1986, 108, 6115. (f) Low, J. J.; Goddard, W. A. Organometallics 1986, 5, 609. (g) Blomberg, M. R. A.; Siegbahn, P. E. M.; Nagashima, U.; Wennerberg, J. J. Am. Chem. Soc. 1991, 113, 424. (h) Svensson, M.; Blomberg, M. R. A.; Siegbahn, P. E. M. J. Am. Chem. Soc. 1991, 113, 7076. (i) Blomberg, M. R. A.; Siegbahn, P. E. M.; Svensson, M. J. Am. Chem. Soc. 1992, 114, 6095. (j) Siegbahn, P. E. M.; Blomberg, M. R. A.; Svensson, M. J. Am. Chem. Soc. 1993, 115, 4191. (k) Blomberg, M. R. A.; Siegbahn, P. E. M.; Svensson, M. J. Phys. Chem. 1994, 98, 2062. (l) Siegbahn, P. E. M.; Blomberg, M. R. A. Organometallics 1994, 13, 354. (m) Siegbahn, P. E. M. Organometallics 1994, 13, 2833. (n) Siegbahn, P. E. M.; Svensson, M. J. Am. Chem. Soc. 1994, 116, 10124. (o) Siegbahn, P. E. M. J. Am. Chem. Soc. 1996, 118, 1487. (3) (a) Sakaki, S.; Ieki, M. J. Am. Chem. Soc. 1993, 115, 2373. (b) Sakaki, S.; Biswas, B.; Sugimoto, M. J. Chem. Soc., Dalton Trans. 1997, 803. (c) Sakaki, S.; Biswas, B.; Sugimoto, M. Organometallics 1998, 17, 1278. (d) Sakaki, S.; Biswas, B.; Sugimoto, M. J. Am. Chem. Soc. 2000, 19, 3895. (e) Murahashi, S.-I.; Hirano, T.; Yano, T. J. Am. Chem. Soc. 1978, 100, 348. (f) Murahashi, S.-I.; Watanabe, J. J. Am. Chem. Soc. 1979, 101, 7429. (g) Murahashi, S.-I.; Naota, T.; Taki, H.; Mizuno, M.; Takaya, H.; Komiya, S.; Mizuno, Y.; Oyasato, N.; Hiraoka, M.; Hirano, M.; Fukuoka, A. J. Am. Chem. Soc. 1995, 117, 12436. (h) Murahashi, S.-I.; Naota, T. Bull. Chem. Soc. Jpn. 1996, 69, 1805. (4) (a) Jun, C.-H. Chem. Soc. Rev. 2004, 33, 610. (b) Rybtchinski, B.; Milstein, D. Angew. Chem., Int. Ed. 1999, 38, 870. (c) Storsberg, J.; Yao, M.-L.; Ocal, N.; De Meijere, A.; Adam, A. E. W.; Kaufmann, D. E. Chem. Commun. 2005, 5665. (d) Nakazawa, H.; Kawasaki, T.; Miyoshi, K.; Suresh, C. H.; Koga, N. Organometallics 2004, 23, 117. (e) Chianese, A. R.; Zeglis, B. M.; Crabtree, R. H. Chem. Commun. 2004, 2176. (f) Terao, Y.; Wakui, H.; Satoh, T.; Miura, M.; Nomura, M. J. Am. Chem. Soc. 2001, 123, 10407. (g) Ateşin, T. A.; Li, T.; Lachaize, S.; Brennessel, W. W.; Garcia, J. J.; Jones, W. D. J. Am. Chem. Soc. 2007, 129, 7562. (h) Ateşin, T. A.; Li, T.; Lachaize, S.; Garcia, J. J.; Jones, W. D. Organometallics 2008, 27, 3811. (5) (a) Lei, Y.; Wrobleski, A. D.; Golden, J. E.; Powell, D. R.; Aubé, J. J. Am. Chem. Soc. 2005, 127, 4552. (b) Fran, L.; Yang, L.; Guo, C.; Foxman, B. M.; Ozerov, O. V. Organometallics 2004, 23, 4778. (c) Yao, M.-L.; Adiwidjaja, G.; Kanfmann, D. E. Angew. Chem., Int. Ed. 2002, 41, 3375. (d) Niklas, N.; Heinemann, F. W.; Hample, F.; Alsfasser, R. Angew. Chem., Int. Ed. 2002, 41, 3386. (6) (a) Maerker, A. Angew. Chem., Int. Ed. 1987, 26, 972. (b) Kabalka, G. W.; Yao, M.-L.; Borella, S.; Goins, L. K. Organometallics 2007, 26, 4112. (c) Szu, P.-h.; He, X.; Zhao, L.; Liu, H.-w. Angew. Chem., Int. Ed. 2005, 44, 6742. (d) Casado, F.; Pisano, L.; Farriol, M.; Gallardo, I.; Marquet, J.; Melloni, G. J. Org. Chem. 2000, 65, 322.

ASSOCIATED CONTENT

S Supporting Information *

The structures of some intermediates and transition states for the reactions of H2NCN and Me2NCN, dissociation energies for the E−CN, E−Mo, and E−SiMe3 bonds, localized orbitals demonstrating back-donation in side-on cyanamide complexes, and Cartesian coordinates of all the optimized structures are available free of charge via the Internet at http://pubs.acs.org.

■

Article

AUTHOR INFORMATION

Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS Some of the calculations were performed at the Research Center for Computational Science of Okazaki National Research Institutes, Japan, and the information technology center of Nagoya University. A.A.D. thanks Japan Student Services Organization for a Fiscal 2012 JASSO Follow-up Research Fellowship. J

dx.doi.org/10.1021/om400179y | Organometallics XXXX, XXX, XXX−XXX

Organometallics

Article

(18) (a) Gonzalez, C.; Schlegel, H. B. J. Chem. Phys. 1989, 90, 2154. (b) Gonzalez, C.; Schlegel, H. B. J. Phys. Chem. 1990, 94, 5523. (19) (a) Flükiger, P.; Lüthi, H. P.; Portmann, S.; Weber, J. MOLEKEL 4.3; Swiss Center for Scientific Computing: Manno, Switzerland, 2000−2002. (b) Portmann, S.; Lüthi, H. P. Chimia 2000, 54, 766. (20) Werner, H.-J.; Knowles, P. J.; Knizia, G.; Manby, F. R.; Schütz, M.; Celani, P.; Korona, T.; Lindh, R.; Mitrushenko, A.; Rauhut, G.; Shamasundar, K. R.; Adler, T. B.; Amos, R. D.; Bernhardsson, A.; Berning, A.; Cooper, D. L.; Deegan, M. J. O.; Dobbyn, A. J.; Eckert, F.; Goll, E.; Hampel, C.; Hesselmann, A.; Hetzer, G.; Hrenar, T.; Jansen, G.; Köppl, C.; Liu, Y.; Lloyd, A. W.; Mata, R. A.; May, A. J.; McNicholas, S. J.; Meyer, W.; Mura, M. E.; Nicklaß, A.; O’Neill, D. P.; Palmieri, P.; Peng, D.; Pflüger, K.; Pitzer, R.; Reiher, M.; Shiozaki, T.; Stoll, H.; Stone, A. J.; Tarroni, R.; Thorsteinsson, T.; Wang, M. MOLPRO version 2012.1; http://www.molpro.net: Cardiff, United Kingdom, 2012.

(7) Vliet, E. B. Organic Syntheses; Wiley: New York, Collect. Vol. 1, 1941; p 201. (8) Fukumoto, K.; Oya, T.; Itazaki, M.; Nakazawa, H. J. Am. Chem. Soc. 2009, 131, 38. (9) Fukumoto, K.; Dahy, A. A.; Oya, T.; Hayasaka, K.; Itazaki, M.; Koga, N.; Nakazawa, H. Organometallics 2012, 31, 787. (10) Dahy, A. A.; Koga, N.; Nakazawa, H. Organometallics 2012, 31, 3995. (11) (a) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.; Kudin, K. N.; Burant, J. C.; Millam, J. M.; Iyenger, S. S.; Tomasi, J.; Barone, V.; Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.; Nakatsuji, N.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Comperts, R.; Startmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth, G. A.; Salvador, P.; Dannenberg, D. D.; Zakrzewiski, V. G.; Dapprich, S.; Daniels, A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian 03, Revision D.02; Gaussian, Inc.: Pittsburgh, PA, 2004. (b) 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., Jr.; 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, Ö .; Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian 09, Revision C.01; Gaussian, Inc.: Wallingford, CT, 2009. (12) (a) Becke, A. D. J. Chem. Phys. 1993, 98, 5648. (b) Becke, A. D. J. Chem. Phys. 1996, 104, 1040. (13) (a) Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37, 785. (b) Mielich, B.; Savin, H.; Preuss, H. Chem. Phys. Lett. 1989, 157, 200. (c) Burke, K.; Perdew, J. P.; Wang, Y. In Electronic Density Functional Theory: Recent Progress and New Directions; Dobson, J. F.; Vignale, G.; Das, M. P., Eds.; Plenum: New York, 1998. (d) Perdew, J. P. In Electronic Structure of Solids 91; Ziesche, P.; Eschrig, H., Eds.; Akademie Verlag: Berlin, 1991; p 11. (e) Perdew, J. P.; Chevary, J. A.; Vosko, S. H.; Jakson, K. A.; Pederson, M. R.; Singh, D. J.; Fiolhais, C. Phys. Rev. B 1992, 46, 6671. (f) Perdew, J. P.; Chevary, J. A.; Vosko, S. H.; Jakson, K. A.; Pederson, M. R.; Singh, D. J.; Fiolhais, C. Phys. Rev. B 1993, 48, 4978. (g) Perdew, J. P.; Burke, K.; Wang, Y. Phys. Rev. B 1996, 54, 16533. (14) Hay, P. J.; Wadt, W. R. J. Chem. Phys. 1985, 82, 299. (15) (a) Hehre, W. J.; Ditchfield, R.; Pople, J. A. J. Chem. Phys. 1972, 56, 2257. (b) Hariharan, P. C.; Pople, J. A. Theor. Chem. Acta 1973, 28, 213. (c) Francl, M. M.; Pietro, W. J.; Hehre, W. J.; Binkley, J. S.; Defrees, D. J.; Gordon, M. S. J. Chem. Phys. 1982, 77, 3654. (16) (a) Krishnan, R.; Binkley, J. S.; Seeger, R.; Pople, J. A. J. Chem. Phys. 1980, 72, 650. (b) Wachters, A. J. H. J. Chem. Phys. 1970, 52, 1033. (c) Hay, P. J. J. Chem. Phys. 1977, 66, 4377. (d) McLean, A. D.; Chandler, G. S. J. Chem. Phys. 1980, 72, 5639. (17) Foresman, J. B.; Frisch, A. E. Exploring Chemistry with Electronic Structure Methods, 2nd ed.; Gaussian Inc.: Pittsburgh, PA, 1995; Chapter 7, p 158. K

dx.doi.org/10.1021/om400179y | Organometallics XXXX, XXX, XXX−XXX