Mechanistic Insights into Scandium-Catalyzed ... - ACS Publications

Apr 10, 2017 - Prediction Model. Fan Liu,. †. Gen Luo,. †,‡. Zhaomin Hou,*,†,‡ and Yi Luo*,†. †. State Key Laboratory of Fine Chemicals,...
1 downloads 22 Views 3MB Size
This is an open access article published under an ACS AuthorChoice License, which permits copying and redistribution of the article or any adaptations for non-commercial purposes.

Article pubs.acs.org/Organometallics

Mechanistic Insights into Scandium-Catalyzed Hydroaminoalkylation of Olefins with Amines: Origin of Regioselectivity and Charge-Based Prediction Model Fan Liu,† Gen Luo,†,‡ Zhaomin Hou,*,†,‡ and Yi Luo*,† †

State Key Laboratory of Fine Chemicals, School of Chemical Engineering, Dalian University of Technology, Dalian 116024, People’s Republic of China ‡ Organometallic Chemistry Laboratory and RIKEN Center for Sustainable Resource Science, RIKEN, 2-1 Hirosawa, Wako, Saitama 351-0198, Japan S Supporting Information *

ABSTRACT: DFT calculations have been carried out on the Sccatalyzed regioselective hydroaminoalkylation of olefins with amines. It has been revealed that the true active species is an amine-coordinated η2azametallacyclic complex, and electronic factors play a crucial role in achieving regioselectivity. The charge dispersion and charge alternation account well for the stability of the olefin insertion transition states and products. The charge distribution of 15 olefin substrates used previously in the experimental studies correlates well with the observed regioselectivity and could thus provide a potential model for regioselectivity prediction. The reactivity of different types of C−H bonds was also explored by modeling the reaction of (CH2C6H4NMe2-o)2Sc+ with iPrN(Me)(Et). The suggested reactivity trend for the Sc-catalyzed hydroaminoalkylation follows the order of primary C−H bond > secondary C−H bond > tertiary C−H bond.



INTRODUCTION The synthesis of amine compounds has received intense attention for a long time due to the critical function of amine groups in many biologically active compounds and industrial functional materials.1 Catalytic C(sp3)−H bond activation at the position α to a nitrogen atom has recently attracted much interest from synthetic organic chemists because this offers new and promising synthetic approaches for the functionalization of simple amines.2 In particular, the catalytic hydroaminoalkylation of olefins via C(sp3)−H bond activation of alkyl amines is an atom-economical (100%) strategy and thus it is an excellent reaction to target for advances in green chemistry.3,4 The control of regioselectivity is highly important in organic synthesis and theoretical organic chemistry in general. In the hydroaminoalkylation of olefins, both linear and branched products could be formed. So far, a majority of the hydroaminoalkylation catalysts have been based on group 4 and 5 (Ti, Zr, Nb, Ta) metal complexes.5,6 To the best of our knowledge, the reported early-transition-metal catalysts for this transformation usually give branched or a mixture of both linear and branched isomers.3−8 In contrast to early transition metals, late-transition-metal catalysts (Ru, Ir) often give linear products, but the reactions have been limited to amine substrates with a pyridyl directing group.9 The regioselective hydroaminoalkylation of olefins with simple amines without an extra directing group remains scarce. Very recently, Hou and co-workers reported that a rare-earth alkyl catalyst could © XXXX American Chemical Society

promote the intermolecular hydroaminoalkylation of olefins with a variety of tertiary amines under relatively mild conditions and, in particular, with an exclusively substrate dependent regioselectivity (Scheme 1).10 This reaction could offer an ideal Scheme 1. Scandium-Catalyzed Hydroaminoalkylation of Olefins with Amines

model for mechanistic studies with the purpose of developing regioselective catalytic systems. On the basis of previous studies of C−H functionalization by rare-earth alkyl catalysts,11−15 a plausible mechanism for this reaction was proposed, as shown in Scheme 2.10 The deprotonation of a methyl group in an amine substrate by one aminobenzyl ligand (R*) on the cationic scandium precatalyst 1 16 could give the η 2 azametallacyclic intermediate 2 as a catalytically active species,17 along with release of 1 equiv of N,N-dimethyl-o-toluidine (*RH). Subsequently, an olefin could insert into the Sc−C Received: February 15, 2017

A

DOI: 10.1021/acs.organomet.7b00116 Organometallics XXXX, XXX, XXX−XXX

Organometallics



RESULTS AND DISCUSSION In the mechanistic discussions, the hydroaminoalkylation of Nmethylpiperidine (a1) with n-hexene was chosen as a model

Scheme 2. Proposed Mechanism of Scandium-Catalyzed Intermolecular Hydroaminoalkylation

bond of 2 via 1,2- and 2,1-insertion manners (depending on the R group of the olefin) to give the corresponding ring-expanded azametallacyclic intermediate 3 or 4. Finally, the C−H activation of another molecule of amine substrate could give the corresponding branched or linear products and complete the catalytic cycle. Despite the aforementioned mechanistic suggestions, many details remain ambiguous. Several pertinent questions have arisen from this reaction. (1) Is the η2-azametallacyclic intermediate 2 the true active species? (2) What is the ratedetermining step in the catalytic cycle? (3) Why is the C(sp3)− H hydroaminoalkylation generally limited to the α-methyl C− H bond of amines rather its α-methylene C−H bond? (4) What is the origin of exclusively substrate dependent regioselectivity (branched or linear products)? To answer these questions, which are generally important for the development of a highly regioselective hydroaminoalkylation catalytic system, DFT calculations on the reaction of Sccatalyzed α-C(sp3)−H addition of amines to olefins have been carried out in this study. It has been found that electronic factors play a crucial role in the regioselectivity. The charge distribution of olefin substrates has been proposed to serve as a potential prediction model for the regioselectivity of such reactions.



Article

Figure 1. Gibbs energy profile for the formation of active species via the reaction of 1 with a1.

reaction. On the basis of the aforementioned mechanism (Scheme 2), the following mechanistic discussions are generally divided into three steps: the generation of active species, alkene insertion, and C−H bond activation of another amine. Generation of Active Species. The calculated Gibbs energy profile and important optimized structures for the formation of active species are shown in Figures 1 and 2, respectively. As shown in Figure 1, the reaction proceeds through the coordination of N-methylpiperidine to the cationic scandium species 1, leading to the more stable complex A (ΔG = −16.3 kcal/mol). Then A goes through the hydrogen transfer transition state TS1 to generate complex B with an N,Ndimethyl-o-toluidine moiety. In TS1 (Figure 2), the C1, Sc, and C2 atoms and the transferring H construct a four-center structure, which features a typical σ-bond metathesis (σ-BM) event.25,26 This C−H bond activation step has an energy barrier of 25.7 kcal/mol. Subsequently, complex B could release N,Ndimethyl-o-toluidine to give 2 with a three-membered-ring moiety, as proposed previously.10 However, 2 has a higher free energy than A by 22.5 kcal/mol and is relatively unstable. Therefore, it seems that species 2 may not be the real active species in this reaction. Alternatively, since the amine substrate is abundant in the reaction system, the ligand exchange between N,N-dimethyl-o-toluidine and N-methylpiperidine in complex B could easily occur to form the N-methylpiperidinecoordinated species 2′, which is significantly more stable than 2 by 19.2 kcal/mol (see Figure S1 in the Supporting Information). The species 2 and 2′ both have a η2azametallacycle moiety, and the bond lengths indicate that the interaction among Sc and adjacent atoms in 2 is stronger than that in 2′ (Figure 2). The other possible active species with two η2-azametallacycle moieties formed by further C(sp3)−H bond activation in 2′ will also be discussed and is found to have higher energy (vide infra). Therefore, 2′ is more likely the real active species during the catalytic cycle, and its formation is endergonic by 3.3 kcal/mol relative to A.

COMPUTATIONAL METHODS

All calculations were performed with the Gaussian 09 program.18 The B3PW91 functional19 was used for geometry optimizations together with the 6-31G(d) basis set for C, H, and N atoms and Stuttgart/ Dresden relativistic effective core potentials (RECPs)20 as well as the associated valence basis sets for the Sc atom. Each optimized structure was subsequently analyzed by harmonic vibrational frequencies at the same theoretical level to characterize each stationary point as a minimum (NImag = 0) or a transition state (NImag = 1) and to obtain the thermodynamic corrections to Gibbs free energy (298.15 K). To obtain more reliable relative energies, single-point energy calculations were carried out with a larger basis set. In such single-point calculations, the M06 functional, which has good performance in the treatment of transition-metal systems,21 was used together with the 6311+G(d,p) basis set for all atoms and the CPCM solvation model22 (in toluene solution with UFF atomic radii23) for consideration of solvation effects. The reported Gibbs free energy in solution includes the gas-phase Gibbs free energy correction. To reduce the overestimation of the entropy contribution derived from the gas-phase model, corrections for free energies were made by −2.6 (or +2.6) kcal/mol for 2:1 (or 1:2) molecule transformations.24 The relative enthalpies and uncorrected relative free energies of all stationary points are also provided in Table S1 in the Supporting Information for reference. B

DOI: 10.1021/acs.organomet.7b00116 Organometallics XXXX, XXX, XXX−XXX

Article

Organometallics

Figure 2. Optimized structures of TS1, 2, and 2′.

Sc−C bond. For a comparison, both possibilities were calculated. As illustrated in Figure 3, direct C(sp3)−H bond activation via TS2′ needs to overcome an energy barrier of 27.9 kcal/mol and leads to intermediate D′ with two η2-azametallacycle moieties and a coordinated N,N-dimethyl-o-toluidine. Similar to TS1 shown in Figures 1 and 2, the transition state TS2′ also has a four-center structure, suggesting a one-step σBM process. For the other reaction pathway, the ligand exchange between olefin and N-methylpiperidine first occurs to form the π-complex C (Figure S2 in the Supporting Information). This process is endergonic by 13.4 kcal/mol, which can be traced to the stronger coordination capability of amine in comparison to olefin. Thereafter, the CC double bond in C could insert into the Sc−C bond via a conventional four-center olefin insertion transition state,27 TS2, yielding the ring-expanded azametallocyclic intermediate D. The olefin insertion step has a free energy barrier of 22.2 kcal/mol, which is significantly lower than that for the C(sp3)−H activation (TS2′, ΔG⧧ = 27.9 kcal/mol). Obviously, such an olefin insertion is both kinetically and thermodynamically more favored than the direct C(sp3)−H bond activation (Figure 3). Therefore, after the formation of 2′, olefin insertion could take place. In addition, D′ is unlikely to be generated. C−H Bond Activation of Another Amine Molecule. After olefin insertion, another N-methylpiperidine molecule

Figure 3. Gibbs energy profiles for olefin insertion into 2′ and for intramolecular C−H bond activation of 2′.

Olefin Insertion. Starting with the species 2′, there are two possibilities for further reactions: (1) direct C−H bond activation in 2′, giving a new species with two η2-azametallacycle moieties, and (2) ligand exchange between n-hexene and N-methylpiperidine and subsequent olefin insertion into the

Figure 4. Gibbs energy profile for the C−H bond activation of the second amine molecule. C

DOI: 10.1021/acs.organomet.7b00116 Organometallics XXXX, XXX, XXX−XXX

Article

Organometallics

profile suggests that C1 is a more favorable site for C−H bond activation in comparison with the case of C1′ (Figure 4), which is consistent with the experimental result that the hydroaminoalkylation product was obtained in the present reaction. Along with the favorable pathway, the C−H bond activation via TS3 has the highest energy barrier (ΔG⧧ = 28.3 kcal/mol) and is the rate-determining step in the catalytic cycle, which is also suggested by experiments determining the kinetic isotope effect. The whole catalytic process is exergonic by 23.9 kcal/ mol after one turnover. As discussed above, the computational results generally support the proposed mechanism, which mainly involves the generation of a η2-azametallacyclic intermediate as the active species, the subsequent olefin insertion into the Sc−C bond, and the C−H activation of another molecule of amine to give the hydroaminoalkylation product and regenerate the active species (Scheme 3). This result suggests that the aminecoordinated η2-azametallacyclic complex should be the real active species involved in the catalytic cycle rather than the bare η2-azametallacyclic species proposed previously.10 It is noteworthy that alkene insertion is not the rate-determining step, which is different from our previous work on the alkylation of pyridines.12b This could thus account for the experimental observation that a significant excess of olefins is unnecessary in the current reaction. Obviously, the mechanism presented here is in sharp contrast with previously reported early-/latetransition-metal catalyzed hydroaminoalkylation reactions. In early-transition-metal (such as titanium) systems, the metallaaziridine species (such as II in Chart 1) as a key active intermediate has both M−C and M−N bonds, which is different from the η2-azametallacyclic active species (such as I in Chart 1) containing only an Sc−C bond in the Sc-catalyzed system.4,5d In the case of II, after olefin insertion into the M−C bond, two proton transfer events between N and C atoms (sequential N−H and C−H bond activation) are required to cleave M−C and M−N bonds to achieve the dissociation of the hydroaminoalkylation product. This is in contrast to the Sccatalyzed system, in which only one proton transfer (C−H bond activation) occurred during the catalytic cycle and the proton transfer occurred between two C atoms rather than N and C atoms. Due to the requirement of deprotonation of the N−H bond in the early-transition-metal systems, the reactions are only possibly suitable for primary and secondary amines, while tertiary amines work in the hydroaminoalkylation by the Sc catalyst. For the late-transition-metal catalysts, a mechanism was proposed via a sequential oxidative addition/olefin insertion/reductive elimination process, during which the redox metal center was essential to achieve such a reaction.9a Regioselectivity for the Olefin Insertion. It was experimentally found that different kinds of products (linear or branched products) were obtained when different types of olefins were used.10 Aryl-substituted olefins (styrene and styrene derivatives) and alkyl-substituted olefins would lead to linear and branched products, respectively. In addition to styrenes, vinyltrimethylsilane also gave the linear product. To access the origin of such interesting regioselectivity, three olefin substrates, viz. n-hexene (as a representative of alkyl-substituted olefins), styrene (as a representative of aryl-substituted olefins), and vinyltrimethylsilane (a special case), were chosen for the calculations (Scheme 4). The regioselectivity is determined in the olefin insertion step, which could yield a branched product via 1,2-insertion or a linear product via 2,1-insertion. As shown in Figure 5a, for n-

Scheme 3. Computationally Suggested Mechanism of Scandium-Catalyzed Hydroaminoalkylation

Chart 1

Scheme 4. Hydroaminoalkylation of N-Methylpiperidine with Different Olefins

could coordinate to the Sc center in D, which then undergoes C−H bond activation via a σ-BM mechanism. In D, there are two possible carbon sites (C1 and C1′, Figure 4) for accepting the transferring proton. As shown in Figure 4, the σ-BM (C−H activation) between methyl and C1-methylene via TS3, which has geometric features similar to those of TS1 shown in Figure 1, could form complex F. A ligand exchange could feasibly take place between a1 and F (Figure S3 in the Supporting Information) to finally release the hydroaminoalkylation product (P1) and regenerate the active species 2′. The other pathway for the C−H bond activation between the methyl and C1′-methylene via TS3′ yields the *RH-coordinated complex F′, to finally give N,N-dimethyl-o-toluidine (*RH) instead of the experimentally observed product. The resulting energy D

DOI: 10.1021/acs.organomet.7b00116 Organometallics XXXX, XXX, XXX−XXX

Article

Organometallics

Figure 5. Gibbs energy profiles for (a) n-hexene, (b) styrene, and (c) vinyltrimethylsilane insertion into the active species 2′ via 1,2-insertion and 2,1-insertion manners.

Figure 6. Optimized structures of olefin insertion transition states. Mulliken atomic charges are given in blue with parentheses. |Q| and S denote the unsigned average charge and square error (S = ∑(|Qx| − |Q|)2) of Sc, C1, C2, and C3 atoms, respectively.

hexene, the relative energy of the 1,2-insertion transition state TS212 is found to be lower than that for the 2,1-insertion transition state TS221 by 1.7 kcal/mol and the corresponding 1,2-insertion product D12 is more stable than the 2,1-insertion product D21 by 2.6 kcal/mol, implying that the 1,2-insertion manner is more favorable than 2,1-insertion both kinetically and thermodynamically. In contrast, the energy profiles illustrated in Figure 5b,c show that 2,1-insertion manner is both kinetically and thermodynamically more favored over the 1,2-insertion manner for styrene and vinyltrimethylsilane substrates, respectively. Therefore, the computational results

for all those three olefin substrates are well in accordance with the experimental observations on regioselectivity. Having achieved an agreement between calculation and experiment, we then focused on the origin of regioselectivity. To get more information on steric and electronic factors, the olefin insertion transition states and corresponding insertion products together with selected bond lengths and atomic charges are shown in Figures 6 and 7, respectively. As shown in Figure 6, the Sc−C2 bond lengths in TS221 are longer than those in TS212 (Sc−C1) by 0.05, 0.05, and 0.08 Å for the nhexene, styrene, and vinyltrimethylsilane cases, respectively, due E

DOI: 10.1021/acs.organomet.7b00116 Organometallics XXXX, XXX, XXX−XXX

Article

Organometallics

Figure 7. Optimized structures of olefin insertion products. Mulliken atomic charge are given in blue with parentheses. |Q| and S denote the unsigned average charge and square error (S = ∑(|Qx| − |Q|)2) of Sc, C1, C2, C3, and N1 atoms, respectively.

Figure 8. Mulliken atomic charge distribution on CC double bond moieties of olefin substrates used in experiments.10

Figure 9. Comparison of α-C(cyc)−H and α-C(Me)−H activations of N-methylpiperidine.

to steric repulsion between the R group of the olefin and the aminobenzyl ligand (R*). These results indicate that similar steric repulsions exist in all three olefin cases and steric factors tend to lead to the 1,2-insertion product,28 which could not explain the difference in regioselectivity. Therefore, steric factors could not be the origin of the regioselectivity. As we know, charge dispersion is conducive to the stability of a structure. In olefin insertion transition states, the fourmembered Sc−C1−C2−C3 (or Sc−C2−C1−C3) units are the most important moieties related to chemical changes. To understand the stability of the structures, the unsigned average charges (|Q|) and square errors (S) of Sc, C1, C2, and C3 atoms for transition states TS212 and TS221 were adopted in this study to estimate the degree of charge dispersion. In general, the smaller the values of |Q| and S, the more stable the structure. For n-hexene (Figure 6), |Q| = 0.52 and S = 0.040 for

the 1,2-insertion transition state TS212 are smaller than those (| Q| = 0.66 and S = 0.092) for the 2,1-insertion transition state TS221, suggesting that the former is more stable than the latter. As expected, in contrast to n-hexene, the |Q| and S values of TS212 for the case of styrene and vinyltrimethylsilane are larger than those of TS221 (|Q| = 0.47 and S = 0.283 vs |Q| = 0.38 and S = 0.202 for styrene, |Q| = 0.64 and S = 0.328 vs |Q| = 0.32 and S = 0.203 for vinyltrimethylsilane). This suggests that the 2,1insertion manner is more favorable than the 1,2-insertion manner for styrene and vinyltrimethylsilane substrates. These theoretical results derived from the unsigned average charge and square error of olefin insertion transition states could account for the observed regioselectivity. In addition to the charge dispersion, the charge alternation effect also makes an important contribution to the stability of F

DOI: 10.1021/acs.organomet.7b00116 Organometallics XXXX, XXX, XXX−XXX

Article

Organometallics

Figure 10. Gibbs energy profiles for the primary, secondary, and tertiary C−H activations in the reaction of 1 with iPrN(Me)(Et).

the transition states.28,29 For n-hexene (Figure 6), the atomic charges on Sc(0.68)−C1(−0.52)−C2(0.45) in TS212 shows a charge alternation pattern, while that on Sc(0.63)−C2(0.82)− C1(−0.75)/C4(−0.51) in TS221 does not. Thus, 1,2-insertion is favored for the n-hexene case. For styrene, the charge alternation pattern of Sc(0.56)−C2(−0.08)−C4(0.28) exists in TS221 and therefore has a lower free energy than TS212 with a charge distribution of Sc(0.65)−C1(−0.77)−C2(−0.08). A similar charge distribution pattern is also observed for the vinyltrimethylsilane case, which accounts for its favorable 2,1insertion. All of these results are also in good agreement with the experimental observations on the regioselectivity. Energy profiles suggest that the favored insertion manner also leads to a thermodynamically favorable product. As shown in Figure 7, all of the insertion products have similar geometric parameters of the five-membered Sc−C1−C2−C3−N1 (or Sc−C2−C1−C3−N1) units. Therefore, steric repulsion could not explain the thermodynamic stability of the corresponding products. Like the case of transition states, the stability of insertion products can be also generally explained by charge dispersion and charge alternation principles.30 For instance, the 1,2-insertion product for n-hexene and 2,1-insertion products for styrene and vinyltrimethylsilane have smaller |Q| and S values of the five-membered Sc−C1−C2−C3−N1 units in comparison to those for the other corresponding products. Therefore, the thermodynamic stability of insertion products is also controlled by electronic factors. On the basis of an analysis of olefin insertion transition states and products, it is concluded that electronic factors play a crucial role in the regioselectivity. This result motivated us to wonder whether the regioselectivity can be directly predicted by the charge distribution on the CC double bond moiety of olefin substrates. As we know, the nucleophilic center possessing more negative charge would more easily attack the electrophilic Sc center. As shown in Figure 8, the atomic charge distribution on CC double bond moieties of 15 olefin substrates used in experiments10 is displayed and the result suggests that the atomic charge distribution of all substrates could give a good explanation for their regioselectivity. That is, C2 atoms of the CC double bond possess more negative charge for aryl-substituted olefins and vinyltrimethylsilane and thus undergo a 2,1-insertion manner to give linear products, while C1 atoms of the CC double bond possess more negative charge for alkyl-substituted olefins and thus undergo 1,2-insertion manner to give branched products. This result

demonstrates again that electronic factors play an essential role in the regioselectivity. Additionally, the charge distribution information could provide a potential model for predicting the regioselectivity of such rare-earth catalyzed hydroaminoalkylation reactions. Reactivity of C−H Bonds. It is experimentally found that hydroaminoalkylation catalyzed by a scandium catalyst exclusively occurs at α-methyl C−H bonds in N-methylpiperidine rather than α-methylene C−H bonds.10 To get a better understanding of this regiochemical control, the activation of the α-position secondary C(cyc)−H bond of methylene in a1 was also investigated and the result is shown in Figure 9 (αC(Me)−H activation is illustrated again for comparison). The computational result shows that α-C(cyc)−H bond activation needs a high energy barrier of 32.4 kcal/mol, which is significantly more unfavorable than α-C(Me)−H bond activation (ΔG⧧ = 25.7 kcal/mol). This result could give a better understanding of the experimental observation. In contrast, previous studies on C−H bond reactivity often show the order of tertiary C−H bond > secondary C−H bond > primary C−H bond in processes such as alkane hydroxylation31 and C−H amination.32 To further examine the reactivity of C−H bonds in scandium-catalyzed hydroaminoalkylation, iPrN(Me)(Et) was chosen as an amine model compound for the calculation of various C−H activations, in accordance with the reaction of 1 with iPrN(Me)(Et). As shown in Figure 10, the energy barriers for the activation of primary, secondary, and tertiary C−H bonds are 25.2, 29.4, and 30.9 kcal/mol, respectively. This result indicates that the reactivity of C−H bonds in Sc-catalyzed hydroaminoalkylation generally follows the order of primary C−H bond > secondary C−H bond > tertiary C−H bond, which is sharply different from previous studies on C−H bond reactivity involved in alkane hydroxylation and C−H amination.



CONCLUSION The reaction mechanism and the origin of regioselectivity of the scandium-catalyzed α-C(sp3)−H addition of amines to olefins have been investigated by DFT calculations. The computational results suggest that the reaction mainly involves three steps: (a) generation of an active species, (b) olefin insertion into the active species, and (c) subsequent C−H bond activation of an amine. It has been found that the true active species is an amine-coordinated η2-azametallacyclic complex rather than the base-free η2-azametallacyclic analogue proposed G

DOI: 10.1021/acs.organomet.7b00116 Organometallics XXXX, XXX, XXX−XXX

Article

Organometallics

(4) For recent reviews on the hydroaminoalkylation of olefins, see: (a) Roesky, P. W. Angew. Chem., Int. Ed. 2009, 48, 4892−4894. (b) Chong, E.; Garcia, P.; Schafer, L. L. Synthesis 2014, 46, 2884− 2896. (5) For examples of group 4 metal catalysts, see: (a) Bexrud, J. A.; Eisenberger, P.; Leitch, D. C.; Payne, P. R.; Schafer, L. L. J. Am. Chem. Soc. 2009, 131, 2116−2118. (b) Kubiak, R.; Prochnow, I.; Doye, S. Angew. Chem., Int. Ed. 2009, 48, 1153−1156. (c) Kubiak, R.; Prochnow, I.; Doye, S. Angew. Chem., Int. Ed. 2010, 49, 2626−2629. (d) Prochnow, I.; Zark, P.; Müller, T.; Doye, S. Angew. Chem., Int. Ed. 2011, 50, 6401−6405. (6) For examples of group 5 metal catalysts, see: (a) Herzon, S. B.; Hartwig, J. F. J. Am. Chem. Soc. 2007, 129, 6690−6991. (b) Herzon, S. B.; Hartwig, J. F. J. Am. Chem. Soc. 2008, 130, 14940−14941. (c) Eisenberger, P.; Ayinla, R. O.; Lauzon, J. M. P.; Schafer, L. L. Angew. Chem., Int. Ed. 2009, 48, 8361−8365. (d) Reznichenko, A. L.; Hultzsch, K. C. J. Am. Chem. Soc. 2012, 134, 3300−3311. (7) Garcia, P.; Lau, Y.; Perry, M.; Schafer, L. Angew. Chem., Int. Ed. 2013, 52, 9144−9148. (8) For examples of linear isomers as the major products, see: (a) Dörfler, J.; Doye, S. Angew. Chem., Int. Ed. 2013, 52, 1806−1809. (b) Dörfler, J.; Preuß, T.; Schischko, A.; Schmidtmann, M.; Doye, S. Angew. Chem., Int. Ed. 2014, 53, 7918−7922. (9) (a) Chatani, N.; Asaumi, T.; Yorimitsu, S.; Ikeda, T.; Kakiuchi, F.; Murai, S. J. Am. Chem. Soc. 2001, 123, 10935−10941. (b) Bergman, S. D.; Storr, T. E.; Prokopcová, H.; Aelvoet, K.; Diels, G.; Meerpoel, L.; Maes, B. U. W. Chem. - Eur. J. 2012, 18, 10393−10398. (c) Schinkel, M.; Wang, L.; Bielefeld, K.; Ackermann, L. Org. Lett. 2014, 16, 1876− 1879. (d) Pan, S.; Endo, K.; Shibata, T. Org. Lett. 2011, 13, 4692− 4695. (10) Nako, A. E.; Oyamada, J.; Nishiura, M.; Hou, Z. Chem. Sci. 2016, 7, 6429−6434. (11) Nishiura, M.; Guo, F.; Hou, Z. Acc. Chem. Res. 2015, 48, 2209− 2220. (12) (a) Guan, B. T.; Hou, Z. J. Am. Chem. Soc. 2011, 133, 18086− 18089. (b) Luo, G.; Luo, Y.; Qu, J.; Hou, Z. Organometallics 2012, 31, 3930−3937. (c) Guan, B. T.; Wang, B.; Nishiura, M.; Hou, Z. Angew. Chem., Int. Ed. 2013, 52, 4418−4421. (d) Song, G.; O, W. N. W.; Hou, Z. J. Am. Chem. Soc. 2014, 136, 12209−12212. (e) Song, G.; Wang, B.; Nishiura, M.; Hou, Z. Chem. - Eur. J. 2015, 21, 8394−8398. (13) Oyamada, J.; Hou, Z. Angew. Chem., Int. Ed. 2012, 51, 12828− 12832. (14) Song, G.; Luo, G.; Oyamada, J.; Luo, Y.; Hou, Z. Chem. Sci. 2016, 7, 5265−5270. (15) Shi, X.; Nishiura, M.; Hou, Z. J. Am. Chem. Soc. 2016, 138, 6147−6150. (16) The cationic scandium alkyl species 1 could be obtained through the reaction of [Sc(CH2C6H4NMe2-o)3] with 1 equiv of [Ph3C][B(C6F5)4]. For the X-ray crystal structure of a cationic halfsandwich scandium complex, see: Li, X.; Nishiura, M.; Mori, K.; Mashiko, T.; Hou, Z. Chem. Commun. 2007, 4137−4139. (17) A cationic Sc-η2-pyridyl species via deprotonation of pyridine by the aminobenzyl ligand (R*) of a cationic scandium species was isolated and confirmed to be a key catalytic species; see ref 12e. (18) 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, N. J.; 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 .;

previously. C−H bond activation is found to be the ratedetermining step in the catalytic cycle, in agreement with the experimental KIE results. Having achieved an agreement between theory and experiment in regioselectivity, the detailed charge analyses suggest that the regioselectivity is under electronic control in the scandium-catalyzed hydroaminoalkylation reaction. In addition, the Mulliken charge distribution of 15 olefin substrates used in previous experiments correlates well with the observed regioselectivity and thus provides a potential model to predict the regioselectivity in such reactions. It has been also found that the activation of the α-C(Me)−H bond of methyl in N-methylpiperidine is both kinetically and thermodynamically more favorable than that of its α-C(cyc)− H bond. In addition, the model reaction of (CH2C6H4NMe2o)2Sc+ with iPrN(Me)(Et) indicates that the reactivity of the C−H bond generally follows the order of primary C−H bond > secondary C−H bond > tertiary C−H bond. These theoretical results could be helpful for developing regioselective catalytic hydroaminoalkylation reactions.



ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.organomet.7b00116. Details of the ligand exchange processes, energy corrections from gas-phase frequency analysis, singlepoint electronic energies, relative enthalpies, and relative free energies without corrections (PDF) All optimized Cartesian coordinates of stationary points as well as imaginary frequencies of transition states (XYZ)



AUTHOR INFORMATION

Corresponding Authors

*E-mail for Z.H.: [email protected]. *E-mail for Y.L.: [email protected]. ORCID

Zhaomin Hou: 0000-0003-2841-5120 Yi Luo: 0000-0001-6390-8639 Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS This work was partially supported by the NSFC (Nos. 21429201, 21674014) and a Grant-in-Aid for Scientific Research (S) from the JSPS (No. 26220802). The authors also thank the RICC (RIKEN Integrated Cluster of Clusters) and the Network and Information Center of the Dalian University of Technology for part of the computational resources.



REFERENCES

(1) Salvatore, R. N.; Yoon, C. H.; Jung, K. W. Tetrahedron 2001, 57, 7785−7811. (2) For reviews on catalytic C−H activation in the position α to a nitrogen atom, see: (a) Campos, K. R. Chem. Soc. Rev. 2007, 36, 1069−1084. (b) Doye, S. Angew. Chem., Int. Ed. 2001, 40, 3351−3353. (3) (a) Clerici, M. G.; Maspero, F. Synthesis 1980, 1980, 305−306. (b) Nugent, W. A.; Ovenall, D. W.; Holmes, S. J. Organometallics 1983, 2, 161−162. H

DOI: 10.1021/acs.organomet.7b00116 Organometallics XXXX, XXX, XXX−XXX

Article

Organometallics Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian 09, Revision A.01; Gaussian, Inc., Wallingford, CT, 2009. (19) (a) Becke, A. D. J. Chem. Phys. 1993, 98, 5648−5652. (b) Lee, C. T.; Yang, W. T.; Parr, R. G. Phys. Rev. B: Condens. Matter Mater. Phys. 1988, 37, 785−789. (c) Perdew, J. P.; Burke, K.; Wang, Y. Phys. Rev. B: Condens. Matter Mater. Phys. 1996, 54, 16533−16539. (20) (a) Dolg, M.; Wedig, U.; Stoll, H.; Preuss, H. J. Chem. Phys. 1987, 86, 866−872. (b) Schwerdtfeger, P.; Dolg, M.; Schwarz, W. H. E.; Bowmaker, G. A.; Boyd, P. D. W. J. Chem. Phys. 1989, 91, 1762− 1774. (c) Dolg, M.; Stoll, H.; Savin, A.; Preuss, H. Theor. Chim. Acta 1989, 75, 173−194. (d) Andrae, D.; Haeussermann, U.; Dolg, M.; Stoll, H.; Preuss, H. Theor. Chim. Acta 1990, 77, 123−141. (e) Dolg, M.; Stoll, H.; Preuss, H. Theor. Chim. Acta 1993, 85, 441−450. (f) Bergner, A.; Dolg, M.; Kuechle, W.; Stoll, H.; Preuss, H. Mol. Phys. 1993, 80, 1431−1441. (21) (a) Zhao, Y.; Truhlar, D. G. Acc. Chem. Res. 2008, 41, 157−167. (b) Zhao, Y.; Truhlar, D. G. Theor. Chem. Acc. 2008, 120, 215−241. (22) (a) Barone, V.; Cossi, M. J. Phys. Chem. A 1998, 102, 1995− 2001. (b) Cossi, M.; Rega, N.; Scalmani, G.; Barone, V. J. Comput. Chem. 2003, 24, 669−681. (23) (a) Miertus, S.; Scrocco, E.; Tomasi, J. Chem. Phys. 1981, 55, 117−129. (b) Cammi, R.; Mennucci, B.; Tomasi, J. J. Phys. Chem. A 2000, 104, 5631−5637. (24) (a) Wang, B.; Luo, G.; Nishiura, M.; Hu, S.; Shima, T.; Luo, Y.; Hou, Z. J. Am. Chem. Soc. 2017, 139, 1818−1821. (b) Kang, X.; Luo, G.; Luo, L.; Hu, S.; Luo, Y.; Hou, Z. J. Am. Chem. Soc. 2016, 138, 11550−11559. (c) Bai, W.; Lee, K.-H.; Sung, H. H. Y.; Williams, I. D.; Lin, Z.; Jia, G. Organometallics 2016, 35, 3808−3815. (d) Xie, H.; Lin, Z. Organometallics 2014, 33, 892−897. (e) Yuan, R.; Lin, Z. Organometallics 2014, 33, 7147−7156. (f) Liu, Q.; Lan, Y.; Liu, J.; Li, G.; Wu, Y.-D.; Lei, A. J. Am. Chem. Soc. 2009, 131, 10201−10210. (25) Selected reviews: (a) Vastine, B. A.; Hall, M. B. Coord. Chem. Rev. 2009, 253, 1202−1218. (b) Lin, Z. Coord. Chem. Rev. 2007, 251, 2280−2291. (c) Lin, Z. Acc. Chem. Res. 2010, 43, 602−611. (d) Balcells, D.; Clot, E.; Eisenstein, O. Chem. Rev. 2010, 110, 749− 823. (e) Waterman, R. Organometallics 2013, 32, 7249−7263. (f) Kefalidis, C. E.; Castro, L.; Perrin, L.; Del Rosal, I.; Maron, L. Chem. Soc. Rev. 2016, 45, 2516−2543. (26) For an example of multimetal-cooperated σ-bond metathesis C−H activation, see: Luo, G.; Luo, Y.; Zhang, W.; Qu, J.; Hou, Z. Organometallics 2014, 33, 1126−1134. (27) (a) Niu, S.; Hall, M. B. Chem. Rev. 2000, 100, 353−406. (b) Luo, G.; Luo, Y.; Hou, Z.; Qu, J. Organometallics 2016, 35, 778− 784. (c) Kang, X.; Song, Y.; Luo, Y.; Li, G.; Hou, Z.; Qu, J. Macromolecules 2012, 45, 640−651. (d) Luo, Y.; Luo, Y.; Qu, J.; Hou, Z. Organometallics 2011, 30, 2908−2919. (28) Bi, S.; Lin, Z.; Jordan, R. F. Organometallics 2004, 23, 4882− 4890. (29) (a) Klein, J. Tetrahedron 1988, 44, 503−518. (b) Becker, J. Y.; Klein, J. Tetrahedron 1988, 44, 2289−2292. (c) Stolow, R. D.; Samal, P. W.; Giants, T. W. J. Am. Chem. Soc. 1981, 103, 197−199. (d) Cohen, Y.; Klein, J.; Rabinovitz, M. J. Am. Chem. Soc. 1988, 110, 4634−4640. (e) Frenking, G.; Gobbi, A. Chem. Phys. Lett. 1992, 197, 335. (30) The stability of the styrene insertion products is mainly governed by charge dispersion rather than charge alternation principles. (31) Postils, V.; Company, A.; Solà, M.; Costas, M.; Luis, J. M. Inorg. Chem. 2015, 54, 8223−8236. (32) Paradine, S. M.; White, M. C. J. Am. Chem. Soc. 2012, 134, 2036−2039.

I

DOI: 10.1021/acs.organomet.7b00116 Organometallics XXXX, XXX, XXX−XXX