Article Cite This: J. Org. Chem. 2018, 83, 10959−10973
pubs.acs.org/joc
A Molecular Electron Density Theory Study of the Role of the Copper Metalation of Azomethine Ylides in [3 + 2] Cycloaddition Reactions Luis R. Domingo,*,† Mar Ríos-Gutiérrez,† and Patricia Pérez‡ †
Department of Organic Chemistry, University of Valencia, Dr. Moliner 50, 46100 Burjassot, Valencia, Spain Universidad Andres Bello, Facultad de Ciencias Exactas, Departamento de Ciencias Químicas, Av. República 498, 8370146, Santiago, Chile
Downloaded via SAN FRANCISCO STATE UNIV on September 21, 2018 at 11:06:29 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.
‡
S Supporting Information *
ABSTRACT: The copper metalation of azomethine ylides (AYs) in [3 + 2] cycloaddition (32CA) reactions with electrondeficient ethylenes has been studied within the Molecular Electron Density Theory (MEDT) at the MPWB1K/6-311G(d,p) level, in order to shed light on the electronic effect of the metalation in the course of the reaction. Analysis of the Conceptual Density Functional Theory reactivity indices indicates that the metalation of AYs markedly enhances the nucleophilicity of these species given the anionic character of the AY framework. These 32CA reactions take place through stepwise mechanisms characterized by the formation of a molecular complex. Both nonmetalated and metalated 32CA reactions present similar activation energies. While metalated 32CA reactions are completely regioselective, their stereoselectivity depends on the bulk of the ligand as well as the nature of the ethylene derivative. The metalation of the AY slightly increases the asynchronicity of the C−C single bond formation. Electron Localization Function topological analysis of the C−C bond formation processes makes it possible to characterize the mechanism of these 32CA reactions as a two-stage one-step mechanism. The present MEDT study rules out any catalytic role of the Cu(I) cation in the kinetics of the 32CA reactions of metalated AYs.
1. INTRODUCTION
Scheme 1. Synthesis of Pyrrolidines by 32CA Reactions of AYs
The [3 + 2] cycloaddition (32CA) reaction between a threeatom component (TAC) and an ethylene derivative is one of the most efficient synthetic methods for the obtainment of fivemembered heterocyclic compounds in a highly regio- and stereoselective manner.1,2 Pyrrolidines are important fivemembered heterocyclic units which have only one nitrogen in their core framework, of great pharmaceutical importance.3−5 The simplest pyrrolidine 3 can be obtained by a 32CA reaction of azomethine ylide (AY) 1 with ethylene 2 (see Scheme 1). The simplest AY 1 is a very reactive species participating in pseudodiradical-type (pdr-type) 32CA reactions.6,7 The pseudodiradical electronic structure of AY 1 (see Scheme 1) is accountable for the very low activation energy of the nonpolar 32CA reaction with ethylene 2, 1.0 kcal·mol−1.7 Although substitution in the simplest AY 1 increases the activation energies, the corresponding 32CA reactions remain very fast. © 2018 American Chemical Society
Consequently, due to the high reactivity of AYs, the use of Lewis acid catalysts is not required in their 32CA reactions. Most TACs involved in 32CA reactions are generated in situ as transient species. In this context, there are at least three methods for the generation of AYs (see Scheme 2): the first one Received: June 26, 2018 Published: July 27, 2018 10959
DOI: 10.1021/acs.joc.8b01605 J. Org. Chem. 2018, 83, 10959−10973
Article
The Journal of Organic Chemistry
silver vs copper, these 32CA reactions involving substituted AYs derived from α-aminoester 4 are completely meta regioselective (see Scheme 3). In 2000, Cossio studied, for the first time, theoretically as well as experimentally the 32CA reaction of the lithium-metalated AY 10 with nitroethylene 11, finding a stepwise mechanism involving the formation of a strong stabilized zwitterionic intermediate (see Scheme 4).29 In acetonitrile, the B3LYP/631+G(d) activation energy associated to the first nucleophilic attack was 7.0 kcal·mol−1, while that for the ring closure was 10.7 kcal·mol−1. A high endo stereoselectivity was observed as a result of the strong electrostatic interaction present between the lithium cation and an oxygen of the nitro group, which is only feasible along the endo reaction path.29 Very recently, the reactivity and selectivities in the 32CA reaction of the lithium-metalated AY 14 with MA 5 were theoretically studied by Emamian at the MPWB1K/6-31G(d) computational level (see Scheme 5).30 This 32CA reaction showed to be entirely regioselective and endo stereoselective affording pyrrolidine 15, in complete agreement with the experimental outcomes31 and Cossio’s theoretical findings.29 In THF, the reaction turned stepwise to allow a weak stabilization of the corresponding zwitterionic intermediate, but the barrier for the subsequent ring closure was unappreciable, 1.3 kcal mol−1. Lewis acids in 32CA reactions of nucleophilic TACs with weak electrophilic ethylenes, or in 32CA reactions of weak electrophilic TACs with strong nucleophilic ethylenes, considerably accelerate 32CA reactions, which has been explained in terms of a strong electrophilic activation of the species to which the Lewis acid is coordinated, thus favoring the 32CA reaction through a more polar process. Although the diastereo- and enantioselectivities in 32CA reactions of N-metalated AYs have been theoretically studied to a great extent,18,20,22,32 the role of the metal in the course of these 32CA reactions has been considered to a much lesser extent. Thus, in order to shed light on the electronic effect of the metalation of AYs in 32CA reactions, the reactions of AY 16, broadly described in the literature, and the copper-metalated AY-Cu 17 and AY-CuL 18 with MA 5 are studied herein within the Molecular Electron Density Theory (MEDT)33 (see Scheme 6). While, in the absence of bulk bidentated ligands, the formation of an early strong Cu(I)-carboxyl electronic interaction determines the reactivity and endo stereoselectivity, the use of a hindered bidentated ligand such as bisoxazoline 19, which sterically prevents this electronic interaction, can provide relevant information about the reactivity of metalated AYs involving chiral bidentated ligands in 32CA reactions.
Scheme 2. In Situ Generation of AYs
consists of the thermal ring aperture of a properly substituted aziridine;8 the second method involves the isomerization of an imine derivative of an α-aminoester containing at least one αhydrogen;9−11 and in the third method, the treatment of an αaminoester with a basic species in the presence of an appropriate salt generates an N-metalated AY.12,13 Interestingly, the generation of N-metalated AYs in the presence of chiral ligands has allowed the enantioselective synthesis of pyrrolidines.5,14−28 In 1991, Grigg and co-workers24 demonstrated, for the first time, that the addition of a stoichiometric amount of chiral cobalt or manganese complexes with ephedrine derivatives as the chiral ligand in the generation of AYs derived from imines of glycine alkyl esters could give pyrrolidines with up to 96% ee. It was also reported that silver(I) salts in combination with chiral phosphane ligands can catalyze the 32CA reaction of AYs. Later, Jorgensen described the 32CA reactions of N-benzylidene glycinates with methyl acrylate (MA) 5 and Et3N as the base in the presence of chiral ligands such as bisoxazoline 7 (R = tBut) and Lewis acids such as Zn(OTf)2, yielding pyrrolidine 6 in high yield and as a diastereomerically pure product with 78% ee and an improvement in the enantioselectivity to 91% ee at −20 °C (see Scheme 3).25 Scheme 3. 32CA Reactions of N-Metalated AYs in the Presence of Chiral Bidentated Ligands with Electrophilic Ethylenes
2. COMPUTATIONAL METHODS DFT calculations were performed using the MPWB1K functional34 together with the 6-311G(d,p) basis set.35 Optimizations were performed using the Berny method.36,37 The transition state structures (TSs) were characterized by frequency computations. The intrinsic reaction coordinate (IRC) paths38 were carried out using the secondorder González−Schlegel integration method39,40 Solvent effects of dichloromethane (DCM) were considered by single-point energy calculations using the polarizable continuum model (PCM) developed by Tomasi’s group41,42 in the framework of the self-consistent reaction field (SCRF).43−45 The electronic structures were characterized by natural population analysis (NPA),46,47 and by the Electron Localization Function (ELF)48 and Quantum Theory of Atoms in Molecules (QTAIM)49 topological analyses of the electron density. All computations were
Different metals, such as Zn,17,25 Ag,14,17,18,20,22 and Cu, 5,17−19,21,23,25−28 and chiral ligands, such as 7, 25 8,5,14,18,19,23,26,27 or 9,17,22,23 have been used in the 32CA reactions of N-metalated AYs with electrophilic ethylene derivatives, improving diastereo- and enantioselectivities. Interestingly, while stereoselectivities have been found to be dependent on the chiral ligands, the substitution of the electrophilic ethylene and even the nature of the metal, e.g. 10960
DOI: 10.1021/acs.joc.8b01605 J. Org. Chem. 2018, 83, 10959−10973
Article
The Journal of Organic Chemistry Scheme 4. 32CA Reaction of the Lithium-Metalated AY 10 with Nitroethylene 11
Scheme 5. 32CA Reaction of the Lithium-Metalated AY 14 with MA 5
Scheme 6. 32CA Reactions of Substituted AY 16 and the Copper-Metalated AY-Cu 17 and AY-CuL 18 with MA 5
as the reference, was first performed. Figure 1 shows the localization domains of the ELF, the positions of the ELF basin attractors, the populations of the most significant valence basins, and the proposed ELF-based Lewis structures as well as the natural atomic charges. The topological ELF analysis of the simplest AY 1 reflects the molecular symmetry of this compound. Thus, two pairs of V(C) and V′(C) monosynaptic basins, integrating a total population of 1.04e each one, and two V(C,N) disynaptic basins, integrating 2.58e each one, are found. Each pair of V(C) and V′(C) monosynaptic basins can be related to a pseudoradical carbon center, and each V(C,N) disynaptic basin, to a C−N partial double bond (see 1 in Figure 1). Consequently, AY 1 possesses a pseudodiradical electronic structure7 that enables this TAC to participate in pdr-type 32CA reactions.6 The presence of a carboxyl group at the C1 carbon and a phenyl group at the C3 carbon in AY 16 provokes the depopulation of the four V(C1), V′(C1), V(C3), and V′(C3) monosynaptic basins by a total of 0.16e (C1) and 0.72e (C3), and the increase of the population of the V(C1,N2) and V(N2,C3) disynaptic basins by a total of 0.31e. The larger population of the V(N2,C3) disynaptic basin than that of the V(C1,N2) one (see 16 in Figure 1), together with the stronger depopulation of the C3 pseudoradical center than of the C1 one, suggests an electron delocalization mainly toward the phenyl substituent bound to the C3 carbon. However, AY 16 also presents a pseudodiradical structure7 and, therefore, it is expected
carried out with the Gaussian 09 suite of programs.50 ELF and QTAIM studies were performed with the TopMod51 and Multiwfn52 programs, respectively, using the corresponding optimized MPWB1K/6-31G(d) monodeterminantal wave functions. Global electron density transfer (GEDT)53 values were computed by the sum of the natural atomic charges (q) of the atoms belonging to each framework at the TSs; GEDT = Σqf. Conceptual DFT54,55 (CDFT) global reactivity indices and Parr functions56 were computed using the equations given in ref 55.
3. RESULTS AND DISCUSSION The present MEDT study is organized as follows: in section 3.1, we study the electronic structures of AY 16 and coppermetalated AY-Cu 17 through a topological analysis of the electron density. In section 3.2, we present an analysis of the CDFT reactivity indices at the ground state (GS) of the reagents involved in the 32CA reactions under study. In section 3.3, we perform a study of the 32CA reactions of copper-metalated AYCu 17 and AY-CuL 18 with MA 5. In section 3.4, we analyze the origin of the endo stereoselectivity in the 32CA reactions of copper-metalated AY-Cu 17 and AY-CuL 18. And finally, in section 3.5, we perform an ELF topological analysis of the C−C single bond formation along the meta/endo reaction path of the 32CA reaction between copper-metalated AY-Cu 17 and MA 5. The study of the 32CA reaction of AY 14 with MA 5 is given in section 1 of the Supporting Information. 3.1. Topological Analysis at the GS of AY 16 and Copper-Metalated AY-Cu 17. A topological analysis of the ELF48 of AY 16 and copper-metalated AY-Cu 17, characterizing their electronic structure as well as the simplest counterpart AY 1 10961
DOI: 10.1021/acs.joc.8b01605 J. Org. Chem. 2018, 83, 10959−10973
Article
The Journal of Organic Chemistry
Figure 1. MPWB1K/6-31G(d) domains of the ELF localization of AYs 1, 16 and AY-Cu 17, at an isosurface value of ELF = 0.75 (upper part); positions of the ELF basin attractor, populations of the most representative valence basin (center part); and the proposed ELF-based Lewis structures, with the natural atomic charges (lower part). Negative charges are colored in red, and negligible charges, in green. Populations of the ELF valence basin and natural atomic charges are given in number of electrons, e.
that, similar to the simplest AY 1, it will participate in pdr-type 32CA reactions.6 When the Cu(I) cation is coordinated to the N2 nitrogen of AY 16, the most relevant ELF topological change observed at copper-metalated AY-Cu 17 is the disappearance of the V(C1[3]) and V′(C1[3]) monosynaptic basins associated with the C1 and C3 pseudoradical centers present at AYs 1 and 16 (see 17 in Figure 1). Note that the green localization domain observed at the C1 carbon is the consequence of the delocalization of the V(C1,C) disynaptic basin situated between the C1 carbon and the carboxyl group toward the former. Furthermore, while the total electron population of the V(C1,N2) and V(N2,C3) disynaptic basins is again that obtained at AY 1, i.e. 5.15e, a new V(N2) monosynaptic basin, related to N2 nonbonding electron density, can be observed, having a population of 2.67e. Note that the TOPMOD program51 assigns the N2 nonbonding electron density as a V(N2,Cu) disynaptic basin, which would correspond to an N2−Cu single bond. Consequently, in order to correctly assign the synapticity of this valence basin and, thus, to characterize the electronic interaction between the N2 nitrogen and Cu(I) cation in the copper-metalated AY-Cu 17, a QTAIM49 topological analysis of the electron density distribution in the N2−Cu(I) region was performed. The calculated QTAIM parameters of the critical points (cps) found in the Cu(I)−N2 region, as well as in the C1−N2 one for comparison, are gathered in Table 1, while the color-filled maps of the Laplacian of the electron density and the ELF in the molecular plane defined by the Cu(I), N2 and C1′ nuclei are represented in Figure 2.
In the QTAIM framework, negative values of the Laplacian of the electron density, i.e. ∇2ρcp < 0, are associated with covalent bonds, while positive values, i.e. ∇2ρcp > 0, are related to noncovalent interactions. As shown in Table 1, the electron density associated with cp1 presents a low value, ρcp ≤ 0.1 au, and Laplacian ∇2ρcp > 0, indicating that the trajectories of the gradient paths involving cp1 are not associated with a covalent bond. Note that, in contrast, cp2 associated with the C1−N2 covalent bond of the AY framework presents ρcp2 ≥ 0.3 au and Laplacian ∇2ρcp2 < 0. Finally, the two-dimensional color-filled map of the ELF given in Figure 2b clearly shows the monosynaptic behavior of the V(N2) valence basin. Consequently, the noncovalent interaction between the Cu(I) cation and the N2 nitrogen in AY-Cu 17 can be associated with a Cu(I)−N2 ionic bond. Table 1. QTAIM Parameters (in au), Namely, The Electron Density ρcp and Its Laplacian ∇2ρcp, of the cps Associated with the Cu(I)−N2 and C1−N2 Regions in AY-Cu 17 Cu(I)−N2 C1−N2
cp
ρcp
∇2ρcp
cp1 cp2
0.106 0.314
0.422 −0.392
Finally, an NPA was carried out in order to establish the charge distribution at AYs 1, 16, and 17. The NPA indicated that the C1−N2−C3 core framework of the simplest AY 1 presents a total negative charge of −1.26e almost equally shared among the three core nuclei, the hydrogen nuclei gathering the positive charge (see Figure 1). This picture significantly contrasts with that arising from the resonance theory, emphasizing that charges 10962
DOI: 10.1021/acs.joc.8b01605 J. Org. Chem. 2018, 83, 10959−10973
Article
The Journal of Organic Chemistry
Figure 2. Two-dimensional representations of the color-filled maps of (a) the Laplacian of the electron density and (b) the ELF, in the Cu(I)−N2− C1′ molecular plane of AY-Cu 17. (3,-1) cps with ∇2ρcp < 0 are colored in blue, while cps with ∇2ρcp > 0 are colored in red.
AY 1 will never act as an electrophile in a polar process, with N > 5 eV it will act as a very strong nucleophile. The presence of a carboxyl group at the C1 carbon and a phenyl substituent at the C3 carbon of the simplest AY 1 considerably enhances the ω index of AY 16, ω = 1.25 eV, and diminishes the N index, N = 4.66 eV; although the substitution considerably increases the electrophilicity of AY 16, it is still classified as a strong nucleophile. Metalation of AY 16 by the Cu(I) cation diminishes the electrophilicity ω index of AY-Cu 17 to 0.70 eV, classifying it as a marginal electrophile, and enhances its nucleophilicity N index to 5.44 eV, categorizing it as a supernucleophile. More drastic changes are observed when bidentated ligand 19 is coordinated to the Cu(I) cation in AY-CuL 18; now, the nucleophilicity N index of complex AY-CuL 18 rises to 5.90 eV, a very high value. The strong nucleophilic character of these metalated AYs can be attributed to the anionic character of the AY framework in these species. Note that the usual catalytic role of Lewis acids produces a rise of the electrophilic character of organic molecules; hence, more polar processes are favored. On the other hand, MA 5 shows ω = 1.28 eV and N = 1.33 eV allowing its classification as a strong electrophile and as a moderate nucleophile. It is worth mentioning that although MA 5 is classified as a strong electrophile within the electrophilicity scale, it is one of the weakest electrophilic species.55 Analysis of the global CDFT reactivity indices at the GS of the reagents indicates that, in a polar 32CA reaction, MA 5 will act as an electrophile and AYs 16−18 will act as strong nucleophiles. In general, it is observed that the polar character of 32CA reactions increases by increasing the electrophilic character of the ethylene; however, in these 32CA reactions, the supernucleophilic behavior of the AYs controls the polar character of these reactions. The most favorable reaction path given between nonsymmetric electrophilic/nucleophilic pairs along polar processes is that associated with the initial two-center interaction between the most electrophilic center of the electrophile and the most nucleophilic center of the nucleophile. In 2013, the electrophilic Pk+ and nucleophilic Pk+ Parr functions56 were proposed. These indices were derived from the spin electron density changes
are not the result of distributing electrons by pairs in Lewis structures.57 The carboxyl and phenyl groups bound to C1 and C3, respectively, at AY 16 decrease the corresponding negative charges to −0.25e (C1) and −0.01e (C3), suggesting that the electron-withdrawing power of the latter is approximately twice that of the former. Note that the N2 nitrogen remains negatively charged by 0.42e. Finally, coordination of the Cu(I) cation to the N2 nitrogen produces no significant changes in the charge distribution of AY-Cu 17. The negative charge of the N2 nitrogen increases by 0.14e, but those of the C1 and C3 carbons increase by negligible amounts. 3.2. Analysis of the CDFT Reactivity Indices at the GS of the Reagents. It is well-known that reactivity indices defined within the CDFT54,55 are very useful tools to predict the polar and ionic reactivity of chemical processes. Global descriptors, specifically, the chemical potential, μ, hardness, η, electrophilicity, ω, and nucleophilicity, N, of MA 5 and AYs 16− 18 are given in Table 2. Table 2. MPWB1K/6-311G(d,p) Electronic Chemical Potential (μ), Chemical Hardness (η), Global Electrophilicity (ω), and Global Nucleophilicity (N), in eV MA 5 AY 16 AY-Cu 17 Ethylene 2 AY-CuL 18 Simplest AY 1
μ
η
ω
N
−4.72 −3.41 −2.58 −3.67 −2.06 −2.03
8.69 4.65 4.75 9.90 4.88 6.19
1.28 1.25 0.70 0.68 0.43 0.33
1.33 4.66 5.44 1.78 5.90 5.27
It may be seen that the electronic chemical potentials58,59 μ of the AYs, between −2.06 (AY-CuL 18) and −3.41 (AY 16) eV, are higher than that of MA 5, μ = −4.72 eV, pointing out that, along a polar 32CA reaction, the GEDT53 will flux from these AYs toward the electrophilic MA 5. The ω60 and N61 indices of the simplest AY 1 are 0.33 and 5.27 eV, respectively, allowing it to be classified as a marginal electrophile within the electrophilicity scale and as a strong nucleophile within the nucleophilicity scale.55 Therefore, while 10963
DOI: 10.1021/acs.joc.8b01605 J. Org. Chem. 2018, 83, 10959−10973
Article
The Journal of Organic Chemistry
Figure 3. 3D representations of the Mulliken atomic spin densities of the radical anion of MA 5 and radical cations of AYs 16−18. Electrophilic Pk+ and nucleophilic Pk− Parr functions of MA 5 and AYs 16−18 are also given.
(see Scheme 4)29 and Emamian (see Scheme 4),30 is a consequence of the favorable electronic interaction taking place between the Cu(I) cation and the carboxyl oxygen of MA 5 in the s-cis conformation, which is not feasible on the other reaction paths. However, when MA 5 approaches in the s-trans conformation, the cycloaddition process takes place in one-step (see Scheme 7). The MPWB1K/6-311G(d,p) relative electronic energies with respect to MC-Cu, in DCM, of the stationary points involved in the 32CA reaction of coppermetalated AY-Cu 17 with MA 5 are given in Scheme 7, while the total electronic energies, in gas phase and in DCM, are given in the Supporting Information. An exploration of the reaction paths between the separated reagents and the TSs allowed finding MC-Cu, in which the two reagents are close in a parallel rearrangement. In DCM, MC-Cu is 8.6 kcal·mol−1 more stable than the separated reagents due to the favorable electronic interactions taking place between the basic carboxyl oxygen of MA 5 and the acidic Cu(I) cation. From MC-Cu, the relative energies in DCM of the stationary points found along the most favorable meta/endo reaction path are 4.1 (TS1-Cu-mn), −5.0 (IN-Cu-mn), −4.5 (TS2-Cu-mn), and −16.0 (CA-Cu-mn) kcal·mol−1. Along the other four reaction paths, the relative energies of the TSs with respect to MC-Cu are 8.6 (TS′-Cu-mn), 10.8 (TS-Cu-mx), 9.2 (TS-Cuon), and 13.6 (TS-Cu-ox), the reaction paths being exothermic by 14−16 kcal·mol−1. Some significant conclusions can be inferred from these energy results: (i) The 32CA reaction of copper-metalated AY-Cu 17 with MA 5 presents a low activation energy, 4.1 (TS1-Cu-mn) kcal·mol−1. In fact, TS1-Cu-mn is found below the separated reagents. (ii) Formation of IN-Cumn is exothermic by 5.0 kcal·mol−1, but the subsequent ring closure has an unappreciable barrier, 0.5 kcal·mol−1; i.e., once IN-Cu-mn is formed, it closes very quickly yielding CA-Cu-mn. Consequently, the finding of this two-step cycloaddition mechanism does not have any relevance. (iii) TS′-Cu-mn, in which MA 5 presents the s-trans conformation, is found 4.5 kcal· mol−1 above TS1-Cu-mn as a consequence of the disappearance of the favorable electronic interaction appearing in the endo TS1-Cu-mn; (iv) interestingly, the activation energy associated
attained through the GEDT process from the nucleophile to the electrophile, as a powerful tool to study the local reactivity in polar and ionic processes. Consequently, both the electrophilic Pk+ and the nucleophilic Pk− Parr functions of MA 5 and AYs 16−18, respectively, were calculated and analyzed to identify the most electrophilic and nucleophilic centers of the species involved in these metalated 32CA reactions (see Figure 3). After the Pk+ Parr functions analysis at the reactive sites of MA 5, it can be concluded that the β-conjugated C4 carbon is the most electrophilic center of this species with a Pk+ value of 0.61. Otherwise the nucleophilic Pk− Parr functions analysis at the reactive sites of AY 16 reveals that the carboxyl substituted C1 carbon, Pk− = 0.69, is significantly more nucleophilically activated than the phenyl substituted C3 carbon, Pk− = 0.26. Metalation of AY 16 does not substantially modify the nucleophilic Pk− Parr functions of AY-Cu 17 and AY-CuL 18 (see Figure 3). In both metalated AYs, the C1 carbon is approximately twice as nucleophilically activated as the C3 one. 3.3. Study of the 32CA Reaction of Copper-Metalated AY-Cu 17 and AY-CuL 18 with MA 5. The study of the nonmetalated 32CA reaction of AY 16 with MA 5 is given in the Supporting Information. For the 32CA reaction of coppermetalated AY 16 with MA 5, two computational models were chosen: (i) in Model I, the Cu(I) cation was explicitly solvated with two ether molecules; (ii) in Model II, the Cu(I) cation was coordinated to the bidentated ligand 19 (see Scheme 6). 3.3.1. Study of the 32CA Reaction of Copper-Metalated AYCu 17 with MA 5. First, the 32CA reaction of copper-metalated AY-Cu 17, Model I, is analyzed. Similar to the 32CA reaction between AY 16 and MA 5 (see Scheme S1 in the Supporting Information), this 32CA reaction can take place through two pairs of reaction paths: one pair of stereoisomeric paths, namely endo and exo, and another pair of regioisomeric paths, namely meta and ortho (see Scheme 7). Analysis of the stationary points found along the four reaction paths reveals that this 32CA reaction takes place via a stepwise mechanism. An exhaustive analysis of the meta/endo reaction path made it possible to find two TSs and one intermediate associated with the cycloaddition process. This behavior, which is similar to that found by Cossio 10964
DOI: 10.1021/acs.joc.8b01605 J. Org. Chem. 2018, 83, 10959−10973
Article
The Journal of Organic Chemistry Scheme 7. Competitive Reaction Paths Associated with the Reaction of Copper-Metalated AY-Cu 17 with MA 5a
MPWB1K/6-311G(d,p) relative electronic energies with respect to MC-Cu, in DCM, are given in kcal·mol−1.
a
with this copper-metalated 32CA reaction via TS1-Cu-mn is only slightly superior to that associated with the reaction of AY 16, 0.8 kcal·mol−1 (see TS-mn in the Supporting Information). Consequently, the copper metalation of AY 16 in Model I does not play any catalytic role in the reaction rate. Note that this energy difference even increases to 5.3 kcal·mol−1 along TS′-Cumn; (v) this reaction is completely endo stereoselective as TSCu-mx is 6.7 kcal·mol−1 above TS1-Cu-mn and completely meta regioselective as TS-Cu-on is 5.1 kcal·mol−1 above TS1Cu-mn. Consequently, although the copper metalation of AY 16 has no significant role in the reaction rate, it has a determining role in the regio- and stereoselectivity of this 32CA reaction. Finally, (vi) the copper metalation of AY 16 reduces the exothermic character of the reaction considerably. The geometries of the TSs are given in Figure 4. Along the most favorable meta/endo reaction path, the distances between the pairs of C1/C4 and C3/C5 interacting carbons are 2.121 and 2.778 Å at TS1-Cu-mn, 1.602 and 2.520 Å at IN-Cu-mn, and 1.596 and 2.311 Å at TS2-Cu-mn, while at the meta/endo TS′-Cu-mn and the meta/exo TS-Cu-mx these values are 2.031 and 2.753, and 2.042 and 2.779 Å, respectively. At the ortho TSs, the distances between the pairs of C1/C5 and C3/C4 interacting carbons are 2.690 and 2.093 Å at TS-Cu-on and 2.559 and 2.026 Å at TS-Cu-ox, respectively. Several interesting conclusions can be achieved from these geometrical parameters: (i) these TSs are more asynchronous and slightly later than those associated with the 32CA reaction between AY 16 and MA 5 (see Figure S1 in the Supporting Information); (ii) the geometry of TS2-Cu-mn is very similar to that of IN-Cu-mn, explaining the very flat energy surface around these stationary points; (iii) as expected, the comparable C−C distances at TS′Cu-mn and TS-Cu-mx suggest a very similar bonding pattern at
Figure 4. MPWB1K/6-311G(d,p) gas phase geometries of the TSs associated with the 32CA reaction of AY-Cu 17 with MA 5. The distances at IN-Cu-mn are given in parentheses, while those at TS′-Cumn are give in brackets. Distances are given in angstroms, Å. 10965
DOI: 10.1021/acs.joc.8b01605 J. Org. Chem. 2018, 83, 10959−10973
Article
The Journal of Organic Chemistry
Scheme 8. Competitive Reaction Paths Associated with the 32CA Reaction of Copper-Metalated AY-CuL 18 with MA 5a
MPWB1K/6-311G(d,p) relative electronic energies with respect to MC-CuL, in DCM, are given in kcal·mol−1.
a
both stereoisomeric TSs; and finally, (iv) in the four TSs, the C− C bond formation process involving the β-conjugated C4 carbon of MA 5 is more advanced than that at the α position, in agreement with the similar electrophilic Parr functions of AY 16 and AY-Cu 17 (see section 3.2). The polar nature of the 32CA reaction of AY 14 with MA 5 was evaluated by computing the GEDT at the corresponding TSs. GEDT values of 0.0e are associated with nonpolar processes, while values higher than 0.2e are associated with polar processes. The values of the GEDT, which fluxes from the AY-Cu to the MA frameworks, are 0.35e at TS1-Cu-mn, 0.61e IN-Cu-mn, 0.56e at TS2-Cu-mn, 0.38e at TS′-Cu-mn, 0.35e at TS-Cu-mx, 0.35e at TS-Cu-on, and 0.28e at TS-Cu-ox. These high values caused by the strong nucleophilic character of AYCu 17 allow establishing the strong polar character of this 32CA reaction. 3.3.2. Study of the 32CA Reaction of Copper-Metalated AYCuL 18 with MA 5. In the reaction of copper-metalated AY CuL 18 with MA 5, Model II, the Cu(I) cation was coordinated to the bidentated ligand 19. Similar to the two aforementioned reactions, this 32CA reaction can take place through four reaction paths (see Scheme 8). This 32CA reaction takes place via a stepwise mechanism. The MPWB1K/6-311G(d,p) relative electronic energies with respect to MC-CuL, in DCM, of the stationary points involved in the 32CA reaction of coppermetalated AY-CuL 18 with MA 5 are given in Scheme 8, while the total electronic energies, in gas phase and in DCM, are given in the Supporting Information. The reaction begins with the formation of MC-CuL, in which the two reagents are close in a parallel rearrangement. In DCM, MC-CuL is 3.2 kcal·mol−1 more stable than the separated
reagents. Note that MC-CuL is 5.4 less stable than MC-Cu as a consequence of the disappearance of the favorable electronic interactions present in the latter (see section 3.3.1). From MCCuL, the relative energies of the TSs associated with the four reaction paths in DCM are 2.6 (TS-CuL-mn), 4.6 (TS-CuLmx), 8.8 (TS-CuL-on), and 7.2 (TS-CuL-ox) kcal·mol−1, the reaction being exothermic by 17−18 kcal·mol−1. Some appealing conclusions can be drawn from these energy results: (i) The 32CA reaction of copper-metalated AY-CuL 18 with MA 5 presents very low activation energy, 2.6 (TS-CuL-mn) kcal·mol−1. This activation energy is only 0.7 kcal·mol−1 lower than that associated with the metal-free 32CA reaction. Consequently, similar to Model I, the copper metalation of AY 16 in Model II plays no noticeable catalytic role in the reaction rate. Note that considering the relative energies with respect to the separated reagents, TS-mn is located 1.1 kcal/mol below TS-CuL-mn. (ii) This reaction is moderately endo stereoselective, as TS-CuL-mx is 2.0 kcal·mol−1 above TS-CuLmn, and completely meta regioselective, as TS-CuL-ox is 4.6 kcal·mol−1 above TS-CuL-mn. The use of a hindered ligand such as bidentated ligand 19, which prevents the approach of the carboxyl oxygen of MA 5 to the Cu(I) cation, slightly decreases the endo stereoselectivity as a consequence of the disappearance of the favorable electronic interactions present in TS1-Cu-mn, but maintains complete meta regioselectivity. (iii) Although the metalation reduces the exothermic character of the 32CA reaction, at low temperature it is thermodynamically irreversible. These energy results are in complete agreement with the experimental observation than while the stereoselectivities have been found to be dependent on the chiral ligands, the substitution of the electrophilic ethylene, and even the nature 10966
DOI: 10.1021/acs.joc.8b01605 J. Org. Chem. 2018, 83, 10959−10973
Article
The Journal of Organic Chemistry
Figure 5. MPWB1K/6-311G(d,p) gas phase geometries of the TSs associated with the 32CA reaction of metalated AY-CuL 18 with MA 5. Distances are given in angstroms, Å.
of the metal, these 32CA reactions are completely meta regioselective. The geometries of the TSs associated with the 32CA reaction of AY-CuL 18 with MA 5 are given in Figure 5. At the meta TSs, the distances between the pairs of C1/C4 and C3/C5 interacting carbons are 2.050 and 2.758 Å at TS-CuL-mn and 2.023 and 2.686 Å at TS-CuL-mx, while, at the ortho TSs, the distances between the pairs of C1/C5 and C3/C4 interacting carbons are 2.568 and 2.061 Å at TS-CuL-on and 2.622 and 2.016 Å at TS-CuL-ox, respectively. A comparison of the geometries of the TSs involved in reaction Models I and II shows great similarity. Consequently, the change of coordination in these models, i.e. two explicit molecules of dimethyl ether as solvent or a bidentated ligand such as 19, does not significantly modify the electronic structure of the TSs. The main difference between the two reaction models is the presence of an O(carboxyl)−Cu(I)electronic interaction in Model I, which is not feasible in reaction Model II due to the presence of the bulky bidentated ligand which prevents the approach of the carboxyl oxygen of MA 5 to the Cu(I) cation. The values of the GEDT at the TSs, fluxing from the coppermetalated AY-CuL to the MA frameworks, are 0.34e at TS-CuLmn, 0.35e at TS-CuL-mx, 0.30e at TS-CuL-on, and 0.31e at TSCuL-ox. At the more favorable meta TSs, the GEDT is slightly higher than that at the ortho ones. These high values, which are similar to those found at the TSs of the 32CA reaction between AY-Cu 17 and MA 5, account for the strong polar character of
this 32CA reaction as a consequence of the strong nucleophilic character of copper-metalated AY-Cu 17 and AY-CuL 18. In general, the GEDT decreases the activation energies of polar organic reactions favoring the bonding changes demanded to reach the corresponding TS along the reaction path.53 Due to the higher nucleophilic character of copper-metalated AYs 17 and 18 than that of AY 16, the GEDT values of the TSs are significantly higher than that computed for the metal-free 32CA reaction of AY 16. However, despite the higher polar character of these copper-metalated 32CA reactions, the corresponding activation energies remain unchanged. This behavior is a consequence of the fact that, in these polar 32CA reactions, the GEDT is not controlled by the weak electrophilic MA 5 but by the strongly nucleophilic AYs (see section 3.2). 3.4. Origin of the Endo Stereoselectivity in the 32CA Reactions of Copper-Metalated AY-Cu 17 and AY-CuL 18 with MA 5. Endo/exo stereoselectivity may be the result of weak noncovalent interactions, namely, electrostatic interactions, hydrogen bonds, van der Waals interactions, etc. While favorable electrostatic interactions play an important role in the endo stereoselectivity, repulsive steric interactions could be responsible for the exo stereoselectivity. As was mentioned in the introduction, while in the absence of bulk bidentated ligands, the formation of an early strong Cu(I)-carboxyl electronic interaction determines the reactivity and endo stereoselectivity,29,30 the use of a hindered bidentated ligand such as bisoxazoline 19 can change the endo stereoselectivity. In order to characterize the factors controlling the endo stereoselectivity in 10967
DOI: 10.1021/acs.joc.8b01605 J. Org. Chem. 2018, 83, 10959−10973
Article
The Journal of Organic Chemistry the copper-metalated 32CA reactions of AY-Cu 17 and AY-CuL 18 with MA 5, the topology of the noncovalent interactions (NCI)62 taking place at the more favorable TSs, TS1-Cu-mn and TS-CuL-mn, was analyzed (see Figure 6).
Figure 7. Phases in which the IRC associated with the pmr-type 32CA reaction between copper-metalated AY-Cu 17 and MA 5 is topologically divided at Sj structures. The red line indicates the position of TS′-Cu-mn, the blue line indicates the division of the IRC in two stages, and the turquoise lines indicate the structures defining the phases at which the formation of the two new C−C single bonds starts. Relative energies, ΔE, with respect to the first structure of the reaction path, S0, are given in kcal·mol−1.
Figure 6. Noncovalent interaction (NCI) gradient isosurfaces of the TSs involved in the most favorable meta/endo reaction path associated with the 32CA reaction of metalated AY-Cu 17 (TS1-Cu-mn) and AYCuL 18 (TS-CuL-mn) with MA 5.
Cu-mn and TS-mn (see Tables S5 and S6 in the Supporting Information). The populations of the most relevant ELF valence basins, among other pertinent parameters, of these structures are given in Table 3, while their ELF attractor positions are shown in Figure 8. ELF topological analysis of AY-Cu 17 is discussed in section 3.1. On the other hand, the presence of two disynaptic basins, V(C4,C5) and V′(C4,C5), integrating a total population of 3.42e, allows characterizing the C4−C5 double bond of MA 5. At TS′-Cu-mn, d(C1−C4) = 2.057 Å and d(C3−C5) = 2.779 Å, the most relevant topological changes with respect to the separated reagents are the creation of a V(C1) monosynaptic basin, integrating 0.71e, at the most nucleophilic C1 carbon of AY-Cu 17, and the merger of the two V(C4,C5) and V′(C4,C5) disynaptic basins present in MA 5 into a new V(C4,C5) disynaptic basin integrating 3.39e as a consequence of the slight depopulation of the C4−C5 bonding region. The energy cost associated with TS′-Cu-mn, which is mainly related to the formation of the C1 pseudoradical center, is 7.3 kcal·mol−1, while the GEDT is 0.31e. At S4′, d(C1−C4) = 2.009 Å and d(C3−C5) = 2.771 Å, two new V(C4) and V(C5) monosynaptic basins, integrating 0.23e each, are observed at the C4 and C5 carbons of the MA 5 framework (see Figure 8). The electron density of these V(C4[5]) monosynaptic basins, which are associated with two C4 and C5 pseudoradical centers, is a consequence of an electron density reorganization within the C4−C5 bonding region caused by the depopulation of the two V(C4,C5) and V′(C4,C5) disynaptic basins, present in MA 5, by ca. 0.47e. In addition, the V(C1) monosynaptic basin present at TS′-Cu-mn has slightly increased its population to 0.77e. At S4′, which is isoenergetic to TS′-Cu-mn, the GEDT has increased to 0.34e. At S5, d(C1−C4) = 1.995 Å and d(C3−C5) = 2.768 Å, the first most relevant topological change along the meta/endo reaction path takes place: while the two V(C1) and V(C4) monosynaptic basins present at S4′ have disappeared, a new V(C1,C4) disynaptic basin is created with an initial population of 0.96e (see Figure 8). This relevant topological change indicates that the formation of the first C1−C4 single bond takes place at a C−C distance of 2.00 Å through the C-to-C coupling of the two C1 and C4 pseudoradical centers.53 At S5, the GEDT has slightly increased to 0.35e.
The dark blue color of the surface in the region between the Cu(I) cation and the carboxyl oxygen of MA 5 indicates a strong attractive NCI in TS1-Cu-mn, which is not present in TS-CuLmn. Conversely, two green surfaces associated with weak favorable NCIs between two methylene hydrogens of bidentated ligand 19 and the carboxyl oxygen of MA 5 are observed in TS-CuL-mn, which may be responsible for the lower endo stereoselectivity found in reaction Model II. NCI analysis of the most favorable meta/endo TS of the 32CA reaction of (Z)-C-phenyl-N-methylnitrone 20 with dimethyl 2benzylidenecyclopropane-1,1-dicarboxylate 21 revealed that a nonclassical CH/O hydrogen bond involving the nitrone C−H hydrogen is responsible for the selectivity experimentally found in that 32CA reaction. This weak hydrogen bond interaction accounts for the endo stereoselectivity in the 32CA reaction involving AY-CuL 18; however, the use of a more hindered ligand, or ethylene derivatives not having a basic oxygen, can turn the 32CA reaction exo selective. 3.5. ELF Topological Analysis of the C−C Single Bond Formation along the Meta/Endo Reaction Path Associated with the Copper-Metalated 32CA Reaction between AY-Cu 17 and MA 5. In order to understand the role of the copper metalation in the reaction mechanism of the 32CA reactions of AY 16, an ELF topological analysis of the structures associated with the C−C single bond formation along the meta/ endo reaction path of the 32CA reaction between AY-Cu 17 and MA 5, in the s-trans conformation, was performed. In addition, a topological analysis of the ELF of the corresponding structures associated with the meta/endo reaction path of the 32CA reaction between AY 16 and MA 5 was also carried out in order to perform a comparative analysis of the two reaction mechanisms (see section 2 in the Supporting Information). The 13 phases in which the IRC associated with the 32CA reaction between copper-metalated AY-Cu 17 and MA 5 is topologically divided are shown in Figure 7. This figure also shows where the two C−C single bonds are formed along the IRC. The selected structures, i.e. those defining the phases at which the formation of the new C−C single bonds starts, Sj, and the last structure of the previous phases, S′j, were chosen after performing a BET63 study along the IRCs associated with TS′10968
DOI: 10.1021/acs.joc.8b01605 J. Org. Chem. 2018, 83, 10959−10973
Article
The Journal of Organic Chemistry
Table 3. ELF Valence Basin Populations of the Reagents, TS′-Cu-mn and CA-Cu-mn, as well as of the Selected Structures of the IRC Involved in the C−C Single Bond Formation, i.e. S4′, S5, S10′, and S11, along the meta/endo Reaction Path Associated with the 32CA Reaction between AY-Cu 17 and MA 5 (Distances Are Given in Angstroms, Å; MPWB1K/6-31G(d) Gas Phase Relativea Energies in kcal·mol−1; and Electron Populations and GEDT in Average Number of Electrons, e)
structures
AY-Cu 17
TS′-Cu-mn
S4′
S5
d(C1−C4) d(C3−C5)
2.057 2.779
2.009 2.771
1.995 2.768
ΔE
7.3
7.3
7.3
GEDT
0.31
0.34
0.35
0.27
0.26
0.05
2.07 2.85
2.04 2.87
2.03 2.86
1.76 3.79
1.76 3.78
2.85 3.39
2.83 2.92
2.83 2.98
2.05 2.02
2.04 2.01
1.71 2.61 1.58 1.75 1.87
0.71
0.77
1.77 1.29
1.83 1.81
V(C1,N2) V(N2) V′(N2) V(N2,C3) V(C4,C5) V′(C4,C5) V(C1) V(C3) V(C4) V(C5) V′(C5) V(C1,C4) V(C3,C5)
MA 5
2.11 2.67 3.03 1.64 1.78
S10′
S11
1.581 2.037
1.580 2.025
−3.1
−3.4
CA-Cu-mn 1.551 1.591 −17.8
0.45 0.23 0.23
0.24
0.82
0.96
1.75
Relative to the first point of the reaction path, S0.
a
At S10′, d(C1−C4) = 1.581 Å and d(C3−C5) = 2.037 Å, while the V(C5) monosynaptic basin has reached 0.82e, a new V(C3) monosynaptic basin, integrating 0.45e, related to the C3 pseudoradical center present in AY-Cu 17, is created. Note that, at this structure, there is a strong electron reorganization between the five centers involved in the cycloaddition process. The electron density of the V(C3) monosynaptic basin partly comes from the depopulation of the V(N2,C3) disynaptic basin by 0.78e, and concomitantly that depopulation increases the electron density of the V(N2) monosynaptic basin to 3.79e. It is worth mentioning that the populations of the V(C5) monosynaptic basin and V(C1,C4) disynaptic basin increase as the population of V(C4,C5) decreases by 0.96e. Note that the new V(C1,C4) disynaptic basin has already reached a population of 1.75e, being 96% of the electron density of the C1−C4 single bond in CA-Cu-mn. At S10′, the GEDT has decreased to 0.27e. At S11, d(C1−C4) = 1.580 Å and d(C3−C5) = 2.025 Å, the second most significant topological change along the reaction
path occurs: the two V(C3) and V(C5) monosynaptic basins present in S10′ have merged into a new V(C3,C5) disynaptic basin integrating a population of 1.29e (see Figure 8). This important topological change indicates that formation of the second C3−C5 single bond begins at a C3−C5 distance of 2.03 Å through the C-to-C coupling of the two C3 and C5 pseudoradical centers present in S10′. Note that, at S11, the V(C1,C4) disynaptic basin has a population of 1.77e, being 97% of the electron density of the C1−C4 single bond in CA-Cu-mn. This behavior permits characterizing the two-stage one-step mechanism.64 The bonding changes taking place toward the formation of the new C3−C4 single bond are slightly exothermic by 3.4 kcal·mol−1 (see Table 3). Finally, at CA-Cu-mn, d(C1−C4) = 1.551 Å and d(C3−C5) = 1.591 Å, the V(C1,C4), V(C3,C5), and V(C4,C5) disynaptic basins end up with ca. 1.80e, while the V(N2) monosynaptic basin present at S11 has split into two monosynaptic basins, V(N2) and V’(N2), integrating 2.61e and 1.71e, respectively. Interestingly, the population of the V(N2) monosynaptic basin 10969
DOI: 10.1021/acs.joc.8b01605 J. Org. Chem. 2018, 83, 10959−10973
Article
The Journal of Organic Chemistry
Figure 8. Attractor positions of the ELF valence basins for the structures of the IRC involved in the formation of the C−C single bonds along the meta/ endo reaction path associated with the 32CA reaction between AY-Cu 17 and MA 5. The electron populations, in e, are given in brackets.
of CA-Cu-mn is the same as that of the V(N2) monosynaptic basin of AY-Cu 17, displaying the strong ionic character of the Cu(I)−N2 interaction (see section 3.1). The strong depopulation of the V(N2,C3) disynaptic basin by 1.29e, compared with that of the V(C1,N2) one to 1.71e, i.e. 0.47e, at the end of the reaction path emphasizes the stronger contribution of the V(N2,C3) disynaptic basin to the nonbonding N2 electron density assisted by the Cu presence (see Table 3). The bonding changes taking place toward the formation of CA-Cu-mn are exothermic by 17.8 kcal·mol−1, while the GEDT decreases to 0.05e. From this ELF topological analysis of the C−C single bond formation along the meta/endo reaction path associated with the 32CA reaction of copper-metalated AY-Cu 17 and MA 5, the following conclusions can be drawn: (i) The activation energy of reaction Model I via TS′-Cu-mn, 7.3 kcal·mol−1 from S0, can be mainly associated with the creation of the C1 pseudoradical center in AY-Cu 17. (ii) Formation of the two new C1−C4 and C3−C5 single bonds takes place at Phases VI and XII, respectively (see Figure 7), at C−C distances of 2.00 and 2.03 Å via the C-to-C coupling of two C1[3] and C4[5] pseudoradical centers created along the reaction.53 Formation of the first C1− C4 single bond in the copper-metalated 32CA reaction, 2.00 Å, takes place later than that in the 32CA reaction of AY 16, 2.23 Å (see Tables S4 and S5 in the Supporting Information). (iii) Formation of the first C1−C4 single bond involving the most nucleophilic center of AY-Cu 17 and the most electrophilic center of MA 5 is anticipated by the analysis of the Parr functions (see section 3.2). (iv) Formation of the second C3−C5 single bond begins at Phase XII (see Figure 7) when formation of the first C1−C4 single bond is completed by up to 97% and with a quite high 27.0% degree of asynchronicity, measured as the relative difference between the IRC values at which the
formation of both single bonds takes place. Accordingly, the 32CA reaction between AY-Cu 17 and MA 5 occurs through a nonconcerted two-stage one-step mechanism.64 Interestingly, structure S7, d(C1−C4) = 1.656 Å and d(C3−C5) = 2.671 Å (see Table S6 in the Supporting Information), is close to the structure of the IRC corresponding to the first maximum of the second derivative of the energy with respect to the reaction coordinate (see the blue line in Figure 7).65 This structure, which is very similar to IN-Cu-mn (see Figure 4), splits the IRC into the two stages in which the mechanism is characterized. (v) Along the C1−C4 single bond formation, the C1 pseudoradical center of AY-Cu 17 contributes to a larger extent than the C4 one created at the MA 5. This behavior is a consequence of the facility for the creation of the C1 pseudoradical center in AY-Cu 17. Note that, at the GS of AYs 1 and 16, the C1 pseudoradical center is already present (see section 3.1). And last, (vi) a comparative analysis of the bonding changes taking place along the 32CA reactions of AY 16 (see section 2 in the Supporting Information) and AY-Cu 17 with MA 5 allows the establishment of a great similarity between the two mechanisms: one pseudoradical center in the AY framework along the reaction path before the formation of the first C1−C4 single bond suggests that both 32CA reactions follow a pmr-type mechanism. The main difference is found in the formation of the first C1−C4 single bond; the later character of TS′-Cu-mn than that of TSmn, as well as the higher activation energy found in the coppermetalated 32CA reaction, can be associated with the higher energy cost demanded for the creation of the C1 pseudoradical center in AY-Cu 17 (see Tables S5 and S6 in the Supporting Information). 10970
DOI: 10.1021/acs.joc.8b01605 J. Org. Chem. 2018, 83, 10959−10973
Article
The Journal of Organic Chemistry
■
CONCLUSIONS The 32CA reactions of copper-metalated AY-Cu 17, in which the Cu(I) cation was explicitly solvated with two ether molecules, and AY-CuL 18, in which the Cu(I) cation was coordinated to the bidentated bisoxazoline 19, with MA 5 have been studied within MEDT at the MPWB1K/6-311G(d,p) computational level in order to shed light on the role of the metalation of AYs in these 32CA reactions. ELF comparative analysis of the electronic structure of AY 16 and copper-metalated AY-Cu 17 indicates that metalation of AY 16 provokes the loss of the pseudodiradical character of AY 16, thus decreasing the high reactivity of AY 16. On the other hand, QTAIM analysis of copper-metalated AY-Cu 17 reveals the ionic nature of the Cu(I)−N interaction in metalated AYs, thereby characterizing the AY framework as an anionic structure. Analysis of the CDFT reactivity indices indicates that metalation of AY 16 notably increases the nucleophilicity of AY-Cu 17 and AY-CuL 18 due to the anionic character of the AY framework. All these 32CA reactions take place through stepwise mechanisms, starting with the formation of an earlier MC. In general, after formation of the MC, the cycloaddition process takes place in a single elementary step. Only the meta/endo reaction path of the 32CA reaction of AY-Cu 17 with MA 5 takes place in two consecutive elementary steps given the strong Cu(I)−O interaction already present at the corresponding MC. However, the ring closure step associated to this mechanism has an unappreciable energy barrier, 0.5 kcal·mol−1. An analysis of the relative energies involved in the three studied 32CA reactions makes it possible to obtain some relevant conclusions: (i) The activation energies associated with the two metalated reaction models, 4.1 (AY-Cu 17) and 2.6 (AY-CuL 18) kcal·mol−1, are close to that found in the nonmetalated process, 3.3 (AY 16) kcal·mol−1, indicating that metalation does not have any effect on the reaction rate. (ii) The two metalated 32CAs reaction are completely regioselective, yielding only the experimentally observed 4-pyrrolidines. (iii) The 32CA reaction involving the solvated Model I presents high endo stereoselectivity due to the favorable electronic interaction present between the basic carboxyl oxygen of MA 5 and the acidic Cu(I) cation at the endo TS-Cu-mn. However, the use of bulky bidentated ligands or bulky ethylene derivatives can turn these 32CA reactions exo stereoselective. Finally, (iv) the copper metalation of AY 16 considerably reduces the exothermic character of the reaction. Analysis of the geometric parameters of the TSs involved in these 32CA reactions indicates that metalation of AY 16 slightly increases the asynchronicity and the late character of the TS. The high asynchronicity found at the more favorable regioisomeric meta TSs indicates that they are associated to a nonconcerted two-stage one-step mechanism, in which the formation of the first C−C single bond involves the most nucleophilic center of these AYs, i.e. the carboxyl substituted C1 carbon, and the most electrophilic center of MA 5, the βconjugated C4 carbon, a behavior anticipated by the analysis of the Parr functions. The GEDT values at the TSs indicate that the metalation enhances the polar character of these 32CA reactions because of the increase of the nucleophilic character of AY-Cu 17 and AYCuL 18 due to their anionic character, but this behavior does not have any effect on the reaction rate.
The ELF topological analysis of the selected structures of the IRC involved in the C−C bond formation processes along the most favorable meta/endo regioisomeric reaction paths of the nonmetalated and metalated reaction Model I permits establishing the two-stage one-step nature of the mechanism of these 32CA reactions yielding 4-pyrrolidines. ELF topological analysis of the corresponding TSs points out that the formation of the first C1− C4 single bond has not started yet. Although AY-Cu 17 does not have any pseudoradical center, both the activation energy associated to TS′-Cu-mn and the BET study along the reaction path indicate that these metalated reactions are associated to pmr-type 32CA reactions. Consequently, both the substitution in the simplest AY 1 and the copper metalation change the mechanism from pdr-type for the simplest AY 1 to pmr-type for AY-Cu 17, explaining the moderate deceleration of these 32CA reactions. This finding rules out any catalytic role of the Cu(I) cation in the kinetics of 32CA reactions of metalated AYs. From an experimental point of view we can conclude that the Cu(I) metalation of the AYs does not produce any catalytic effect on the reaction rate of these 32CA reactions, but it has a significant influence on the regio- and stereoselectivity. While the 32CA reaction of Cu(I) metalated AYs with electrophilic ethylenes becomes completely regioselective, the stereoselectivity depends on several factors such as the bulk of the ligand and the electronic structure of the ethylene derivative.
■
ASSOCIATED CONTENT
S Supporting Information *
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.joc.8b01605. Study of the 32CA reaction of AY 16 with MA 5. Tables with MPWB1K/6-311G(d,p) total electronic energies, in gas phase and in solvent, for the stationary points involved in the 32CA reactions of AY 16; copper-metalated AY-Cu 17; copper-metalated AY-CuL 18 with MA 5. BET studies of the 32CA reactions of the AY 16 with MA 5 and copper-metalated AY-Cu 17 and MA 5. MPWB1K/6311G(d,p) gas phase total energies, the only imaginary frequency of the TSs, and Cartesian coordinates of the stationary points involved in the 32CA reactions of AY 16, AY-Cu 17, and AY-CuL 18 with MA 5 (PDF)
■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. ORCID
Luis R. Domingo: 0000-0002-2023-0108 Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS Work supported by the Ministry of Economy and Competitiveness (MINECO) of the Spanish Government, Project CTQ2016-78669-P (AEI/FEDER, UE) and Fondecyt (Chile) Grant 1180348. Cooperación Internacional of Fondecyt is also thanked for continuous support (L.R.D.). M.R.-G. also thanks MINECO for a postdoctoral contract cofinanced by the European Social Fund (BES-2014-068258).
■
REFERENCES
(1) Padwa, A. 1,3-Dipolar Cycloaddition Chemistry; Wiley-Interscience: New York, 1984; Vols. 1−2. 10971
DOI: 10.1021/acs.joc.8b01605 J. Org. Chem. 2018, 83, 10959−10973
Article
The Journal of Organic Chemistry (2) Padwa, A.; Pearson, W. H. Synthetic Applications of 1,3-Dipolar Cycloaddition Chemistry Toward Heterocycles and Natural Products; John Wiley & Sons, Inc.: New York, NY, USA, 2002; Vol. 59. (3) Bailly, C. Lamellarins, from A to Z: A Family of Anticancer Marine Pyrrole Alkaloids. Curr. Med. Chem.: Anti-Cancer Agents 2004, 4, 363− 378. (4) Bellina, F.; Rossi, R. Synthesis and biological activity of pyrrole, pyrroline and pyrrolidine derivatives with two aryl groups on adjacent positions. Tetrahedron 2006, 62, 7213−7256. (5) Narayan, R.; Potowski, M.; Jia, Z.-J.; Antonchick, A. P.; Waldmann, H. Catalytic Enantioselective 1,3-Dipolar Cycloadditions of Azomethine Ylides for Biology-Oriented Synthesis. Acc. Chem. Res. 2014, 47, 1296−1310. (6) Domingo, L. R.; Emamian, S. R. Understanding the mechanisms of [3 + 2] cycloaddition reactions. The pseudoradical versus the zwitterionic mechanism. Tetrahedron 2014, 70, 1267−1273. (7) Domingo, L. R.; Chamorro, E.; Pérez, P. Understanding the High Reactivity of the Azomethine Ylides in [3 + 2] Cycloaddition Reactions. Lett. Org. Chem. 2010, 7, 432−439. (8) Coldham, I.; Collis, A. J.; Mould, R. J.; Robinson, D. E. Pyrrolidines by 1,3-Dipolar Cycloaddition of Conjugated Azomethine Ylides. Synthesis 1995, 1995, 1147−1150. (9) Grigg, R.; Kemp, J.; Sheldrick, G.; Trotter, J. 1,3-Dipolar cycloaddition reactions of imines of α-amino-acid esters: X-ray crystal and molecular structure of methyl 4-(2-furyl)-2,7-diphenyl-6,8-dioxo3,7-diazabicyclo[3.3.0]octane-2-carboxylate. J. Chem. Soc., Chem. Commun. 1978, 109−111. (10) Grigg, R.; Kemp, J. X = Y-ZH systems as potential 1,3-dipoles the stereochemistry and regioselectivity of cycloaddition reactions of imines of alfa-amino-acid esters. Tetrahedron Lett. 1980, 21, 2461− 2464. (11) Grigg, R. New Prototropic processes in the synthesis of heterocyclic compounds. Bull. Soc. Chim. Belg. 1984, 93, 593−603. (12) Tsuge, O.; Kanemasa, S. Advances in Heterocyclic Chemistry; Academic Press: San Diego, 1989; Vol. 45. (13) Shuji, K.; Hidetoshi, Y. Asymmetric 1,3-dipolar cycloadditions of azomethine ylides with a chiral electron-deficient olefinic dipolarophile. Tetrahedron Lett. 1990, 31, 3633−3636. (14) Longmire, J. M.; Wang, B.; Zhang, X. Highly Enantioselective Ag(I)-Catalyzed [3 + 2] Cycloaddition of Azomethine Ylides. J. Am. Chem. Soc. 2002, 124, 13400−13401. (15) Cabrera, S.; Arrayás , R. G.; Carretero, J. C. Highly Enantioselective Copper(I)−Fesulphos-Catalyzed 1,3-Dipolar Cycloaddition of Azomethine Ylides. J. Am. Chem. Soc. 2005, 127, 16394− 16395. (16) Kim, H. Y.; Shih, H.-J.; Knabe, W. E.; Oh, K. Reversal of Enantioselectivity between the Copper(I)- and Silver(I)- Catalyzed 1,3-Dipolar Cycloaddition Reactions Using a Brucine-Derived Amino Alcohol Ligand. Angew. Chem., Int. Ed. 2009, 48, 7420−7423. (17) Adrio, J.; Carretero, J. C. Novel dipolarophiles and dipoles in the metal-catalyzed 1,3-dipolar cycloaddition of azomethine ylides. Chem. Commun. 2011, 47, 6784−6794. (18) Koizumi, A.; Kimura, M.; Arai, Y.; Tokoro, Y.; Fukuzawa, S. Copper- and Silver-Catalyzed Diastereo- and Enantioselective Conjugate Addition Reaction of 1-Pyrroline Esters to Nitroalkenes: Diastereoselectivity Switch by Chiral Metal Complexes. J. Org. Chem. 2015, 80, 10883−10891. (19) Tang, L.-W.; Zhao, B.-J.; Dai, L.; Zhang, M.; Zhou, Z.-M. Asymmetric Construction of Pyrrolidines Bearing a Trifluoromethylated Quaternary Stereogenic Center via CuI Catalyzed 1,3-Dipolar Cycloaddition of Azomethine Ylides with b-CF3-b,b-Disubstituted Nitroalkenes. Chem. - Asian J. 2016, 11, 2470−2477. (20) Ponce, A.; Alonso, I.; Adrio, J.; Carretero, J. C. Chem. - Eur. J. 2016, 22, 4952−4959. (21) Pascual-Escudero, A.; de Cózar, A.; Cossio, F. P.; Adrio, J.; Carretero, J. C. Alkenyl Arenes as Dipolarophiles in Catalytic Asymmetric 1,3-Dipolar Cycloaddition Reactions of Azomethine Ylides. Angew. Chem., Int. Ed. 2016, 55, 15334−15338.
(22) Cayuelas, A.; Ortiz, R.; Nájera, C.; Sansano, J. M.; Larrañaga, O.; de Cózar, A.; Cossío, F. C. Enantioselective Synthesis of Polysubstituted Spiro-nitroprolinates Mediated by a (R,R)-Me-DuPhos·AgFCatalyzed 1,3-Dipolar Cycloaddition. Org. Lett. 2016, 18, 2926−2929. (23) Corpas, J.; Ponce, A.; Adrio, J.; Carretero, J. C. Cu1-Catalyzed Asymmetric [3 + 2] Cycloaddition of Azomethine Ylides with Cyclobutenones. Org. Lett. 2018, 20, 3179−3182. (24) Allway, P.; Grigg, R. Chiral Co(II) and (Mn(II) catalysts for the 1,3-dipolar cycloaddition reactions of azomethine ylides derived from arylidene imines of glycine. Tetrahedron Lett. 1991, 32, 5817−5820. (25) Gothelf, A. S.; Gothelf, K. V.; Hazell, R. G.; Jorgensen, K. A. Catalytic Asymmetric 1,3-Dipolar Cycloaddition Reactions of Azomethine Ylides−A Simple Approach to Optically Active Highly Functionalized Proline Derivatives. Angew. Chem., Int. Ed. 2002, 41, 4236−4238. (26) He, Z.-L.; Sheong, F. K.; Li, Q.-H.; Lin, Z.; Wang, C.-J. Exoselective 1,3-Dipolar [3 + 6] Cycloaddition of Azomethine Ylides with 2-Acylcycloheptatrienes: Stereoselectivity and Mechanistic Insight. Org. Lett. 2015, 17, 1365−1368. (27) Xu, S.; Zhang, Z.-M.; Xu, B.; Liu, B.; Liu, Y.; Zhang, J. Enantioselective Regiodivergent Synthesis of Chiral Pyrrolidines with Two Quaternary Stereocenters via Ligand-Controlled Copper(I)Catalyzed Asymmetric 1,3-Dipolar Cycloadditions. J. Am. Chem. Soc. 2018, 140, 2272−2283. (28) He, Z.-L.; Sheong, F. K.; Li, Q.-H.; Lin, Z.; Wang, C.-J. Exoselective 1,3-Dipolar [3 + 6] Cycloaddition of Azomethine Ylides with 2-Acylcycloheptatrienes: Stereoselectivity and Mechanistic Insight. Org. Lett. 2015, 17, 1365−1368. (29) Vivanco, S.; Lecea, B.; Arrieta, A.; Prieto, P.; Morao, I.; Linden, A.; Cossío, F. P. Origins of the Loss of Concertedness in Pericyclic Reactions: Theoretical Prediction and Direct Observation of Stepwise Mechanisms in [3 + 2] Thermal Cycloadditions. J. Am. Chem. Soc. 2000, 122, 6078−6092. (30) Emamian, S. How the mechanism of a [3 + 2] cycloaddition reaction involving a stabilized N-lithiated azomethine ylide toward a pdeficient alkene is changed to stepwise by solvent polarity? What is the origin of its regio- and endo stereospecificity? A DFT study using NBO, QTAIM, and NCI analyses. RSC Adv. 2016, 6, 75299−75314. (31) Dondas, H. A.; Durust, Y.; Grigg, R.; Slater, J. M.; Sarker, M. A. B. XYZH systems as potential 1,3-dipoles. Part 62:1 1,3-Dipolar cycloaddition reactions of metallo-azomethine ylides derived from αiminophosphonates. Tetrahedron 2005, 61, 10667−10682. (32) Cayuelas, A.; Larrañaga, O.; Selva, V.; Nájera, C.; Akiyama, T.; Sansano, J. M.; de Cózar, A.; Miranda, J. I.; Cossío, F. P. Cooperative Catalysis with Coupled Chiral Cycloadditions of Azomethine Ylides. Chem. - Eur. J. 2018, 24, 8092−8097. (33) Domingo, L. R. Molecular Electron Density Theory: A Modern View of Reactivity in Organic Chemistry. Molecules 2016, 21, 1319. (34) Zhao, Y.; Truhlar, G. D. Hybrid Meta Density Functional Theory Methods for Thermochemistry, Thermochemical Kinetics, and Noncovalent Interactions: The MPW1B95 and MPWB1K Models and Comparative Assessments for Hydrogen Bonding and van der Waals Interactions. J. Phys. Chem. A 2004, 108, 6908−6918. (35) Hehre, W. J.; Radom, L.; Schleyer, P. v. R.; Pople, J. A. Ab initio Molecular Orbital Theory; Wiley: New York, 1986. (36) Schlegel, H. B. Optimization of equilibrium geometries and transition structures. J. Comput. Chem. 1982, 3, 214−218. (37) Schlegel, H. B. Modern Electronic Structure Theory; Yarkony, D. R., Ed.; World Scientific Publishing: Singapore, 1994. (38) Fukui, K. Formulation of the reaction coordinate. J. Phys. Chem. 1970, 74, 4161−4163. (39) González, C.; Schlegel, H. B. Reaction path following in massweighted internal coordinates. J. Phys. Chem. 1990, 94, 5523−5527. (40) González, C.; Schlegel, H. B. Improved algorithms for reaction path following: Higher-order implicit algorithms. J. Chem. Phys. 1991, 95, 5853−5860. (41) Tomasi, J.; Persico, M. Molecular-interactions in solution - An overview of methods based on continuous distributions of the solvent. Chem. Rev. 1994, 94, 2027−2094. 10972
DOI: 10.1021/acs.joc.8b01605 J. Org. Chem. 2018, 83, 10959−10973
Article
The Journal of Organic Chemistry (42) Simkin, B. Y.; Sheikhet, I. Quantum Chemical and Statistical Theory of Solutions-A Computational Approach; Ellis Horwood: London, 1995. (43) Cossi, M.; Barone, V.; Cammi, R.; Tomasi, J. Ab initio study off solvated molecules: a new implementation of polarizable continuum model. Chem. Phys. Lett. 1996, 255, 327−335. (44) Cances, E.; Mennucci, B.; Tomasi, J. A new integral equation formalism for the polarizable continuum model: Theoretical background and applications to isotropic and anisotropic dielectrics. J. Chem. Phys. 1997, 107, 3032−3041. (45) Barone, V.; Cossi, M.; Tomasi, J. Geometry optimization of molecular structures in solution by the polarizable continuum model. J. Comput. Chem. 1998, 19, 404−417. (46) Reed, A. E.; Weinstock, R. B.; Weinhold, F. Natural-population analysis. J. Chem. Phys. 1985, 83, 735−746. (47) Reed, A. E.; Curtiss, L. A.; Weinhold, F. Intermolecular interactions from a natural bond orbital, donor-acceptor viewpoint. Chem. Rev. 1988, 88, 899−926. (48) Becke, A. D.; Edgecombe, K. E. A simple measure of electron localization in atomic and molecular systems. J. Chem. Phys. 1990, 92, 5397−5403. (49) Bader, R. F. W. Atoms in Molecules. A Quantum Theory; Claredon Press: Oxford, U.K., 1990. (50) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.; Kudin, K. N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.; Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian03; Gaussian, Inc.: Wallingford, CT, 2009. (51) Noury, S.; Krokidis, X.; Fuster, F.; Silvi, B. Computational tools for the electron localization function topological analysis. Comput. Chem. 1999, 23, 597−604. (52) Lu, T.; Chen, F. Multiwfn: A multifunctional wavefunction analyzer. J. Comput. Chem. 2012, 33, 580−592. (53) Domingo, L. R. A New C-C Bond Formation Model Based on the Quantum Chemical Topology of Electron Density. RSC Adv. 2014, 4, 32415−32428. (54) Geerlings, P.; De Proft, F.; Langenaeker, W. Conceptual Density Functional Theory. Chem. Rev. 2003, 103, 1793−1873. (55) Domingo, L. R.; Ríos-Gutiérrez, M.; Pérez, P. Applications of the Conceptual Density Functional Theory Indices to Organic Chemistry Reactivity. Molecules 2016, 21, 748. (56) Domingo, L. R.; Pérez, P.; Sáez, J. A. Understanding the local reactivity in polar organic reactions through electrophilic and nucleophilic Parr functions. RSC Adv. 2013, 3, 1486−1494. (57) Pauling, L. The Nature of the Chemical Bond An Introduction to Modern Structural Chemistry; Cornell University Press: New York, 1960. (58) Parr, R. G.; Pearson, R. G. Absolute hardness - Companion parameter to absolute electronegativity. J. Am. Chem. Soc. 1983, 105, 7512−7516. (59) Parr, R. G.; Yang, W. Density Functional Theory of Atoms and Molecules; Oxford University Press: New York, 1989. (60) Parr, R. G.; von Szentpaly, L.; Liu, S. Electrophilicity index. J. Am. Chem. Soc. 1999, 121, 1922−1924. (61) Domingo, L. R.; Chamorro, E.; Pérez, P. Understanding the Reactivity of Captodative Ethylenes in Polar Cycloaddition Reactions. A Theoretical Study. J. Org. Chem. 2008, 73, 4615−4624.
(62) Johnson, E. R.; Keinan, S.; Mori-Sanchez, P.; Contreras-Garcia, J.; Cohen, J.; Yang, A. W. Revealing Noncovalent Interactions. J. Am. Chem. Soc. 2010, 132, 6498−6506. (63) Krokidis, X.; Noury, S.; Silvi, B. Characterization of elementary chemical processes by catastrophe theory. J. Phys. Chem. A 1997, 101, 7277−7282. (64) Domingo, L. R.; Saéz, J. A.; Zaragozá, R. J.; Arnó, M. Understanding the Participation of Quadricyclane as Nucleophile in Polar [2σ + 2σ + 2π] Cycloadditions toward Electrophilic π Molecules. J. Org. Chem. 2008, 73, 8791−8799. (65) Yepes, D.; Murray, J. S.; Pérez, P.; Domingo, L. R.; Politzer, P.; Jaque, P. Complementarity of reaction force and electron localization function analyses of asynchronicity in bond formation in Diels−Alder reactions. Phys. Chem. Chem. Phys. 2014, 16, 6726−6734.
10973
DOI: 10.1021/acs.joc.8b01605 J. Org. Chem. 2018, 83, 10959−10973