Article Cite This: J. Org. Chem. 2018, 83, 2067−2076
pubs.acs.org/joc
Mechanistic Unveiling of CC Double-Bond Rotation and Origins of Regioselectivity and Product E/Z Selectivity of Pd-Catalyzed Olefinic C−H Functionalization of (E)‑N‑Methoxy Cinnamamide Lingjun Liu, Guojing Pei, Peng Liu, Baoping Ling, Yuxia Liu,* and Siwei Bi* School of Chemistry and Chemical Engineering, Qufu Normal University, Qufu 273165, P.R. China S Supporting Information *
ABSTRACT: Density functional theory (DFT) calculations have been performed to study the Pd-catalyzed CH functionalization of (E)-N-methoxy cinnamamide (E1), which selectively provides the α-CH activation products (EP as minor product and its CC rotation isomer ZP′ as major product). Three crucial issues are solved: (i) The detailed mechanism leading to ZP′ is one issue. The computational analyses of the mechanisms proposed in previously experimental and theoretical literature do not seem to be consistent with the experimental findings due to the high barriers involved. Alternatively, we present a novel oxidation/reduction-promoted mechanism featuring the Pd(0) → Pd(II) → Pd(0) transformation. The newly proposed mechanism involves the initial coordination of the active catalyst PdL2 (L = t-BuCN) with the CC bond in EP, followed by the oxidative cyclization/reductive decyclization-assisted CC double-bond rotation processes resulting in ZP′ and regeneration of PdL2. (ii) The origin of the product E/Z selectivity is the second issue. On the basis of the calculated results, it is found that, at the initial stage of the reaction, EP is certainly completely generated, while no ZP′ formation occurred. Once E1 is used up, EP immediately acts as the partner of the new catalytic cycle and sluggishly evolves into ZP′. A small amount of generated ZP′ would reversibly transform to EP due to the higher barrier involved. (iii) The intrinsic reasons for the regioselectivity are the third issue. The calculated results indicate that the regioselectivity for αCH activation is mainly attributed to the stronger electrostatic attraction between the α-C and the metal center.
1. INTRODUCTION Small four-membered heterocycles, particularly β-lactams, as one of the most important structural motifs, have increasingly obtained widespread concerns.1 As the functionalized skeletons, the β-lactam rings are widely found in various broad-spectrum antibiotic compounds defeating formerly lethal diseases such as the plague, wound sepsis, and tuberculosis.2 Apart from the pharmacological significance, β-lactams, acting as reactive synthons, are also extensively engaged in organic synthesis.3 These intrinsic potentials stimulate the continuing evolution of new synthetic methodologies for the construction of β-lactams. Generally, conventional synthetic methods include (a) [2+2]cycloadditions such as the Staudinger (ketene and imine)4 and Kinugasa reactions (nitrones and alkynes)5 and (b) cyclization with CN or acylN bond formations.6 Recent notable progress was achieved in the transition-metal-catalyzed CH activation. The class of the reactions is intrinsically more fascinating from the point of the ecological and economical view, which could even enable unusual disconnections,7,8 and thus provides one of the most powerful strategies for accessing various β-lactam scaffolds.9 In the utilization of the CH functionalizations catalyzed by the transition-metal complexes, 9e,10 no directed CH activations of simple olefins have been extensively exploited.11 In contrast, a few research studies on the directed © 2018 American Chemical Society
functionalization of the olefinic CH bond are investigated. Recently, Yu and co-workers12 successfully developed the directed CH activation of N-methoxy cinnamamide (Esubstrate E1)13 in 1,4-dioxane with a Pd(II) catalyst, in which air is employed as the sole oxidant. The operationally simple reaction exhibited an excellent regioselectivity and stereochemistry. As indicated in Scheme 1, the cycloisomerization reaction at 100 °C exclusively evolves into the α-CH activation products (four-membered 4-imino-β-lactams), EP and its CC rotation isomer ZP′, while no β-CH activation product EP″(five-membered imino-γ-lactam) is detected. More intriguingly, among these observed α-CH activation products, ZP′ is confirmed to be the major product in 72% yield and only small amount of EP is obtained. With regard to the unusual reaction displayed in Scheme 1, Yu et al. postulated that, as indicated in Scheme 2, the active catalyst, Pd(II)(PhCHCHCONOMe)2, is initially generated in the presence of O2 and E1. A subsequent 1,1-insertion of tBuNC into the PdN bond followed by acyl migration gives intermediate B, which undergoes α-olefinic CH activation to afford the five-membered cyclopalladated species C, and then, reductive elimination occurs to produce EP. Further experReceived: November 28, 2017 Published: January 27, 2018 2067
DOI: 10.1021/acs.joc.7b03007 J. Org. Chem. 2018, 83, 2067−2076
Article
The Journal of Organic Chemistry
Scheme 1. Pd(II)-Catalyzed Cycloismerization of (E)-N-Methoxy Cinnamamide (E1) Reported by Yu and Co-workers12
other unknown pathway? (2) The origin of the product E/Z selectivity is another key issue. It has been experimentally demonstrated by Yu’s group that product EP is the minor one, and its rotation isomer ZP′ acts as the major one. So, what is the intrinsic reason for the preference of ZP′ over EP? (3) The origin of regioselectivity with E1 is another key issue as well. Chen’s group14 has reported the intrinsic reason for the preference of the α-CH activation using Z1, in contrast, when E 1 is used as the substrate, whether the regioselectivity is inherently similar to that with Z1 involved. If not, what should be the actual origin of regioselectivity? These unsolved vital issues above, undoubtedly, would heavily reduce the potential application profile of the intriguing methodology and limit further experimental optimization. Herein, we performed a thorough density functional theory (DFT) study for the Pd(II)catalyzed cycloisomerization of (E)-N-methoxy cinnamamide E 1. It is anticipated that answers to these issues would deepen the understanding for the significant reaction, which is informative for an efficient design of new related catalytic reactions.
Scheme 2. Proposed Mechanistic Pathway by Yu and CoWorkers for the Formation of α-CH Activation Products E P and ZP′, Respectively, from the Pd(II)-Catalyzed Cycloismerization of N-Methoxy Cinnamamide (E1)
imentation12 verified that ZP′ is generated via the CC bond rotation of EP. For the significant reaction, Chen’s group14 has theoretically reported the mechanism leading to the minor product EP with the involvement of E1 and the origin of regioselectivity with Zsubstrate Z1 (using the B3LYP//M06 method). However, to our knowledge, the following three key issues are still unsolved: (1) The detailed mechanism for forming the major product ZP′ is one key issue. It is examined by our calculations that the direct CC bond rotation process requires an activation barrier of 48.1 kcal/mol (Scheme 3), which is unachievable at
2. COMPUTATIONAL DETAILS The geometries of all species were fully optimized at the Becke3LYP level of the density functional theory,20 which has been shown to describe static Pd-catalyzed organometallic systems reasonably well.15,21 The SDD22 and 6-31G(d,p) basis sets23 were used to described Pd and all other atoms, respectively. Vibration frequencies at the same level of theory were conducted to confirm all stationary points as either minima (zero imaginary frequencies) or first-order saddle points (one imaginary frequency) and to provide free energies at 298.15 K. The intrinsic reaction coordinates (IRC)24 for transition states have also been carried out to verify that such structures actually connected two desired minima. Solvation energies were evaluated by a self-consistent reaction field (SCRF) method25 using the polarizable continuum model (CPCM)26 with UAHF atomic radii. The single-point energies in the 1,4-dioxane solvent were obtained at the M06-L level27 by using a larger basis set of SDD for Pd and 6-311++G(d,p) for other atoms based on the gas-phase optimized geometries.28 All of the calculations were implemented with the Gaussian 09 software package.29 It should be emphasized that the gas-phase calculations overestimated the entropic contribution to Gibbs free energy for the real association/dissociation transformations in solution.30 To obtain more accurate relative Gibbs free energies in solution, great effort has been made by some research groups.31 According to the proposal of Pratt and co-workers,32 an additional 4.3 (or −4.3) kcal/mol free energy correction can be applied to each two-to-one (or one-to-two) transformations at 298.15 K and 1 atm. On the basis of the experimental and computational results, Holm and Rybak-Akimova have shown a 4−5 kcal/mol overestimated entropic contribution in their reactions.33 In this paper, we considered applying the method
Scheme 3. Calculated Activation Barrier for the Formation of ZP′ from Direct Rotation of the CC Bond in EP Based on the Proposal of the Yu Group12
the given temperature. Thus, a more favorable pathway is expected to exist, which is a crucial dominated stage of the whole reaction because ZP′ is the preferred major product. In transition-metal-catalyzed systems, the rotation of the CC double bond has been theoretically documented by our group15 and other groups, including Wu,16 Lin,17 Schoenebeck,18 Xia,19 etc. Two typical patterns are involved: either the concerted CC bond rotation assisted by the transition metal3,14−16 or the CC bond transformation into CC σ-bond and then σbond rotation,17 in both of which the oxidative state of the metal center keeps unchanged. In this reaction under investigation, what is the feasible mechanism for the ZP′ formation? Is it one of the two typical patterns mentioned or 2068
DOI: 10.1021/acs.joc.7b03007 J. Org. Chem. 2018, 83, 2067−2076
Article
The Journal of Organic Chemistry
Figure 1. Calculated free energy profile in dioxane for the oxidation of PdL2, deprotonation of substrate E1″, t-BuNC migratory insertion followed by 1,3-acyl migration steps, and structural isomerization. The relative free energies are given in kcal/mol. L = t-BuNC.
developed by Pratt and co-workers to correct the free energies in solution. Such a correction has been confirmed to be reasonable and efficient by massive studies.21c,f
Table 1. Comparison for the Calculated NBO Charges in the C−H Activation Precursors 8 and 11 8 α-C2 α-H O1 Pd
3. RESULTS AND DISCUSSION 3.1. Mechanisms of Forming EP and EP″. Experimentally, Pd2(dba)3/t-BuNC was used as a catalyst precursor. A recent Scheme 4. Tautomerizations of Substrate Amide E1
11 −0.34 0.32 −0.71 0.22
β-C1 β-H O1 Pd
−0.17 0.30 −0.69 0.22
DFT study34 has demonstrated that a Pd(0) monomer, PdL3 (L = t-BuNC) was preferably obtained thermodynamically, and thus PdL3 was considered to initiate the reaction under investigation and also as the reference point in the whole reaction mechanism. Calculations indicated in Figure 1 that one t-BuNC ligand is first dissociated from PdL3 to provide PdL2 as the active catalyst
Figure 2. Calculated free energy profiles in dioxane for forming the products EP and EP″ from intermediate 8. The relative free energies are given in kcal/mol. L = t-BuNC. 2069
DOI: 10.1021/acs.joc.7b03007 J. Org. Chem. 2018, 83, 2067−2076
Article
The Journal of Organic Chemistry
Figure 3. Optimized geometries for species 8, TS8−9, 11, and TS11−12 in Figure 2. The hydrogen atoms, except for those attached to C1 and C2 atoms, have been omitted for clarity. Bond distances are given in angstroms (Å).
Figure 4. Calculated free energy profiles in dioxane for forming the product ZP′ from intermediate 9 via the possible CC σ-bond rotation established in the present work. The relative free energies are given in kcal/mol. L = t-BuNC.
Information), resulting in a four-coordinate square-planar species 3, cis-L2Pd(II)X2 (X = PhCHCHCONOMe). The simultaneously resultant H2O2 is considered as byproduct and further decomposes into H2O and O2 as indicated in previous experimental report.37 From 3, one t-BuNC ligand inserts into the PdN bond to form intermediate 4, in which the N(OMe) trans to L ligand coordinates with the Pd center. The calculations show that the step via TS3−4 requires an activation barrier of 16.7 kcal/mol and is endergonic by 13.7 kcal/mol.
of the reaction. Subsequently, the incorporation of O2 generates peroxopalladium(II) intermediate 2, which has been documented computationally35 and experimentally,36 and then, two molecules substrates will successively react with Pd(II)-peroxo complex 2. As shown in Scheme 4, two isomers of the initial Esubstrate E1, denoted as E1′ and E1″, are located. (See Figure S1 in the Supporting Information.) Although E1″ is less stable than E1 and E1′ in energy, twice E1″-involved deprotonations are most favorable kinetically (Figure S2 in the Supporting 2070
DOI: 10.1021/acs.joc.7b03007 J. Org. Chem. 2018, 83, 2067−2076
Article
The Journal of Organic Chemistry
Figure 5. Calculated free energy profiles in dioxane for forming the product ZP′ from intermediate 9 assisted by the transition metal. The relative free energies are given in kcal/mol. L = t-BuNC.
isomerization takes place and furnishes isomer 8, in which C1C2 π-coordinates with the metal center. Starting from 8, the reaction bifurcates into EP and EP″ and the corresponding calculated free energy profiles are collected in Figure 2. For the EP formation (black line), a concerted metalation deprotonation (CMD) followed by the migration of the α-H (attached to C2 atom) to O1 atom, produces the fivemembered cyclopalladium species 9. The energy demand for the CMD step is computed to be 19.0 kcal/mol. With the replacement of one L ligand by E1″, reductive elimination smoothly occurs with a barrier of 13.3 kcal/mol to achieve product EP and the regenerated PdL2, as the active catalyst, participates in the next catalytic cycle. Alternatively, the rotation of the C2C(O) σ-bond in 8 can give energy comparable to isomer 11, in which the β-H (attached to C1 atom) is adjacent to the acyl O atom. Therefore, the β-CH activation via a CMD process evolves into the six-membered cyclopalladium complex 12. The β-CH activation is found to surmount an energy barrier of 22.6 kcal/mol and to be endergonic by 6.7 kcal/mol relative to 8. The ligand exchange between L and E1″ with a subsequent reductive elimination leads to five-membered cyclization product EP″ after the release of the active catalyst PdL2. 3.2. Origin of Selective CH Activation. It can be seen in Figure 2 that the highest energy point along the β-CH activation pathway (red line), TS11−12, is 3.6 kcal/mol higher than that of TS8−9 along the α-CH activation pathway (black line). The energy difference is obviously smaller than that (more than 25 kcal/mol) for the Z1-involved system by Chen’s theoretical report,14 which originated from both electronic effect of the common precursor intermediate and different coordination modes of Pd with the C(α)C(β) bond in the transition states involved. In this subsection, to explore the origin of the energy difference of 3.6 kcal/mol, we first performed natural bond orbital (NBO) analyses. It is observed in Table 1 that the charge of the α-C2 atom in 8 is −0.34 e, while it is −0.17 e for the β-C atom in 11. Clearly, the attack of the α-C2 to the Pd center in 8 is more favored kinetically due to the stronger electrostatic attraction between the α-C2 and
Figure 6. Optimized geometries for the species 10, TS10−16, 16, and TS17−18 in Figure 5. The hydrogen atoms, except for those attached to C1 atom, have been omitted for clarity. Bond distances are given in angstroms (Å).
The large endergonicity of this process may be a result of configurational distortion of the planar quadrilateral structure in 4. The initially dissociated L ligand now recoordinates to the metal and gives the stable square-planar intermediate 5, a structure featuring two L ligands trans to each other. Through TS5−6, the acyl C atom migrates to the adjacent N(t-Bu) atom (i.e., 1,3-acyl migration) with a barrier of 17.2 kcal/mol, resulting in C-amidinyl Pd(II) complex 6. This 1,3-acyl migration is calculated to exergonic by 10.9 kcal/mol relative to 5. Upon the release of one L ligand in 6, the carbonyl O immediately coordinates to the metal center and gives the unstable Pd···O coordinated intermediate 7. The instability of 7 over 6 may be a result of a relatively weaker O···Pd interaction than the t-BuNC···Pd interaction. A subsequent coordinated 2071
DOI: 10.1021/acs.joc.7b03007 J. Org. Chem. 2018, 83, 2067−2076
Article
The Journal of Organic Chemistry
Figure 7. Calculated free energy profiles in dioxane for forming the product ZP′ from product EP assisted by the PdN coordination established in the present work. The relative free energies are given in kcal/mol. L = t-BuNC.
Figure 8. Optimized geometries for selected species shown in Figure 7. The hydrogen atoms, except for those attached to C1 atoms, have been omitted for clarity. Bond distances are given in angstroms (Å).
3.3. Mechanisms of Forming ZP′. As mentioned in the Introduction, the direct rotation of the CC double bond in E P is theoretically unfeasible for the ZP′ formation. Therefore, further calculations are required to uncover how the intrinsically inert double bond is rotated. Initially, we designed an alternative mechanism (denoted as path a), which features the CC bond rotation turning into the CC σ-bond rotation. The corresponding calculated free energy profile is exhibited in Figure 4. The alternative pathway starts from the α-CH activation product 9 shown in Figure 2. It performs the C1C2 bond insertion into the hydroxyl O1H bond via TS9−14 to afford the bridged-ring intermediate 14, in which the C1C2 σ-bond is achieved. Through rotating the C1C2 σ-bond, the H atom initially ligated to the C1 atom transfers to the O1 atom and provides the bridged-ring complex 16 after the ligand exchange between E1″ and L. The
Pd atoms. This fact is also evidenced by the geometric parameters shown in Figure 3; in TS8−9, the Pd···C2 distance, 2.20 Å, is slightly shorter than the Pd···C1 length in TS11−12, 2.24 Å. However, the instability of 11 over 8 by 1.3 kcal/mol in free energy, as a contributor of the β-CH activation barrier from 8, is partially responsible for the relatively difficult β-C H activation. Besides, in contrast, from the Z1-involved system, it is seen in Figure 3 that the metal center is exclusively coordinating with the activated C atom in TS8−9 and TS11−12, thus bringing similar geometric configurations. Overall, the preference for α-CH activation with E1 involved can be mainly attributed to the electron effect of the precursor intermediates. On the basis of the above facts, it is predicted that the different coordination modes of the metal center in the CH activation transition states importantly influence the energy difference between two CH activations. 2072
DOI: 10.1021/acs.joc.7b03007 J. Org. Chem. 2018, 83, 2067−2076
Article
The Journal of Organic Chemistry
calculated for the CC rotation step is 30.4 kcal/mol, corresponding to the difference between TS17−18 and 17. The finding is in sharp contrast to a recent Ni-catalyzed DFT study by Lin’s group,38a in which the double-bond rotation is very facile in the presence of the unsaturated Ni center. After the recoordination of one L ligand from 18 to provide 16, the reductive elimination smoothly results in product ZP′ and the PdL2 regeneration. For the 10 → 16 transformation, different from path b with the CC bond rotation prior to one L dissociation, the direct double-bond rotation without L ligand dissociation (path c) is also presented in Figure 5. Although the transition state TS10−16 is 9.6 kcal/mol lower than TS17−18 in path b, the direct rotation pathway is still inaccessible under the given temperature due to the high energy demand of 36.2 kcal/mol. The high instability of 16 over 10 is believed to be a key factor leading to the difficult CC rotation, which can be rationalized in terms of the greater steric repulsion between the carbonyl group and the phenyl substituent in 16. As indicated in Figure 6, the (carbonyl)O···H(Ph) distance is 2.02 Å in 16, much shorter than (carbonyl)O···H(C1) in 10 (2.40 Å). The energy difference involved in the two structures can be significantly reduced by replacing the phenyl group with a less bulky Me group. In the less bulky system, the double-bond rotation product is calculated to be only 2.9 kcal/mol higher in free energy than the rotation precursor. Next, we turn our attention to path d in Figure 7, in which the rotation of the C1C2 bond assisted by the transition metal starts from adduct 19, generated by PdL2 π-coordination with EP, and the corresponding geometries with selected structural parameters for the species are given in Figure 8. Upon the delivery of one L ligand from 19, the C1C2 rotation is completed and affords the Z-isomer 21 after the recoordination of the dissociated L ligand. The CC rotation transition state is denoted as TS20−21 and has an incredibly high barrier of 49.6 kcal/mol relative to 19. Finally, the active catalyst PdL2 is released from 21, and the product ZP′ is formed. The high barriers involved in the pathways, paths a−d, are incompatible with the experimental finding that the major product ZP′ is obtained in heated environments. The results imply that a reasonable strategy should be pursued. Here, we proposed a new mechanism for 19 → 21, which involves a stepwise Pd(0) → Pd(II) → Pd(0) process and is referred to as
Figure 9. Orbital diagrams and optimized structures for two ratedetermining CC bond rotation transition states, TS20−21 and TS19−22. The hydrogen atoms, except for those attached to C1 atoms, have been omitted for clarity. The relative free energies are given in kcal/mol. L = t-BuNC.
H-shift process via TS15−16 is calculated to proceed with an activation barrier of 31.0 kcal/mol relative to 14. A subsequent reductive elimination with 11.3 kcal/mol barrier smoothly evolves into product ZP′, and simultaneously, the active catalyst PdL2 is regenerated, thereby completing the whole catalytic cycle. It is worth noting that the resting state of the whole reaction is structure 6 in Figure 1, which is 27.1 kcal/mol below the starting materials. In this case, the overall barrier for forming ZP′ is 47.8 kcal/mol (the difference between TS15−16 and 6). It is clear that the pathway through the CC bond → CC σ bond → σ bond rotation is not viable under heating conditions. Note that a five-membered cyclopalladium is presented in structure 10 (Figure 2). Moreover, 10 is more stable than 9 by 8.2 kcal/mol in free energy. To this end, beginning with 10, an alternative mechanism for the CC double-bond rotation assisted by the transition metal15,38 is designed. As illustrated in Figure 5, the denoted path b is initiated by the dissociation of one L ligand to afford the 14-electron three-coordinated complex 17, which is endergonic by 15.4 kcal/mol. Subsequently, the CC bond undertakes a rotation via TS17−18 to generate the Z-isomer 18 with the phenyl and H substituents switched in positions. The activation barrier
Figure 10. Calculated free energy profiles in dioxane for EP (blue line) and ZP′ (red line) formation, respectively. The relative free energies are given in kcal/mol. 2073
DOI: 10.1021/acs.joc.7b03007 J. Org. Chem. 2018, 83, 2067−2076
Article
The Journal of Organic Chemistry
Scheme 5. Sketch of the Catalytic Cycles for Forming EP and ZP′ from the Pd(II)-Catalyzed Cycloisomerization of N-Methoxy Cinnamamide (E1) Based on the Present Calculations
blue line), the reaction would diverge into two different catalytic cycles: either PdL2 catalyzes the cycloisomerization of E 1″ to provide EP (right part, blue line) or EP evolves into the isomer ZP′ (right part, red line). In the right part, the potential energy profile leading to EP (blue line) lies much lower than that leading to ZP′ (red line). The fact clearly suggests that the E P formation is fast. In other words, during the initial stage of the reaction, product EP is certainly generated, while no ZP′ is obtained. Once E1″ is used up, the catalytic cycles forming EP is terminated. Thereafter, EP would combine with the active catalyst PdL2, resulting in the isomer ZP′. It should be emphasized here that the reverse evolvement ZP′ → EP needs an energy barrier of 34.1 kcal/mol (the difference between TS19−22 and 21 in Figure 7), which is roughly sustainable under 100 °C. Therefore, it is not difficult to understand the small amount of the formation of EP. According to the calculated results above, we schematically present the catalytic cycles for the formation of EP and ZP′, respectively, in Scheme 5, which provides a detailed and consistent view of the mechanism details of the significant catalytic isomerization of N-methoxy cinnamamide E1.
the oxidation/reduction-promoted double-bond rotation mechanism. The new pathway is denoted as path e (black line).39 As illustrated in Figure 7, in the oxidation-promoted step (from 19 to 20), the C1C2 bond in 19 performs a 90° rotation with simultaneous [4+1] oxidative cyclization (EP and Pd(0) active catalyst) to afford the five-membered palladacycle intermediate 22, in which the palladacycle is coplanar to the β-lactam fourmembered ring. This step via transition state TS 19−22 overcomes a barrier of 30.1 kcal/mol and is endergonic by 27.3 kcal/mol; then, a further C1C2 rotation with concomitant reductive decyclization through TS22−21 smoothly results in the Z-isomer 21. The product ZP′ is finally formed by repulsing PdL2 from 21. From the potential energy surface profile shown in Figure 7, we can see that the highest stationary point (TS 19−22 corresponding to 0.4 kcal/mol) along path e is remarkably lower that along path d (TS20−21 corresponding to 19.9 kcal/ mol). The origin of the large energy difference can be understood by comparatively analyzing different geometries of the two 16e transition states involved. It is shown in Figure 9 that the Pd center in TS20−21 π-interacts with the C(Ph)-C1 moiety. Consequently, the d orbitals of Pd can not effectively overlap with the π orbital of the C(Ph)-C1 moiety, and thus a distorted triangular structure is exhibited. In comparison, the coordination mode of the Pd center in TS19−22 features a regular square-planar structure and the Pd d orbitals can effectively overlap with the orbitals of four ligands, which, thereby, causes a low energy barrier. In addition, the kinetic advantage of the stepwise over concerted step is also of key importance for the favorable path e. In summary, path e is more favorable and the overall barrier along the pathway (the difference between TS19−22 and 19) is calculated to be 30.1 kcal/mol, in roughly accordance with the occurrence of the reaction in heated conditions. 3.4. Origin of the Product E/Z Selectivity. On the basis of our theoretical investigations, ZP′ is believed to be formed in a catalytic fashion in which EP serves as the substrate. Meanwhile, a competitive pathway, i.e., the subsequent catalytic cycle leading to EP, should also be considered. As indicated in Figure 10, after EP is formed in the first catalytic cycle (left part,
4. CONCLUSIONS The mechanisms of Pd(II)-catalyzed cycloisomerization of Nmethoxy cinnamamide (E1) have been systematically investigated with the aid of DFT calculations. For the EP formation, the following steps are proposed: the oxidation of the active catalyst PdL2 with O2 (PdL2 → 2), deprotonation (2 → 3), tBuCN migratory insertion (3 → 5), 1,3-acyl migration (5 → 6), α-CH activation (6 → 10), reductive elimination, and regeneration of PdL2 (10 → EP). The regioselectivity of the αCH activation over β-CH activation is attributed to a greater charge density on the α-C atom, leading to a stronger electrostatic attraction between the α-C and Pd atoms. For the Z P′ formation, a novel oxidation/reduction-promoted catalytic mechanism (Pd(0) → Pd(II) → Pd(0)) is proposed, which is initiated by the PdL2 coordination to the CC bond in EP, and then, the stepwise oxidative cyclization/reductive decyclization-assisted CC bond rotation processes is followed to 2074
DOI: 10.1021/acs.joc.7b03007 J. Org. Chem. 2018, 83, 2067−2076
Article
The Journal of Organic Chemistry produce ZP′ and to regenerate PdL2. The overall barrier for the alternative pathway is calculated to be 30.1 kcal/mol. Further theoretical investigations indicate that, at the initial stage of the reaction, EP is certainly completely obtained by the presence of the substrate E1. Once E1 has run out, EP would act as the partner of the new catalytic cycle and slowly evolve into Z P′. Moreover, a small amount of generated ZP′ would reversibly transform to EP due to the slightly higher barrier involved.
■
(6) (a) Bonache, M. A.; Gerona-Navarro, G.; Martín-Martínez, M.; García-López, M. T.; López, P.; Cativiela, C.; González-Muñiz, R. Synlett 2003, 7, 1007−1011. (b) Kanwar, S.; Sharma, S. D. Bull. Chem. Soc. Jpn. 2006, 79, 1748−1752. (7) (a) Anada, M.; Watanabe, N. Chem. Commun. 1998, 15, 1517− 1518. (b) Doyle, M. P.; Kalinin, A. V. Synlett 1995, 10, 1075−1076. (8) (a) Wasa, M.; Yu, J.-Q. J. Am. Chem. Soc. 2008, 130, 14058− 14059. (b) He, G.; Zhao, Y.; Zhang, S.; Lu, C.; Chen, G. J. Am. Chem. Soc. 2012, 134, 3−6. (c) Zhang, Q.; Chen, K.; Rao, W.; Zhang, Y.; Chen, F.; Shi, B. Angew. Chem., Int. Ed. 2013, 52, 13588−13592. (d) Sun, W.-W.; Cao, P.; Mei, R.-Q.; Li, Y.; Ma, Y.-L.; Wu, B. Org. Lett. 2014, 16, 480−483. (e) Wu, X.; Zhao, Y.; Zhang, G.; Ge, H. Angew. Chem., Int. Ed. 2014, 53, 3706−3710. (f) Wang, Z.; Ni, J.; Kuninobu, Y.; Kanai, M. Angew. Chem., Int. Ed. 2014, 53, 3496−3499. (9) (a) Davies, I. W.; Welch, C. J. Science 2009, 325, 701−704. (b) SchÖ nherr, H.; Cernak, T. Angew. Chem., Int. Ed. 2013, 52, 12256−12267. (c) Bryan, M. C.; Dillon, B.; Hamann, L. G.; Hughes, G. J.; Kopach, M. E.; Peterson, E. A.; Pourashraf, M.; Raheem, I.; Richardson, P.; Richter, D.; Sneddon, H. F. J. Med. Chem. 2013, 56, 6007−6021. (d) Lian, Y.; Huber, T.; Hesp, K. D.; Bergman, R. G.; Ellman, J. A. Angew. Chem., Int. Ed. 2013, 52, 629−633. (e) Zhu, R. Y.; He, J.; Wang, X. C.; Yu, J. Q. J. Am. Chem. Soc. 2014, 136, 13194− 13197. (f) Shang, R.; Ilies, L.; Asako, S.; Nakamura, E. J. Am. Chem. Soc. 2014, 136, 14349−14352. (10) Lian, Y.; Huber, T.; Hesp, K. D.; Bergman, R. G.; Ellman, J. A. Angew. Chem., Int. Ed. 2013, 52, 629−633. (11) (a) Chen, Y.; Wang, F.; Jia, A.; Li, X. Chem. Sci. 2012, 3, 3231− 3226. (b) Chen, F.; Feng, Z.; He, C.-Y.; Wang, H.-Y; Guo, Y.-L; Zhang, X. Org. Lett. 2012, 14, 1176−1179. (c) Kim, Y. W.; Niphakis, M. J.; Georg, G. I. J. Org. Chem. 2012, 77, 9496−9503. (d) Carrë r, A.; Brion, J.-D.; Messaoudi, S.; Alami, M. Org. Lett. 2013, 15, 5606−5609. (e) Yu, Y.-Y.; Bi, L.; Georg, G. I. J. Org. Chem. 2013, 78, 6163−6169. (f) Guchhait, S. K.; Priyadarshani, G. J. Org. Chem. 2015, 80, 8482− 8488. (12) Kong, W.-J.; Liu, Y.-J.; Xu, H.; Chen, Y.-Q.; Dai, H.-X.; Yu, J.-Q. J. Am. Chem. Soc. 2016, 138, 2146−2149. (13) In the following section, the capital letter “E” or “Z” is marked in the left superscript of the substrates and products in order to distinguish E or Z configurations. (14) Fan, W.; Sun, C.; Huang, F.; Liu, J.; Zhao, X.; Sheng, X.; Chen, D. Dalton Trans. 2017, 46, 5288−5296. (15) Wu, W.; Bi, S.; Liu, Y.; Tang, Y.; Li, B. Org. Biomol. Chem. 2015, 13, 11539−11549. (16) Song, L.-J.; Wang, T.; Zhang, X.; Chung, L. W.; Wu, Y.-D. ACS Catal. 2017, 7, 1361−1368. (17) Zhang, J.; Quan, Y.; Lin, Z.; Xie, Z. Organometallics 2014, 33, 3556−3563. (18) Sperger, T.; Le, C. M.; Lautens, M.; Schoenebeck, F. Chem. Sci. 2017, 8, 2914−2922. (19) Guo, W.; Xia, Y. J. Org. Chem. 2015, 80, 8113−8121. (20) (a) Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B: Condens. Matter Mater. Phys. 1988, 37, 785−789. (b) Miehlich, B.; Savin, A.; Stoll, H.; Preuss, H. Chem. Phys. Lett. 1989, 157, 200−206. (c) Becke, A. D. J. Chem. Phys. 1993, 98, 5648−5652. (d) Stephens, P. J.; Devlin, F. J.; Frisch, M. J. J. Phys. Chem. 1994, 98, 11623−11627. (21) (a) Yu, H. Z.; Fu, Y.; Guo, Q. X.; Lin, Z. Y. Organometallics 2009, 28, 4507−4512. (b) Wang, H. L.; Yang, X.; Liu, Y. X.; Bi, S. W. Organometallics 2014, 33, 1404−1415. (c) Li, B. W.; Bi, S. W.; Liu, Y. X.; Ling, B. P.; Li, P. Organometallics 2014, 33, 3453−3463. (d) Bi, S. W.; Wang, H. L.; Liu, Y. X.; Yang, X. Curr. Org. Chem. 2014, 18, 1661−1471. (e) Dang, Y.; Qu, S. L.; Wang, Z. X.; Wang, X. T. J. Am. Chem. Soc. 2014, 136, 986−998. (f) Xie, H. J.; Zhao, L. J.; Yang, L.; Lei, Q. F.; Fang; Xiong, C. H. J. Org. Chem. 2014, 79, 4517−4527. (g) Yang, D. S.; Liu, Y. X.; Sun, P. F.; Zhang, N.; Wei, W.; Sun, M. Y.; Chen, G.; Bi, S. W.; Wang, H. Org. Biomol. Chem. 2016, 14, 895−904. (h) Zhang, X.; Liu, Y.; Chen, G.; Pei, G.; Bi, S. Organometallics 2017, 36, 3739−3749.
ASSOCIATED CONTENT
S Supporting Information *
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.joc.7b03007. Calculated free energy barriers in dioxane using different functionals for selected key steps, calculated free energy profiles for the tautomerization of amide E1 to imidic acid E1″, for the oxidation of PdL2 and deprotonation of different N-methoxy amides, and for the transformation from 9 to 16 assisted by an H2O molecule, and Cartesian coordinates of all of the species involved (PDF)
■
AUTHOR INFORMATION
Corresponding Authors
*E-mail:
[email protected]. Phone: +86-537-4458308. Fax: +86-537-4456305. *E-mail:
[email protected]. ORCID
Yuxia Liu: 0000-0003-1139-8563 Siwei Bi: 0000-0003-3969-7012 Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS This work was jointly supported by the National Natural Science Foundation of China (nos. 21403123, 21473100, and 21603116), the China Postdoctoral Science Foundation (2016M600531), the Opening Foundation of Shandong Provincial Key Laboratory of Detection Technology for Tumor Markers (KLDTTM2015-9), and the Doctoral StartUp Scientific Research Foundation of Qufu Normal University (Grant no. BSQD2012018).
■
REFERENCES
(1) (a) Morin, R. B.; Gorman, M. Chemistry and Biology of β-Lactam Antibiotics; Academic Press: New York, 1982; Vols. 1−3. (b) Nehaus, F. C.; Georgopadaku, N. H. In Emerging Targets in Antibacterial and Antifungal Chemotherapy; Sutcliffe, J., Georgopapadakou, N. H., Eds.; Springer Science: Berlin, Germany, 1992. (c) Pitts, C. R.; Lectka, T. Chem. Rev. 2014, 114, 7930−7953. (2) (a) Manhas, M. S.; Bose, A. K. â-lactams: Natural and Synthetic, Part 1; Wiley: New York, 1971. (b) Page, M. I. In Adv. Phys. Org. Chem.; Bethell, D., Ed.; Academic Press: London, UK, 1987. (3) (a) Alcaide, B.; Almendros, P. Synlett 2002, 3, 381−393. (b) Singh, G. S. Tetrahedron 2003, 59, 7631−7649. (c) Fisher, J. F.; Meroueh, S. O.; Mobashery, S. Chem. Rev. 2005, 105, 395−424. (d) Alcaide, B.; Almendros, P.; Aragoncillo, C. Chem. Rev. 2007, 107, 4437−4492. (4) (a) D’hooghe, M.; Van Brabandt, W.; Dekeukeleire, S.; Dejaegher, Y.; De Kimpe, N. Chem. - Eur. J. 2008, 14, 6336−6340. (b) Tuba, R. Org. Biomol. Chem. 2013, 11, 5976−5988. (5) Marco-Contelles, J. Angew. Chem., Int. Ed. 2004, 43, 2198−2200. 2075
DOI: 10.1021/acs.joc.7b03007 J. Org. Chem. 2018, 83, 2067−2076
Article
The Journal of Organic Chemistry (22) (a) Dolg, M.; Wedig, U.; Stoll, H.; Preuss, H. J. Chem. Phys. 1987, 86, 866−872. (b) Andrae, D.; Hä ussermann, U.; Dolg, M.; Stoll, H.; Preuss, H. Theor. Chim. Acta. 1990, 77, 123−141. (23) Andzelm, J.; Huzinaga, S. Gaussian Basis Sets for Molecular Calculations; Elsevier Science: New York, 1984. (24) Fukui, K. Acc. Chem. Res. 1981, 14, 363−368. (25) Tomasi, J.; Persico, M. Chem. Rev. 1994, 94, 2027−2094. (26) (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. (27) Zhao, Y.; Truhlar, D. G. Acc. Chem. Res. 2008, 41, 157−167. (28) Calculated free energy barriers in dioxane using different functionals for selected key steps are given in the Supporting Information. It can be seen that all of the functionals give comparable barriers and consistent trends for selected processes. These facts indicate that the B3LYP//M06-L method employed in our calculations is suitable for the present study. (29) 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, O.; Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian 09, revision D.01; Gaussian, Inc.: Wallingford, CT, 2009. (30) (a) Braga, A. A. C.; Ujaque, G.; Maseras, F. Organometallics 2006, 25, 3647−3658. (b) Tuttle, T.; Wang, D. Q.; Thiel, W.; et al. Organometallics 2006, 25, 4504−4513. (31) (a) Mammen, M.; Shakhnovich, E. I.; Deutch, J. M.; Whitesides, G. M. J. Org. Chem. 1998, 63, 3821−3830. (b) Tamura, H.; Yamazaki, H.; Sato, H.; Sakaki, S. J. Am. Chem. Soc. 2003, 125, 16114−16126. (c) Liang, Y.; Liu, S.; Xia, Y.; Li, Y.; Yu, Z.-X. Chem. - Eur. J. 2008, 14, 4361−4373. (d) García-Melchor, M.; Pacheco, M. C.; Nájera, C.; Lledós, A.; Ujaque, G. ACS Catal. 2012, 2, 135−144. (e) Poater, A.; Pump, E.; Vummaleti, S. V. C.; Cavallo, L. J. Chem. Theory Comput. 2014, 10, 4442−4448. (32) Martin, R. L.; Hay, P. J.; Pratt, L. R. J. Phys. Chem. A 1998, 102, 3565−3573. (33) (a) Huang, D.; Makhlynets, O. V.; Tan, L. L.; Lee, S. C.; RybakAkimova, E. V.; Holm, R. H. Proc. Natl. Acad. Sci. U. S. A. 2011, 108, 1222−1227. (b) Huang, D.; Makhlynets, O. V.; Tan, L. L.; Lee, S. C.; Rybak- Akimova, E. V.; Holm, R. H. Inorg. Chem. 2011, 50, 10070− 10081. (34) Dang, Y.; Deng, X.; Guo, J.; Song, C.; Hu, W.; Wang, Z.-X. J. Am. Chem. Soc. 2016, 138, 2712−2723. (35) (a) Popp, B. V.; Morales, C. M.; Landis, C. R.; Stahl, S. S. Inorg. Chem. 2010, 49, 8200−8207. (b) Konnick, M. M.; Decharin, N.; Popp, B. V.; Stahl, S. S. Chem. Sci. 2011, 2, 326−330. (c) Decharin, N.; Popp, B. V.; Stahl, S. S. J. Am. Chem. Soc. 2011, 133, 13268−13271. (d) Keith, J. M.; Goddard, W. A., III. Organometallics 2012, 31, 545− 552. (36) (a) Gligorich, K. M.; Sigman, M. S. Angew. Chem., Int. Ed. 2006, 45, 6612−6615. (b) Cornell, C. N.; Sigman, M. S. Inorg. Chem. 2007, 46, 1903−1909. (c) Gligorich, K. M.; Sigman, M. S. Chem. Commun. 2009, 3854−3867. (d) Wendlandt, A. E.; Suess, A. M.; Stahl, S. S. Angew. Chem., Int. Ed. 2011, 50, 11062−11087. (e) Campbell, A. N.; Stahl, S. S. Acc. Chem. Res. 2012, 45, 851−863. (f) Wu, W.; Jiang, H. Acc. Chem. Res. 2012, 45, 1736−1748. (37) (a) Konnick, M. M.; Stahl, S. S. J. Am. Chem. Soc. 2008, 130, 5753−5762. (b) Ingram, A. J.; Walker, K. L.; Zare, R. N.; Waymouth, R. M. J. Am. Chem. Soc. 2015, 137, 13632−13646.
(38) (a) Zhang, J.; Quan, Q.; Lin, Z.; Xie, Z. Organometallics 2014, 33, 3556−3563. (b) Song, L. J.; Wang, T.; Zhang, X.; Chung, L. W.; Wu, Y. D. ACS Catal. 2017, 7, 1361−1368. (c) Sperger, T.; Le, C. M.; Lautens, M.; Schoenebeck, F. Chem. Sci. 2017, 8, 2914−2922. (39) It is noted that the stepwise steps 19 → 21 involve the Pd coordination with the N(OMe) atom. The corresponding stepwise steps with weaker PdO(carbonyl) coordination are also theoretically performed and collected in the Supporting Information, which are found to be energetically higher than those with PdN(OMe) coordination.
2076
DOI: 10.1021/acs.joc.7b03007 J. Org. Chem. 2018, 83, 2067−2076
