Aliphatic C(sp3)–H Bond Activation Using Nickel Catalysis

Aug 15, 2017 - Topological analysis of electron density using atoms in molecule analysis are provided in Figure S6 and Table S4 in the Supporting Info...
0 downloads 7 Views 2MB Size
Article pubs.acs.org/joc

Aliphatic C(sp3)−H Bond Activation Using Nickel Catalysis: Mechanistic Insights on Regioselective Arylation Sukriti Singh, Surya K, and Raghavan B. Sunoj* Department of Chemistry, Indian Institute of Technology Bombay, Powai, Mumbai 400076, India S Supporting Information *

ABSTRACT: Transition-metal-catalyzed C(sp3)−H bond activation in aliphatic compounds are of current interest. Lack of mechanistic insights on Ni-catalyzed C(sp3)−H activation using 8-aminoquinoline as a directing group motivated us to examine an interesting direct arylation of an aliphatic tertiary amide by using density functional theory. The catalysis employed Ni(II) precatalyst, 4-iodoanisole as an arylating agent, sodium carbonate, and mesitylenic acid as additives in DMF solvent. Examination of a comprehensive set of mechanistic pathways helped us learn that the most preferred route begins with a bidentate chelate binding of deprotonated substrate to the Ni. The C−H activation in the catalyst− substrate complex via a cyclometalation deprotonation provides a fivemembered nickelacycle intermediate, which upon the rate-limiting oxidative insertion to aryl iodide forms a Ni(IV)−aryl intermediate. The ensuing reductive elimination furnishes the desired arylated product. We note that the explicit inclusion of sodium carbonate, mesitylenic acid, and solvent molecules on sodium ion all are critical in identifying the most favorable pathway. Of the two types of C(sp3)−H bonds in the substrate [2-methyl-2phenyl-N-(quinolin-8-yl)heptanamide], the energies for the regiocontrolling reductive elimination is predicted to be more in favor of the methyl group than the methylene of the pentyl chain, in excellent agreement with the previous experimental observation.



reactions.2 Even though the exact role of this bidentate directing group is not firmly established, it is generally proposed that 8-AQ could stabilize various high-valent metal intermediates formed during the course of the reaction. Although there have been considerable developments in catalytic C(sp2)−H bond functionalization reactions, the direct and site selective functionalization of C(sp3)−H bonds is relatively underdeveloped.7 This is mainly due to the inherently poor acidity and lack of stabilizing orbital interactions between the metal center and such C(sp3)−H bonds. Most of the C(sp3)−H activation reactions bank on the use of palladium catalysts, while the effort toward the employing other transition metals continues to gain popularity.8 One such endeavor has been to employ more cost-effective Ni catalysts,9 which are known to lower the propensity for β-hydride elimination in addition to its ability to bind to carbon−carbon double and triple bonds. In a very recent study, the Chatani group reported the first bidentate directing group assisted Ni-catalyzed C(sp3)−H bond functionalization of unactivated bonds (Scheme 1).10 The reaction provides an important protocol for functionalization at β-C(sp3)−H bonds in a highly regioselective manner. The method can be applied to carboxylic acids in a three-step process such as amidification, C−H functionalization, and

INTRODUCTION Catalytic methods for direct functionalization of C−H bonds have garnered considerable interest as they can shorten synthetic protocols to a desired target.1 The transition-metalmediated C−H bond activation reaction revealed itself as a robust technique in organic chemistry owing to the possibility of combining a limitless pool of simpler starting materials.2 However, the very fact that multiple C−H bonds are typically present in a given reactant can result in indiscriminate functionalization and can thus hamper selectivity in such reactions. The different approaches for site-selective functionalization generally involve the use of specific ligands on transition metals or substrate modifications with a suitable directing group.3 The problem of attaining the desired regioselectivity in C−H activation reactions has, in part, been addressed by the use of a directing group, which is a functional group that can coordinate to the transition metal in such a way that it directs the metal center toward the desired C−H bond.2 Some of the popular examples include weakly coordinating directing groups by the Yu group,4 bidentate directing groups by Daugulis and coworkers, and easily removable or “traceless” directing groups by Chatani and Tobisu.5 While the repertoire of directing groups continues to expand,2 8-aminoquinoline (8-AQ) appears to be an interesting choice owing to its compatibility with a range of transition metals and substrates.6 8-AQ has been used in stereoselective as well as regioselective C−H functionalization © 2017 American Chemical Society

Received: July 4, 2017 Published: August 15, 2017 9619

DOI: 10.1021/acs.joc.7b01672 J. Org. Chem. 2017, 82, 9619−9626

Article

The Journal of Organic Chemistry

Scheme 1. Ni-Catalyzed Direct Arylation of C(sp3)−H Bonds with 4-Iodoanisole in Aliphatic Amides Using 8-Aminoquinoline as the Bidentate Directing Group

our computations in the condensed phase. The discussions in the text employ Gibbs free energies (G298K) obtained by incorporating thermal and entropic terms to the electronic energies in the condensed phase. The ionic salts involved in the reaction were microsolvated using explicit DMF molecules so as to capture the specific interaction between the ion and solvent molecules. The limitation of the continuum solvation model in capturing the short-range interactions could be partly overcome by considering explicitly bound solvent molecules.22 The Gibbs free energies of such salt-solvent clusters in the solvent continuum were evaluated using the SMD continuum solvation model to arrive at an optimum number of DMF molecules likely to be present in the first shell of solvation as described in the Supporting Information.23 The energy span was calculated using the energetic span model developed by Shaik and Kozuch that expresses the turnover frequency (TOF) of a catalytic cycle using the energetics of various stationary points such as the turnover-determining intermediate (TDI) and transition state (TDTS).24 Topological analysis of the electron density distribution was performed using Bader’s atoms in molecules (AIM) formalism by using AIM2000 software25 with the wave function generated at the SMD(DMF)/B3LYP-D3/SDD(Ni,I),6-31G** level of theory. This analysis was employed for identifying weak interatomic interactions within a given transition-state structure for the regioselective C−C bond-formation step of the reaction. In addition, we have used a distortion−interaction model to probe the origin of regioselectivity in this reaction.26

subsequent removal of the auxiliary. This seminal report triggered a good number of examples in the years that followed.11 In keeping with our current research efforts toward understanding the mechanism of C−H functionalization reactions,12 we became interested in examining the mechanism of nickel-catalyzed β-arylation of an aliphatic tertiary amide (R1) using an aryl iodide (R-2). We have employed density functional theory (B3LYP-D3) method to examine the mechanism of this reaction. Previous mechanistic studies on Ni-catalyzed reactions have suggested that the catalytic cycle could involve various species with oxidation states ranging from Ni(0) to Ni(III).13 However, there have only been a few reports suggesting the participation of Ni(IV) intermediates.14 In recent years, Pd(II)/Pd(IV) redox manifold in analogous Pd-catalyzed C−H bond functionalization reactions has been demonstrated as significant in the presence of most common oxidants such as hypervalent iodines and other electrophilic halogenating reagents.15 Interestingly, the Sanford group recently reported the synthesis and characterization of some Ni(IV) complexes.15 Surprisingly, mechanistic studies seldom invoke Ni(IV) intermediates in catalytic cycles. Herein, we wish to report (a) the first detailed mechanistic study of Ni-catalyzed C(sp3)−H bond functionalization assisted by 8-aminoquinoline bidentate directing group by explicitly considering a Ni(II)/Ni(IV) redox manifold,16a (b) the role of additives such as mesityl carboxylic acid and sodium carbonate in the catalytic cycle, and (c) the origin of the regioselective C−H arylation between a methyl and methylene positions on the substrate.





RESULTS AND DISCUSSION

The reaction involves a direct arylation of one of the C(sp3)−H bonds of the substrate by the action of 4-iodoanisole under the catalytic condition as broadly described in Scheme 1. The precatalyst nickel triflate Ni(OTf)2 is employed in this reaction in conjunction with mesitylenic acid additive (denoted as MesCOOH). Under such one-pot reaction conditions with DMF as the solvent, several likely active catalytic species could be envisaged. Hence, we have first investigated various possible ligand combinations around the central nickel atom. One such possibility is to consider another nickel triflate species generated by the uptake of two molecules of neutral carboxylic acid ligands (MesCOOH) by changing the coordination mode of the native triflate on nickel from κ-2 to κ-1. The Gibbs free energy of formation of Ni(OTf)2(MesCOOH)2 from Ni(OTf)2 is found to be about −25 kcal/mol. The Ni(II) complex Ni(OTf)2(MesCOOH)2 (C1), with a trans disposition between triflate and carboxylic acid, appears to be the most probable active catalytic species.16b For improved clarity in discussions, the key mechanistic events are organized into five consecutive steps, which are described in the following sections. Formation of Catalyst−Substrate Complex. The substrate, tethered with an 8-aminoquinoline directing group, can exhibit different modes of coordination to Ni. For instance, it can bind through the amido nitrogen, either in its native form as a neutral amide or as an amidate anion (generated by the action of sodium carbonate through amide deprotonation, the

COMPUTATIONAL DETAILS

The Gaussian 09 program was employed for all the calculations in this study.17 All the stationary points such as reactants, intermediates, and transition states were optimized in the condensed phase using dispersion-corrected hybrid density functional B3LYP-D318 with the 6-31G** basis set19 for all atoms except for nickel and iodine. For Ni and I, the SDD basis set with an effective core potential (ECP) was used.20 For nickel atom, 10 core electrons were represented using an ECP while for iodine ECP represents 46 core electrons, respectively, leaving 18 and 7 valence electrons which were explicitly treated using standard basis sets. Frequency calculations on all of the stationary points were carried out to characterize the nature of those stationary points and also to evaluate the respective molecular partition functions and entropic terms. The transition states were characterized by a unique imaginary frequency, characteristic of first-order saddle points on the potential energy surface, and found to pertain to the desired reaction coordinate. Intrinsic Reaction Coordinate (IRC) calculations were further performed on the transition-state geometries thus obtained to ascertain that the transition state connected to reactants and products on either side of the first-order saddle point. The effect of solvent was incorporated using the continuum solvation model SMD developed by Truhlar and Cramer.21 The solvent used in the reaction is N,N-dimethylformamide (DMF), and hence, we have used the continuum dielectric of DMF (ε = 36.7) in 9620

DOI: 10.1021/acs.joc.7b01672 J. Org. Chem. 2017, 82, 9619−9626

Article

The Journal of Organic Chemistry

Scheme 2. Most Favorable Pathway for the Formation of Catalyst−substrate Complex 2 and Corresponding Relative Gibbs Free Energies (in kcal/mol) in Parentheses

sodium bicarbonate (generated in the formation of 2) to form another intermediate 4, which provides the most conducive situation for the C−H activation step, as described in detail in the next section. A nickelacycle 4a thus generated can now interact with 4-iodoanisole, initially in the form of a prereacting complex 5 formed through the expulsion of sodium triflate. An important oxidative insertion to the Ar−I bond results in a Ni(IV) intermediate 6 via a transition state denoted as [5−6]⧧. A ligand exchange wherein nickel-bound iodide is substituted by triflate provided by sodium triflate (leading to 6a, not explicitly shown in this scheme) sets the stage for reductive elimination. The reductive elimination in intermediate 6a results in arylation of the methyl group leading to a product Ni(II) intermediate 7. At this state, regeneration of the active catalyst 2 as well as the formation of final product is considered by the action of carbonic acid (generated earlier in 4 to 4a conversion). In connection with the next step, a prereacting complex 7a between the carbonic acid and 7 is identified. In [7a-8]⧧, carbonic acid protonates the amidate nitrogen to furnish the product-catalyst complex 8. Uptake of an unreacted substrate b and a neutral MesCOOH ligand regenerates 2 to sustain the catalytic cycle. More molecular details and energetics associated with each of these steps are provided in the following sections. C−H Activation. In view of various additives such as MesCOOH and Na2CO3 used in conjunction with Ni(OTf)2 in DMF, we envisaged several ligand-exchange possibilities in each step of the catalytic cycle so as to identify the energetically most preferred mechanistic route.29 While we focus on the most favorable C−H activation pathway here, an interested reader can find the details of several other possibilities in the Supporting Information.28a To begin with, a concerted metalation−deprotonation (CMD) pathway in the catalyst− substrate complex 2 is considered wherein the C−H activation is facilitated either by the native Ni-bound triflate or by a bicarbonate ligand (generated through ligand exchange).28b The displacement of the triflate by a bicarbonate ligand is found to be energetically favorable by 13 kcal/mol. The various modes for ligand-assisted C−H activation transition states, generally denoted as [4−4a]⧧, are considered. Among these, the lowest energy TS is identified as consisting of a combination of bicarbonate and triflate in C−H activation.28a Certain key features can be noticed from the optimized geometry of this CMD TS as shown in Figure 1. First, the additive sodium bicarbonate is directly involved in the deprotonation of the methyl group as an external ligand. The Ni-bound triflate maintains a hydrogen-bonding interaction with the external bicarbonate.12d,28a Second, we have

relative Gibbs free energy for the deprotonation is found to be 3.4 kcal/mol). The most favorable mode of substrate binding to the catalyst is found to be the one when the amidate nitrogen binds to nickel, the other ligands being triflate and MesCOOH as shown in Scheme 2.16c Alternatively, the neutral substrate can bind to the Ni(II) center in C1 as an amide by displacing one of the neutral MesCOOH ligands. Subsequently, the Nibound triflate can abstract the amide proton and depart as a triflic acid. This will enable the coordination of the pyridine nitrogen of the directing group to the Ni center and lead to the formation of catalyst−substrate complex 2. However, this route is found to be of higher energy. This situation suggests that one of the important roles of the external base (Na2CO3) is to facilitate the formation of the catalyst−substrate complex (2) by deprotonating the amido N−H bond.27 In the following sections, we discuss reaction pathways with the catalyst− substrate complex 2 as the starting point of the catalytic cycle. While several other possibilities were considered in this study, herein we present only the most favorable mechanistic pathway.28 The formation of the catalyst-substrate complex 2 can now trigger a cascade of significant steps as shown in Scheme 3. We describe certain interesting aspects of the catalytic cycle here. In 2, the amidate and quinolone nitrogen atoms of the substrate act as a chelate to the Ni(II) center while a triflate and MesCOOH occupy the remaining two coordination sites. As one can readily note, there are two types of C−H bonds on the substrate that could be activated by the proximal nickel center; one is of the methyl group and the other one of the internal methylene of the pentyl chain. The optimized geometry of the lowest energy conformer of 2 reveals that the C−H bond of the methyl group and the methylene of the pentyl chain are at, respectively, 2.78 and 2.73 Å from the Ni(II) center (These distances are between the corresponding H atoms to the Ni center.) In this section, the C−H activation of only the methyl group is presented as we wish to make an explicit comparison between methyl and pentyl and on the regiochemical preference toward the end of this manuscript. General Description of the Catalytic Cycle. As shown in Scheme 3, the catalytic event can be considered to begin with the catalyst−substrate chelate denoted as intermediate 2. The proximity of the methyl C−H bond in 2 can be improved by displacing the neutral ligand MesCOOH to give another intermediate 3. We have also optimized a few more additional geometric possibilities for intermediate 3.28a The C−H activation in 3 would be more facile as the partially open coordination site at the metal center can exhibit enhanced interaction with the C−H bond. We considered uptake of 9621

DOI: 10.1021/acs.joc.7b01672 J. Org. Chem. 2017, 82, 9619−9626

Article

The Journal of Organic Chemistry

Scheme 3. Full Mechanistic Cycle Shown Using the Most Favorable Pathway Identified on the Basis of Gibbs Free Energies for Ni-Catalyzed β-C(sp3)−H Arylationa

a

Alternative mechanistic possibilities are shown in Scheme S3 and Figure S1 in the Supporting Information.

Oxidative Addition and C−C Reductive Coupling. We focus on a Ni(II)−Ni(IV) redox pathway in this study on the basis of the following factors. There have been proposals on the potential involvement of Ni(II)/Ni(IV) redox manifold in certain C−H bond activation reactions and experimental evidence on Ni(IV) complexes,14 and the directing group 8AQ employed in this reaction is known to stabilize high-valent transition-metal complexes. At this stage of the mechanism, the oxidative addition of the nickelacycle intermediate 4a to 4iodoanisole (Ar−I) is considered. Through a desirable ligand exchange, the nickel-bound triflate gets displaced (in the form of sodium triflate) by the action of Ar−I to form 5. In 5, the Ar−I bond is closer to the Ni(II) center, and an oxidative

considered explicit solvation of the sodium ion in the transition state by using three solvent molecules (DMF).23 It should be noted that sodium ion in sodium bicarbonate is more likely to maintain a tetracoordinate environment in DMF solvent, particularly when it remains in the vicinity of the substrate as shown in Figure 1. The relative Gibbs free energy of [4−4a]⧧ was found to be −5.7 kcal/mol. The TSs with only triflate or only bicarbonate in the CMD process are, respectively, 32.3 and 4.6 kcal/mol higher. Relatively higher basicity of bicarbonate as compared to triflate appears to have a direct role in providing additional stabilization to the corresponding CMD TSs. 9622

DOI: 10.1021/acs.joc.7b01672 J. Org. Chem. 2017, 82, 9619−9626

Article

The Journal of Organic Chemistry

Figure 1. Optimized geometries and corresponding relative Gibbs free energies (in parentheses, kcal/mol) of the transition state for the C−H bond activation of the methyl [4−4a]⧧ and the methylene of the pentyl [4′-4a′]⧧ groups. Distances are shown in Å. Only selected hydrogen atoms are shown for improved clarity.

Figure 2. Optimized geometries and corresponding relative Gibbs free energies (given parentheses, kcal/mol) of the important transition states. Distances are shown in Å. Only selected hydrogen atoms are shown for improved clarity.

addition via the transition state [5−6]⧧ can now give a pentacoordinate Ni(IV) intermediate 6. The geometric features of [5−6]⧧ can be gleaned from Figure 2. The relative Gibbs free energy of the oxidative addition transition state is 13.3 kcal/mol, and the corresponding activation barrier for the elementary step is 14.7 kcal/mol (Figure 3). To set the stage for the most important step, namely, the arylation of the methyl group via a reductive elimination, we examined couple of likely scenarios: (i) a direct reductive elimination30 from the iodide-bound intermediate 6 through [6−6b]⧧ or (ii) a substitution of iodide first by a triflate and an ensuing reductive elimination through [6a−7]⧧. The Gibbs free

energies of the C−C bond formation transition states [6−6b]⧧ and [6a−7]⧧, respectively, with iodide and triflate ligand on Ni(IV) are found to be 5.2 and 0.4 kcal/mol, suggesting that triflate is a better ligand than iodide in the reductive elimination step. Similarly, the corresponding activation barriers for [6− 6b]⧧ and [6a−7]⧧ computed with respect to the preceding intermediate are 6.6 and 4.1 kcal/mol, respectively.30a The Ni(II) complex 7 obtained at the end of the reductive elimination exhibits a Ni−η2-phenyl interaction as shown in Scheme 3. The reaction scheme from intermediate 5 to 7 involves the participation of Ni(II) and Ni(IV) intermediates, which represents a relatively rare example of computational 9623

DOI: 10.1021/acs.joc.7b01672 J. Org. Chem. 2017, 82, 9619−9626

Article

The Journal of Organic Chemistry

Figure 3. Gibbs free energy profile for the lower energy pathway.

In addition, participation of ionic salts (Na2CO3, NaHCO3, NaOTf, and NaI) is considered in various steps of the reaction. We have investigated the mechanism using two models.23b One model is with explicit DMF molecules (3 or 4) bound to the sodium ion as shown in Figure 1 and another one without any explicit solvent molecules. The Gibbs free energies of all these stationary points have been computed in a continuum dielectric of DMF using the SMD solvation model (see the Computational Details). On average, the Gibbs free energies are found to be 8 kcal/mol lower due to microsolvation of sodium ion with explicit DMF molecules. The model with explicit DMF molecules predicted the most preferred transition state for C−H activation as [4−4a⧧]. In this case, the ligand involved in the CMD transition state is a bicarbonate, which is not directly bound to the nickel center. It is noticed that the removal of the bound DMF molecules makes this CMD transition state higher in energy as compared to the transition state wherein a nickelbound bicarbonate is involved in the CMD process.28a,23b

demonstration for Ni(II)/Ni(IV) redox manifold in Nicatalyzed C−H activation and other C−C coupling reactions. The arylated intermediate 7 formed in the reductive elimination is still an amidate and hence requires protonation to form the final product. The intermediate 7 can first interact with the H2CO3, formed during the earlier C(sp3)−H activation step, through a hydrogen-bonding interaction. In such a hydrogen-bonded complex, the acidic proton from the carbonic acid can protonate the amidate nitrogen coordinated to the Ni(II) center. The barrier to protonation through [7a−8]⧧ is found to be 1.7 kcal/mol (Figure 3). The resulting species 8 is essentially a complex between the product and the transition metal.30b The regeneration of the active species 2 occur with the replacement of nickel bound product (P-1) and bicarbonate in intermediate 8 with amidate b and MesCOOH. The overall reaction can be considered as R-1 + Ar−I + Na2CO3(DMF)6 → P-1 + NaI(DMF)3 + NaHCO3(DMF)3. The Gibbs free energy for this transformation is found to be exergonic by −45.2 kcal/mol.31 The full Gibbs free energy profile for the most favorable pathway is provided in Figure 3. The efficiency of the computed catalytic cycle is evaluated by using the energetic span model as described in the Computational Details. It can be noticed that the TDI and the TDTS are, respectively, intermediate 4 and the oxidative addition transition state [5− 6]⧧. The δE for the catalytic cycle is therefore 33.5 kcal/mol.24b Using this value of the activation span, the theoretical turnover frequency (TOF) is 1.6 × 10−5 s−1, which is broadly similar to the TOF reported for related catalytic reactions.24c A closer look at the energy profile suggests that a number of exoergic intermediates are involved in the catalytic pathway, and hence, those may become detectable with delegent experimentation. Interestingly, the active catalyst 2 as well as intermediate 6a have been experimentally characterized previously.14 Importance of Explicit Solvent Molecules. The implicit solvation model provides an inadequate treatment of shortrange interactions between an ion and coordinating solvents. The catalytic pathways considered in this study involve intermediates and transition states with a bound sodium ion.



REGIOSELECTIVITY

As mentioned previously, it is of interest to compare the relative abilities of the two β-C(sp3)−H bonds in the substrate toward the arylation reaction. As noted in the previous section, in the lowest energy intermediate (4) prior to the C−H activation transition state, the methyl C−H bond is closer (1.75 Å) to the nickel than the internal methylene C−H of the pentyl chain (4.22 Å). These distances are between the corresponding H atoms to the Ni center. However, in another nearly degenerate conformer (4′), the pentyl C−H could be brought closer to the nickel. The optimized geometries as well as the relative Gibbs free energies of the CMD transition states for C(sp3)−H activation for methyl and pentyl positions are shown in Figure 1. Interestingly, the CMD transition state for the C(1)−H bond of methyl is found to be 1.6 kcal/mol higher than for the C(1′)−H bond of the pentyl chain. While an energetic preference of this order suggests the formation of a major product with C(1) arylation, it is not adequate to exclude C(1′) arylation. The experimentally observed exclusive regioselectivity for C(1) arylation might therefore be arising 9624

DOI: 10.1021/acs.joc.7b01672 J. Org. Chem. 2017, 82, 9619−9626

Article

The Journal of Organic Chemistry

Explicit solvation of the sodium ion by using DMF molecules (microsolvation) has been found to be critical in locating the lowest energy cyclometalation deprotonation transition state. The rate-limiting step is found to the oxidative insertion to the Ar−I bond to give a Ni(IV)−aryl intermediate. The regioselectivity is controlled by the energetic preference in the reductive elimination transition state. Of the two potential sites for arylation via C(sp3)−H bond activation, namely a methyl and a methylene of a pentyl chain, we note a distinct energetic advantage (∼5 kcal/mol) for arylation of the methyl position in the reductive elimination transition state. The predicted regioselectivity is in line with the experimental observation reported previously.

due to another mechanistically significant step in the catalytic cycle. Perusal of the Gibbs free energy profile (Figure 3) conveys that the CMD step (4 to 4a, via [4−4a]⧧) is likely to be a reversible step as compared to the reductive elimination (6a to 7) leading to arylation. Hence, the transition states for reductive elimination for both C(1) and C(1′) arylations are located. The relative Gibbs free energy of the transition state for pentyl C(1′) arylation is found to be 5.3 kcal/mol higher than that for the methyl C(1) arylation. Similarly the activation barrier for the pentyl C(2) arylation is 6.8 kcal/mol higher than that for the methyl C(1) position. The transition state [6a-7]⧧ for the methyl C(1) arylation exhibited relatively more number of efficient noncovalent interactions These interactions, as shown in the optimized geometries given in Figure 2B, are O··· H−C (a, b, e, f), C−H···π (g), π···π (c), and F···H−C (h, j) more effective in [6a-7]⧧ as compared to the noncovalent interactions noticed in [6a′−7′]⧧ for pentyl C(1′) arylation (Figure 2D).25d In other words, more efficient noncovalent interactions lead to additional stabilization of the C(1) arylation transition-state. To gain additional understanding on the regiochemical preference, we have analyzed the regiocontrolling reductive elimination transition states by using the Activation Strain model.26 It is noted that the total distortion in the reactants is 3.3 kcal/mol lower in [6a−7]⧧ as compared to that in [6a′− 7′]⧧. The interaction energy between the distorted fragments is found to be more favorable by 5.5 kcal/mol in [6a−7]⧧ than that in [6a′−7′]⧧. In other words, both interaction and distortion energies contribute to the overall activation strain energy, which is higher in the case of [6a′−7′]⧧.26d This prediction that arylation at the methyl C−H bond is energetically more favored offers a firmer basis for the exclusive regiochemical preference noted under the experimental conditions. In a very recent parallel investigation, the Liu group examined two-electon oxidative addition and a one-electron oxidation/radical pathway for this reaction.32 In Liu’s study, various C(sp2)−H and C(sp3)−H bond activation reactions were discussed, while our paper presents detailed mechanistic insights of a C(sp3)−H arylation reaction. Although conclusions of both the studies suggests a Ni(II)/Ni(IV) catalytic cycle, the choice of active catalysts are different. As described in the earlier sections of this manuscript, we have explicitly considered the role of added ligand, additives and the importance of solvation. Focus on the regioselectivity aspects is another key point of our study.



ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.joc.7b01672. Cartesian coordinates of all stationary points (reactants, products, intermediates, and transition states) and alternative higher energy pathways and other relevant information (PDF)



AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Raghavan B. Sunoj: 0000-0002-6484-2878 Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS The SpaceTime supercomputing facility at IIT Bombay is gratefully acknowledged for providing generous computing time.



REFERENCES

(1) (a) Li, B.; Dixneuf, P. H. Chem. Soc. Rev. 2013, 42, 5744. (b) Wencel-Delord, J.; Glorius, F. Nat. Chem. 2013, 5, 369. (c) Yang, X.; Shan, G.; Wang, L.; Rao, Y. Tetrahedron Lett. 2016, 57, 819. (2) Rouquet, G.; Chatani, N. Angew. Chem., Int. Ed. 2013, 52, 11726. (3) Neufeldt, S. R.; Sanford, M. S. Acc. Chem. Res. 2012, 45, 936. (4) (a) Chen, X.; Engle, K. M.; Wang, D.-H.; Yu, J.-Q. Angew. Chem., Int. Ed. 2009, 48, 5094. (b) Engle, K. M.; Mei, T.-S.; Wasa, M.; Yu, J.Q. Acc. Chem. Res. 2012, 45, 788. (5) (a) Zaitsev, V. G.; Shabashov, D.; Daugulis, O. J. Am. Chem. Soc. 2005, 127, 13154. (b) Kinuta, H.; Tobisu, M.; Chatani, N. J. Am. Chem. Soc. 2015, 137, 1593. (6) (a) Chen, K.; Shi, B.-F. Angew. Chem., Int. Ed. 2014, 53, 11950; Angew. Chem. 2014, 126, 12144. (b) Deb, A.; Bag, S.; Kancherla, R.; Maiti, D. J. Am. Chem. Soc. 2014, 136, 13602. (c) Monks, B. M.; Fruchey, E. R.; Cook, S. P. Angew. Chem., Int. Ed. 2014, 53, 11065; Angew. Chem. 2014, 126, 11245. (7) (a) Daugulis, O.; Do, H.-Q.; Shabashov, D. Acc. Chem. Res. 2009, 42, 1074. (b) Jazzar, R.; Hitce, J.; Renaudat, A.; Sofack-Kreutzer, J.; Baudoin, O. Chem. - Eur. J. 2010, 16, 2654. (c) Lyons, T. W.; Sanford, M. S. Chem. Rev. 2010, 110, 1147. (8) Hartwig, J. F. Chem. Soc. Rev. 2011, 40, 1992. For a recent review on dehydrogenation of C(sp3)−H bonds, see: (a) Dobereiner, G. E.; Crabtree, R. H. Chem. Rev. 2010, 110, 681. (b) Haibach, M. C.; Kundu, S.; Brookhart, M.; Goldman, A. S. Acc. Chem. Res. 2012, 45, 947. (c) Gunanathan, C.; Milstein, D. Science 2013, 341, 1229712. (9) Ananikov, V. P. ACS Catal. 2015, 5, 1964. (10) Aihara, Y.; Chatani, N. J. Am. Chem. Soc. 2014, 136, 898.



CONCLUSION The DFT(B3LYP-D3) investigations on a Ni-catalyzed direct arylation of C(sp3)−H bonds of an aliphatic tertiary amide tethered with a bidentate 8-aminoquinoline directing group, by using 4-iodoanisole as the arylating agent, revealed several interesting insights. Computed Gibbs free energies indicated that the C(sp3)−H arylation reaction proceed through a catalytic cycle involving a rare Ni(II)/Ni(IV) redox couple. Explicit inclusion of additives, such as sodium carbonate and mesitylenic acid, in the intermediates and transition states involved in the catalytic pathway has been found to be essential toward identifying the lower energy reaction pathway. Whereas sodium carbonate helps in deprotonation of the amidic N−H group in substrate activation in the initial state of the reaction, mesitylenic acid acts as an effective ligand in the active catalyst. 9625

DOI: 10.1021/acs.joc.7b01672 J. Org. Chem. 2017, 82, 9619−9626

Article

The Journal of Organic Chemistry (11) (a) Liu, Y.; Zhang, Z.; Yan, S.; Liu, Y.; Shi, B. Chem. Commun. 2015, 51, 7899. (b) Wang, X.; Qiu, R.; Yan, C.; Reddy, V. P.; Zhu, L.; Xu, X.; Yin, S. Org. Lett. 2015, 17, 1970. (c) Wu, X.; Zhao, Y.; Ge, H. J. Am. Chem. Soc. 2015, 137, 4924. (12) (a) Athira, C.; Sunoj, R. B. Org. Biomol. Chem. 2017, 15, 246. (b) Patel, C.; Abraham, V.; Sunoj, R. B. Organometallics 2017, 36, 151. (c) Anand, M.; Sunoj, R. B.; Schaefer, H. F. ACS Catal. 2016, 6, 696. (d) Anand, M.; Sunoj, R. B. Org. Lett. 2012, 14, 4584. (13) (a) Tang, H.; Zhou, B.; Huang, Xu-Ri; Wang, C.; Yao, J.; Chen, H. ACS Catal. 2014, 4, 649. (b) Tang, H.; Huang, Xu-Ri; Yao, J.; Chen, H. J. Org. Chem. 2015, 80, 4672. (c) Xu, Z.-Y.; Jiang, Y.-Y.; Yu, H.-Z.; Fu, Y. Chem. - Asian J. 2015, 10, 2479. (14) (a) Camasso, N. M.; Sanford, M. S. Science 2015, 347, 1218. (b) Chong, E.; Kampf, J. W.; Ariafard, A.; Canty, A. J.; Sanford, M. S. J. Am. Chem. Soc. 2017, 139, 6058. (15) (a) Sehnal, P.; Taylor, R. J. K.; Fairlamb, I. J. S. Chem. Rev. 2010, 110, 824. (b) Hickman, A. J.; Sanford, M. S. Nature 2012, 484, 177. (c) Canty, A. J. Dalton Trans. 2009, 47, 10409. (16) (a) See Scheme S5 in the Supporting Information for details of additional possibilities considered.. (b) See Figure S1 in the Supporting Information for details of additional possibilities of active catalysts.. (c) The amidate binding is favorable over charge neutral Ndonor amide by about 33 kcal/mol. For details, see Scheme S1 in the Supporting Information.. (17) (a) Gaussian 09, revision D.01: 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.; Keith, T.; 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, O.; Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian, Inc., Wallingford, CT, 2013. (b) For the additional validation, we performed computations also with (i) B3LYP-D3 in combination with the PCM solvation model and (ii) M06 with the SMD solvation model. The results were similar to what is noted here in the main manuscript. See Tables S7 and S8 in the Supporting Information for additional details.. (18) (a) Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B: Condens. Matter Mater. Phys. 1988, 37, 785. (b) Grimme, S.; Antony, J.; Ehrlich, S.; Krieg, H. J. Chem. Phys. 2010, 132, 154104. (19) (a) Hariharan, P. C.; Pople, J. A. Theor. Chim. Acta 1973, 28, 213. (b) Hehre, W. J.; Ditchfield, R.; Pople, J. A. J. Chem. Phys. 1972, 56, 2257. (20) Hay, P. J.; Wadt, W. R. J. Chem. Phys. 1985, 82, 299. (21) Marenich, A. V.; Cramer, C. J.; Truhlar, D. G. J. Phys. Chem. B 2009, 113, 6378. (22) Sunoj, R. B.; Anand, M. Phys. Chem. Chem. Phys. 2012, 14, 12715. (23) (a) See Table S1 and Figure S2 in the Supporting Information for additional details on different explicit solvation situations and how we identified three DMF molecules as most likely in the present case.. (b) See Figure S5 and Table S2 in the Supporting Information for additional details on the effect of explicit solvation in the energy of the stationary points.. (24) (a) Kozuch, S. WIREs Comput. Mol. Sci. 2012, 2, 795. (b) See Table S3 in the Supporting Information for a compilation of various possible ways to determine the maximum δE for the catalytic cycle presented here.. (c) Kozuch, S.; Lee, S.E.; Shaik, S. Organometallics 2009, 28, 1303.

(25) (a) AIM2000 version 2.0; Buro fur Innovative Software, SBKSoftware, Bielefeld, Germany, 2002. (b) Bader, R. F. W. Chem. Rev. 1991, 91, 893. (c) Biegler-Konig, F.; Schonbohm, J.; Bayles, D. J. Comput. Chem. 2001, 22, 545. (d) Topological analysis of electron density using atoms in molecule analysis are provided in Figure S6 and Table S4 in the Supporting Information.. (26) (a) Fernández, I.; Bickelhaupt, F. M. Chem. Soc. Rev. 2014, 43, 4953. (b) Wolters, L. P.; Bickelhaupt, F. M. WIREs Comput. Mol. Sci. 2015, 5, 324. (c) Bickelhaupt, F. M.; Houk, K. N. Angew. Chem., Int. Ed. 2017, 56, 10070. (d) See Figure S7 and Table S5 in the Supporting Information for more details on how the transition state is partitioned into fragments in Activation Strain analysis.. (27) (a) Wang, L.; Huang, J.; Peng, S.; Liu, H.; Jiang, X.; Wang, J. Angew. Chem., Int. Ed. 2013, 52, 1768. (b) Dang, Y.; Qu, S.; Nelson, J. W.; Pham, H. D.; Wang, Z. X.; Wang, X. J. Am. Chem. Soc. 2015, 137, 2006. (28) (a) See Scheme S2 and Figure S3 in the Supporting Information for details of additional higher energy pathways for C-H activation.. (b) We have earlier reported a similar transition state in CH activation reaction under protic conditions: Anand, M.; Sunoj, R. B. Organometallics 2012, 31, 6466. (29) Jindal, G.; Sunoj, R. B. J. Am. Chem. Soc. 2014, 136, 15998. (30) (a) See Scheme S4 and Figure S4 in the Supporting Information for details of additional higher energy reductive elimination transition state.. (b) See Scheme S3 in the Supporting Information for additional details on oxidative addition, reductive elimination, and product formation.. (31) The Gibbs free energies of various species involved in this equation are provided in Table S6 in the Supporting Information. (32) While this paper was under review, another paper appeared reporting similar results. See: Omer, H. M.; Liu, P. J. Am. Chem. Soc. 2017, 139, 9909.

9626

DOI: 10.1021/acs.joc.7b01672 J. Org. Chem. 2017, 82, 9619−9626