Study on Regio- and Diastereoselectivity of the 1,3-Dipolar Cycloaddition Reaction of Azomethine Ylide with 2-(Benzo[d]thiazol-2-yl)-3-(aryl)acrylonitrile: Synthesis, Spectroscopic, and Computational Approach

Abstract

An unprecedented and efficient three-component 1,3-dipolar cycloaddition reaction using (E)-2-(benzo[d]thiazol-2-yl)-3-(aryl)acrylonitriles 4a–g and an in situ generated azomethine ylide 3 from isatin and N-methylglycine is described. The reaction exhibits exclusive regioselectivity, resulting in the formation of 3′-(benzo[d]thiazol-2-yl)-1′-methyl-2-oxo-4′-(aryl)spiro[indoline-3,2′-pyrrolidine]-3′-carbonitriles regioisomers through exo/endo approaches. The diastereoselectivity of the reaction is highly dependent on the substitution pattern of the phenyl ring in dipolarophiles 4a–g, leading to the formation of exo-/endo-cycloadducts in varying ratios. To understand the stereoselectivity, the transition state structures were optimized using the TS guess geometry with the QST3-based method. The reaction mechanism and regioselectivity were elucidated by evaluating global and local electrophilicity and nucleophilicity descriptors at the B3LYP/cc-pVTZ level of theory, along with considerations based on the HSAB principle. The analysis of global electron density transfer (GEDT) showed that the reactions are polar and electron density fluxes from azomethine ylide 3 toward dipolarophile 4a–g. It was found from the molecular electrostatic potential map (MESP) that at the more favorable transition state, approach of reactants locates the oppositely charged regions over each other resulting in attractive forces between the two fragments. The computational results are consistent with the experimental observations, confirming that the reactions proceed through an asynchronous one-step mechanism.

Introduction

Spirocyclic scaffolds offer significant appeal in drug discovery due to their inherent three-dimensional structure, which facilitates interactions with three-dimensional binding sites more effectively than planar heteroaromatic systems used as ligands. This may explain why a considerable number of spiro compounds constitute a vital class of naturally occurring substances with distinct biological properties, having evolved to interact with proteins.¹ For several decades, the synthesis of pyrrolidine-based heterocycles has garnered significant interest, given their importance as a class of compounds exhibiting distinct biological activities.² Moreover, the incorporation of oxindole derivatives, particularly 3-spirooxindoles conjugated with functionalized pyrrolidines, constitutes a pivotal structural motif found in numerous alkaloids. These compounds serve as compelling targets in organic synthesis because of their diverse range of biological activities.³ Additionally, thiazole and its derivatives hold significant importance in the realms of medicinal and material chemistry. Combining thiazole with other heterocyclic moieties has demonstrated remarkable biological activities, including antimicrobial properties,⁴ antitumor effects,⁵ antispasmodic activity,⁶ and anti-inflammatory potential.⁷

The utilization of 1,3-dipolar cycloaddition involving azomethine ylides (AY) and electron-deficient olefins represents a highly convenient, straightforward, and efficient strategy for constructing highly functionalized pyrrolidine and pyrrolizine rings.⁸ In recent decades, significant advancements have been made in understanding the mechanism and fundamental principles underlying the 1,3-dipolar cycloaddition (32CA) reaction, particularly in terms of selectivity. Numerous studies have explored the interplay between theoretical and experimental aspects, contributing to the evolving comprehension of this process. Nevertheless, elucidating the mechanism of 1,3-dipolar cycloadditions remains a persistent challenge. Notably, Domingo made significant contributions to the understanding of 1,3-dipolar cycloaddition mechanisms by pioneering the use of the molecular electron density theory (MEDT).⁹ According to MEDT, molecular reactivity depends on the ability to change electron density. Thus, reactivity in organic chemistry can be described by a quantum chemical analysis of the changes in electron density and their energies along the reaction pathway.¹⁰ Domingo et al. used this methodology to analyze bond formation along the nonpolar and polar 32CA reactions.¹¹ They show that these reactions follow a two-stage one-step mechanism.¹²

As part of our ongoing exploration of 1,3-dipolar cycloaddition reactions,¹³ we present, for the first time, a straightforward and efficient method to access spirooxindole-pyrrolidine-benzothiazole heterocyclic hybrids TMs. The synthetic approach involves a one-pot, three-component reaction utilizing a 1,3-dipolar cycloaddition reaction of nonstabilized azomethine ylide, which is generated in situ through the decarboxylative condensation of isatin and N-methylglycine, with a series of (E)-2-(benzo[d]thiazol-2-yl)-3-(aryl)acrylonitriles (Scheme 1).

Scheme 1. Retrosynthetic Approach to the Target Spirooxindole-Pyrrolidine-Benzothiazoles TMs.

Theoretical analysis of all possible regio- and stereocycloaddition pathways were carried out by calculating global/local electrophilicity/nucleophilicity reactivity indices and the corresponding transition states (TSs) at the B3LYP/cc-pVTZ level of theory.

Results and Discussion

Synthetic Approach and Spectroscopic Verification

The 1,3-dipolar cycloaddition reaction of nonstabilized azomethine ylide 3, generated in situ via decarboxylative condensation of isatin (1) and N-methylglycine (2), with variously substituted (E)-2-(benzo[d]thiazol-2-yl)-3-(aryl)acrylonitrile in absolute ethanol afforded a series of novel 3′-(benzo[d]thiazol-2-yl)-1′-methyl-2-oxo-4′-(aryl)spiro[indoline-3,2′-pyrrolidine]-3′-carbonitriles in moderate to excellent yields (Scheme 2). Remarkably, the reaction displayed exclusive regioselectivity, yielding one of the two possible regioisomers, and high diastereoselectivity, affording two of the four possible diastereomers (Scheme 2 and Table 1).

Scheme 2. Synthesis of Diastereomeric Mixture of 3′-(Benzo[d]thiazol-2-yl)-1′-methyl-2-oxo-4′-(aryl)spiro[indoline-3,2′-pyrrolidine]-3′-carbonitriles 5 and 5′.

Table 1. 1,3-Dipolar Cycloaddition Reaction of Isatin (1) and N-Methylglycine (2) with Dipolarophiles 4a–g.

Entry	Reactants	Products	R	Time (h)	Yield (%)a	Regioselectivity (%)b	Stereoselectivity (%)b
						C1:C2 attack	5:5′
1	1 + 2 + 4a	5a	-H	4	55.4	100:0	100:0
2	1 + 2 + 4b	5b	-OCH3	5	83.2	100:0	100:0
3	1 + 2 + 4c	5c + 5′c	-CH3	4	92.8	100:0	57:43
4	1 + 2 + 4d	5d	-Cl	3	52.7	100:0	100:0
5	1 + 2 + 4e	5e	-F	2	96.5	100:0	100:0
6	1 + 2 + 4f	5f + 5′f	-CN	3	96.2	100:0	68:32
7	1 + 2 + 4g	5g + 5′g	-NO2	3	86.4	100:0	88:12

^aCombined yield of isolated cycloadducts.

^bSeparated by column chromatography workup.

The reaction was found to be regiospecific, consistently leading to the formation of a single regioisomer through C1-attack in all instances. Furthermore, it demonstrated high diastereoselectivity at the spirocenter, which depended on the substitution pattern of the phenyl ring in dipolarophiles 4a–g, resulting in the formation of exo-/endo-cycloadducts in varying ratios (Table 1). The diastereomeric products displayed distinct retardation factors, allowing for effective separation via column chromatography. In most cases, the exo-cycloadduct was obtained as the sole diastereomer, with the exception of dipolarophiles 4c, 4f, and 4g (Table 1, entries 3, 6, and 7), where the endo-cycloadduct was obtained as a minor diastereomer.

Consequently, to fully establish the regiochemical and stereochemical outcome of the reaction, extensive one-dimension (¹H-, ¹³C-, ¹³C-DEPT-45/90/135-NMR) and two-dimension homonuclear and heteronuclear correlation NMR spectrometry techniques were carried out (¹H–¹H-DQF-COSY, ¹H–¹³C-HSQC, ¹H–¹³C-HMBC, ¹H–¹H-NOESY, ¹H–¹H-ROESY) (see the Supporting Information). Using 3′-(benzo[d]thiazol-2-yl)-1′-methyl-4′-(4-nitrophenyl)-2-oxospiro[indoline-3,2′-pyrrolidine]-3′-carbonitriles (5g) and (5′g) as a representative structural models, the relevant spectra used for conclusive regiochemical and stereochemical structural analysis and elucidation are shown in Figures 1 and 2. Although collecting single-crystal X-ray diffraction data would comprise the proper means to determine the regiochemistry and absolute stereochemistry, this was not possible in this case since none of the products could be grown into crystals suitable for X-ray analysis.

Figure 1.

(a) Truncated one dimension and two-dimension NMR spectra of 3′-(benzo[d]thiazol-2-yl)-1′-methyl-4′-(4-nitrophenyl)-2-oxospiro[indoline-3,2′-pyrrolidine]-3′-carbonitrile (5g) (CDCl₃, 600 MHz): (a) ¹H NMR spectrum; (b) ¹³C-DEPT-135 spectrum, CH’s and CH₃’s (positive phase), CH₂’s (negative phase); (c) ¹H–¹H-gDQCOSY-NMR spectrum; (d) ¹H–¹³C-HSQC-NMR spectrum; (e) ¹H–¹³C-gHMBC-NMR spectrum; and (f) ¹H–¹H-ROESY-NMR spectrum.

Figure 2.

(a) Truncated one dimension and two dimension NMR spectra of 3′-(benzo[d]thiazol-2-yl)-1′-methyl-4′-(4-nitrophenyl)-2-oxospiro[indoline-3,2′-pyrrolidine]-3′-carbonitrile (5′g) (CDCl₃, 600 MHz): (a) ¹³C-DEPT-135 spectrum, CH’s and CH₃’s (positive phase), CH₂’s (negative phase); (b) ¹H–¹H-gDQCOSY-NMR spectrum; (c) ¹H–¹³C-HSQC-NMR spectrum; (d) ¹H–¹³C-gHMBC-NMR spectrum; and (e) ¹H–¹H-ROESY-NMR spectrum.

Regiochemical Assignment of Cycloadduct (5g)

Analysis of the ¹³C (see the Supporting Information, Figure S68), ¹³C-DEPT-45/90 (see the Supporting Information, Figures S69−S70), and ¹³C-DEPT-135 (Figure 1, spectrum b) NMR spectra confirmed the presence of 24 signals (10 aromatic CH’s, 7 aromatic quaternary carbons, 1 N-methyl, 1 nitrile, 1 carbonyl carbon, 2 quaternary sp³-hybridized centers, one methylene, and one methine carbons), which is agreeable with all carbons being magnetically nonequivalent except for C_2″ and C_3″ (Figure 1). The oxindoline carbonyl (C₂) appears at δ 174.5 ppm, and the spirocenter carbon (C₃/C_2′) of the pyrrolidine ring, which represents a key indicator of a successful annulation reaction, resonates farthest downfield (δ 78.6 ppm) amid all the carbon atoms of this ring which incidentally are the only nuclei exhibiting chemical shifts in the aliphatic region of the carbon spectrum besides the N-methyl group. The distinctive peak at δ 46.4 ppm has been attributed to the C_4′-H carbon since it is the sole aliphatic signal appearing in the ¹³C-DEPT-90 and was correlated to the most deshielded aliphatic proton signal appearing as a doublet of doublets at δ 5.93 (J = 10.2, 6.6 Hz, C_4′-H). Meanwhile, the methylene ¹³C_5′ chemical shift (C_5′/δ 56.6 ppm) was clearly identified as the only signal with a negative phase in the ¹³C-DEPT-135 spectrum with its attached diastereotopic protons H_a and H_b appearing as apparent triplet (δ 3.94, J = 10.2 Hz, C_5′-H_a) and doublet of doublets (δ 3.82, J = 9.6, 6.6 Hz, C_5′-H_b), respectively. The methylene protons were correlated in the ¹H–¹³C-HSQC-NMR spectrum by two contours to the corresponding C_5′ carbon (Figure 1, spectrum d). The nonequivalent pyrrolidine protons (C_4′-H, C_5′-H_a and C_5′-H_b) were correlated to the same spin system by ¹H–¹H-DQF-COSY (9-contour square in the aliphatic region, Figure 1, spectrum c), providing irrefutable evidence to support the suggested regiochemistry 5 and disprove the alternative regiochemical outcome 6 where the methylene and methine comprise isolated uncoupled spin systems.

The methine proton of C_4′-H (δ 5.93 ppm) shows strong long-range ¹H–¹³C heteronuclear multiple bond correlation (HMBC) (³J_CH) with spirocenter C₃/C_2′ (δ 78.6 ppm), benzothiazole C_7′′′ (δ 160.5 ppm), CN (δ 117.5 ppm), and C_2″ of the 4-nitrophenyl ring (δ 131.0 ppm).

Meanwhile, the methylene protons of C_5′H₂ show strong long-range ¹H–¹³C-HMBC (³J_CH) with N_1′-Me (δ 35.5 ppm), spirocenter C₃/C_2′, pyrrolidine C_3′ (δ 61.7 ppm), and C_1″ of the 4-nitrophenyl ring (δ 144.6 ppm). The ¹H chemical shifts of the 4-nitrophenyl protons (C_2″: δ 7.91; C_3″: δ 8.19 ppm) were identified through ¹H–¹H-COSY (4-contour square in the aromatic region, Figure 1, spectrum c), and their respective ¹³C chemical shifts were verified through correlation contours at δ 131.0 (C_2″) and 123.4 (C_3′′) ppm in the ¹H–¹³C-gHSQC NMR. The ¹³C chemical shift of the remaining C_4″ (δ 147.6 ppm) was confirmed through the ¹H–¹³C-HMBC cross peak (³J_CH) with C_2″-H.

Stereochemical Assignment of "exo" Cycloadduct (5g)

It was critical to assign all the chemical shifts of the aromatic protons of the benzothiazole and oxindoline rings ahead of attempting any kind of stereochemical analysis and determination. Pleasingly, the −NH group of the oxindoline moiety (δ 7.38 ppm/¹H NMR, Figure 1, spectrum a) provided the key entry point that ultimately led to unequivocal determination of the correct stereochemistry for the isolated diastereomer. Specifically, in the ¹H–¹H-ROESY spectrum (Figure 1, spectrum f) there exists a strong correlation contour between the −NH proton (δ 7.38 ppm) and H-7 (doublet, δ 6.73 ppm), and recognition of the latter triggered the identification of the remaining oxindoline protons in the same spin system and assignment of their relative position on the aromatic ring. Hence, while ¹H–¹H-COSY correlated H-7 to H-6 (triplet, δ 7.40 ppm), H-6 to H-5 (triplet, δ 7.26 ppm), and H-5 to H-4 (doublet, δ 8.00 ppm), the ¹H–¹³C-HSQC correlated the H-4 (δ 8.00/127.0), H-5 (δ 7.26/123.5), H-6 (δ 7.40/131.1), and H-7 (δ 6.73/109.9) to the corresponding ¹³C chemical shifts. Further strong proof for the preceding assignment order stems from a strong ¹H–¹³C-HMBC (³J_CH) correlation of H-4 with spirocenter carbon (C₃/C_2′). The remaining two quaternary carbons, C₈ (δ 141.9 ppm) and C₉ (δ 122.8 ppm), were identified through ¹H–¹³C-HMBC cross peaks (³J_CH) with C₄-H/C₆-H and C₅-H/C₇-H, respectively. The chemical shifts of the benzothiazole protons and carbons were identified using H_2′′′ as an entry point after recognizing H_2′′′ as the doublet at δ 7.69 ppm based on strong ¹H–¹³C-HMBC cross peak (³J_CH) with N-C_6′′′ (δ 7.69/151.9), which is anticipated to be the most downfield shift of the benzothiazole ring. Thus, while ¹H–¹H-COSY correlated H_2′′′ to H_3′′′ (triplet, δ 7.37 ppm), H_3′′′ to H_4′′′ (triplet, δ 7.47 ppm), and H_4′′′ to H_5′′′ (doublet, δ 8.06 ppm), ¹H–¹³C-HSQC correlated the H_2′′′ (δ 7.69/121.5), H_3′′′ (δ 7.37/126.2), H_4′′′ (δ 7.47/126.5), and H_5′′′ (δ 8.06/124.1) to the corresponding ¹³C chemical shifts. Additional verification for the foregoing assignment order stems from strong ¹H–¹³C-HMBC (³J_CH) correlations of H_4′′′ with N-C_6′′′ (δ 7.47/151.9) and both H_3′′′ and H_5′′′ with S–C_1′′′ (δ 8.06/7.37/135.6). The chemical shifts of the 4-nitrophenyl protons and carbons were identified using C_4′H as an entry point after identifying H_2″ as the doublet at δ 7.91 ppm based on strong ¹H–¹³C-HMBC (Figure 1, spectrum e) cross peak (³J_CH) of C_4′H with C_2″ (δ 5.93/131.0), where the C_2″ chemical shift was correlated in the ¹H–¹³C-HSQC (Figure 1, spectrum d) to H_2″ (δ 7.91) and H_2″ was correlated in the ¹H–¹H-COSY to H_3″ (δ 8.19). The ¹³C chemical shift of the latter was identified through the ¹H–¹³C-HSQC spectrum (δ 123.4). The remaining two quaternary carbons, C_1″ (δ 144.6 ppm) and C_4″ (δ 147.6 ppm), were identified through ¹H–¹³C-HMBC cross peaks (³J_CH) with C_3″-H/C_5′H₂ and C_2″-H, respectively. The stereochemistry of the nonequivalent methylene protons of C_5′, H_a (δ 3.94, app t, J = 10.2 Hz) and H_b (δ 3.82, 9.6, 6.6 Hz) was determined relative to C_4′-H based on two very strong cross peaks in the ¹H–¹H-ROESY spectrum between H_b/C_4′-H and the 4-nitrophenyl-H_2″, indicating the close proximity of the latter to H_b and C_4′-H compared to H_a.

Having assigned the proton and carbon chemical shifts for the benzothiazole, oxindoline, 4-nitrophenyl, and pyrrolidine rings, what remained was to inspect the ¹H–¹H-ROESY spectrum for any important spatial proximity cross peaks that could be utilized to assign the relative stereochemistry of the three stereogenic centers. As may be anticipated, cycloaddition reactions are inherently diastereoselective and the relative stereochemistry of the two stereogenic centers C_4′ and C_3′ are predictable based on the initial trans geometry of the dipolarophile, (E)-2-(benzo[d]thiazol-2-yl)-3-(aryl)acrylonitrile. Thus, the benzothiazole and 4-nitrophenyl rings are expected to be trans. Indeed the ¹H–¹H-ROESY spectrum shows no correlation contours between any of the protons of these two rings, ruling out syn stereochemistry. One, however, still expects the formation of two diastereomeric products due to the creation of a second spirocenter (C_2′/3) during the process. Notably, the ¹H NMR signals of C_4′-H, C_5′-H_a, and C₄-H, which incidentally have been previously used in related systems^13e to assign the relative stereochemistry of all chiral centers, were most informative and served also herein as key in assigning the relative stereochemistry. Thus, the complete absence of cross peaks between C_4′-H and C₄-H as well as between C_5′-H_a and C₄-H indicated a lack of proximity and an anti-relationship of C₄-H with both, C_5′-H_a and C_4′-H.

Regiochemical Assignment of "endo" Cycloadduct (5′g)

Having set up a dependable spectroscopic tactic to identify and correlate all proton and carbon signals for the preceding pure “exo” diastereomer of 3′-(benzo[d]thiazol-2-yl)-1′-methyl-4′-(4-nitrophenyl)-2-oxospiro[indoline-3,2′-pyrrolidine]-3′-carbonitrile (5g), regio- and stereochemical analysis of the second isomeric product isolated from the reaction followed a similar strategy. To this end, 1D and 2D NMR spectroscopy were employed to verify the structure obtained from the second fraction of 3′-(benzo[d]thiazol-2-yl)-1′-methyl-4′-(4-nitrophenyl)-2-oxospiro[indoline-3,2′-pyrrolidine]-3′-carbonitrile (5′g). The subsequent discussion provides detailed insights and accompanying NMR spectra that lend support to the proposed regio- and stereochemistry of 5′g.

Examination of the ¹³C-DEPT-135 (Figure 2, spectrum a), ¹³C (see the Supporting Information, S78), ¹³C-DEPT-90 (see the Supporting Information, Figure S79) NMR spectra confirmed the presence of 24 signals (10 aromatic CH’s, 7 aromatic quaternary carbons, 1 N-methyl, 1 nitrile, 1 carbonyl carbon, 2 quaternary sp³-hybridized centers, one CH₂, and one alicyclic methine), which is consistent with all carbon atoms being magnetically nonequivalent except for the 4-nitrophenyl carbons, C_2″ and C_3″ (Figure 2). The oxindoline carbonyl (C₂) appears at δ 176.4 ppm, and the spirocenter carbon (C₃/C_2′) of the pyrrolidine ring, which indicates a successful cyclization reaction, absorbed furthest downfield (δ 80.1 ppm) among all the carbon atoms of this ring which, incidentally, are the only nuclei exhibiting chemical shifts in the aliphatic region of the carbon spectrum besides the N-methyl group. The distinctive peak at δ 51.1 ppm has been attributed to the C_4′-H carbon since it is the sole aliphatic signal appearing in the ¹³C-DEPT-90 and was correlated to the most deshielded aliphatic proton signal appearing as doublet of doublets at δ 5.60 (dd, J = 9.6, 8.4 Hz, 1H, C_4′-H). Meanwhile, the methylene ¹³C_5′ chemical shift (C_5′/δ 57.3 ppm) was clearly recognized as the only signal with a negative phase in the ¹³C-DEPT-135 spectrum (Figure 2, spectrum a) with its attached nonequivalent diastereotopic protons H_a and H_b appearing as apparent triplet (δ 4.28, J = 10.2 Hz, C_5′-H_b) and doublet of doublets (δ 3.70, J = 9.6, 8.4 Hz, C_5′-H_a), respectively (Figure 2, spectrum c). The methylene protons were correlated in the ¹H–¹³C-HSQC-NMR spectrum by two contours to the corresponding C_5′ carbon (Figure 2, spectrum c). Strong diamagnetic anisotropic effect caused unexpected large separation between C_5′-H_a and C_5′-H_b. The nonequivalent pyrrolidine protons (C_4′-H, C_5′-H_a, and C_5′-H_b) were traced to the same spin system by ¹H–¹H-DQF-COSY (9-contour square in the aliphatic region, Figure 2, spectrum b), delivering undisputable evidence to back the suggested regiochemistry 5′ and negate the other regiochemical outcome 6′ where the methylene and methine comprise isolated uncoupled spin systems. The methine proton of C_4′-H (δ 5.60 ppm) shows strong long-range ¹H–¹³C heteronuclear multiple bond correlation (HMBC) (³J_CH) with benzothiazole C_7′′′ (δ 165.7 ppm), CN (δ 118.0 ppm), and C_2″ of the 4-nitrophenyl ring (δ 130.9 ppm) as well as HMBC correlation (²J_CH) with C_1″ (δ 143.9 ppm), C_3′ (δ 60.8 ppm), and C_5′ (δ 57.3 ppm). Meanwhile, the methylene protons of C_5′H₂ show strong long-range ¹H–¹³C-HMBC (³J_CH) with N_1′-Me (δ 35.4 ppm), spirocenter C₃/C_2′ (δ 80.1 ppm), pyrrolidine C_3′ (δ 60.8 ppm), and C_1″ of the 4-nitrophenyl ring (δ 143.9 ppm) as well as C_4′ (δ 51.1 ppm; ²J_CH). The ¹H chemical shifts of the 4-nitrophenyl protons (C_2″: δ 7.82; C_3′′: δ 8.21 ppm) were identified through ¹H–¹H-COSY (4-contour square in the aromatic region, Figure 2, spectrum b), and their respective ¹³C chemical shifts were verified through correlation contours at δ 130.9 (C_2″) and 122.7 (C_3″) ppm in the ¹H–¹³C-gHSQC NMR (Figure 2, spectrum c). The ¹³C chemical shift of the remaining C_4″ (δ 147.8 ppm) was confirmed through the ¹H–¹³C-HMBC cross peak (³J_CH) with C_2″-H.

Stereochemical Assignment of "endo" Cycloadduct (5′g)

It was crucial to determine all the chemical shifts of the aromatic protons of the benzothiazole and oxindoline rings ahead of attempting any sort of stereochemical analysis and determination. Pleasingly, the −NH group of the oxindoline moiety (δ 8.07 ppm/¹H NMR, see the Supporting Information) provided the key entry point that ultimately led to unequivocal determination of the correct stereochemistry for the isolated diastereomer. Specifically, in the ¹H–¹H-ROESY spectrum (Figure 2, spectrum e), there exists a strong correlation contour between the −NH proton (δ 8.07 ppm) and H-7 (doublet, δ 6.81 ppm), and identification of the latter initiated the recognition of the remaining oxindoline protons in the same spin system and assignment of their relative positions on the aromatic ring. Henceforth, while ¹H–¹H-COSY (Figure 2, spectrum b) correlated H-7 to H-6 (triplet, δ 7.10 ppm), H-6 to H-5 (triplet, δ 6.55 ppm), and H-5 to H-4 (doublet, δ 6.45 ppm), the ¹H–¹³C-HSQC (Figure 2, spectrum c) correlated the H-4 (δ 6.45/126.7), H-5 (δ 6.55/123.8), H-6 (δ 7.10/130.5), and H-7 (δ 6.81/110.7) to the corresponding ¹³C chemical shifts. Further strong proof for the preceding assignment order stems from a strong ¹H–¹³C-HMBC (³J_CH) correlation of H-4 with spirocenter carbon (C₃/C_2′) (Figure 2, spectrum d).

The remaining two quaternary carbons, C₈ (δ 141.4 ppm) and C₉ (δ 123.3 ppm), were identified through ¹H–¹³C-HMBC cross peaks (³J_CH) with C₄-H/C₆-H and C₅-H/C₇-H, respectively. In addition, the NH proton exhibited ¹H–¹³C-HMBC (³J_CH) correlation with C₉ and (²J_CH) correlation with C₈. The chemical shifts of the benzothiazole protons and carbons were identified using H_2′′′ as an access point after identifying H_2′′′ as the doublet at δ 7.74 ppm based on strong ¹H–¹³C-HMBC cross peak (³J_CH) with N-C_6′′′ (δ 7.74/151.8), which is anticipated to be the most downfield shift of the benzothiazole benzene ring. Thus, while ¹H–¹H-COSY correlated H_2′′′ to H_3′′′ (triplet, δ 7.42 ppm), H_3′′′ to H_4′′′ (triplet, δ 7.55 ppm), and H_4′′′ to H_5′′′ (doublet, δ 8.17 ppm), ¹H–¹³C-HSQC correlated the H_2′′′ (δ 7.74/121.6), H_3′′′ (δ 7.42/126.1), H_4′′′ (δ 7.55/126.7), and H_5′′′ (δ 8.17/123.8) to the corresponding ¹³C chemical shifts (Figure 2, spectrum b). Additional verification for the foregoing assignment order stems from strong ¹H–¹³C-HMBC (³J_CH) correlations of H_4′′′ with N-C_6′′′ (δ 7.55/151.8) and both H_3′′′ and H_5′′′ with S–C_1′′′ (δ 7.42/8.17/135.5) (Figure 2, spectrum d). The chemical shifts of the 4-nitrophenyl protons and carbons were identified using C_4′H as an entry point after identifying H_2″ as the doublet at δ 7.82 ppm based on strong ¹H–¹³C-HMBC (Figure 2, spectrum d) cross peak (³J_CH) of C_4′H with C_2″ (δ 5.60/130.9), where the C_2″ chemical shift was correlated in the ¹H–¹³C-HSQC (Figure 2, spectrum c) to H_2″ (δ 7.82) and H_2″ was correlated in the ¹H–¹H-COSY to H_3′′ (δ 8.21). The ¹³C chemical shift of the latter was identified through the ¹H–¹³C-HSQC spectrum (δ 122.8). The remaining two quaternary carbons, C_1″ (δ 143.9 ppm) and C_4″ (δ 147.7 ppm), were identified through ¹H–¹³C-HMBC cross peaks (³J_CH) with C_5′H₂ and C_2″-H, respectively. The stereochemistry of the nonequivalent methylene protons of C_5′, H_a (δ 3.70, dd, J = 9.6, 8.4 Hz), and H_b (δ 4.28, app t, J = 10.2 Hz) was determined relative to C_4′-H based on two very strong cross peaks in the ¹H–¹H-ROESY spectrum between H_b/C_4′-H and the 4-nitrophenyl-H_2″, indicating the close proximity of the latter to H_b and C_4′-H more so than H_a. Following the successful assignment of the proton and carbon chemical shifts for the benzothiazole, oxindoline, 4-nitrophenyl, and pyrrolidine rings, the ¹H–¹H-ROESY spectrum was inspected for crucial proximity correlation contours that could be utilized to assign the relative stereochemistry of the three stereogenic centers. As expected, the trans stereochemistry of the benzothiazole and 4-nitrophenyl (stereogenic centers C_4′ and C_3′) are predictable based on the initial E-geometry of the dipolarophile. Indeed the ¹H–¹H-ROESY spectrum shows no correlation cross peaks between any of the protons of these two rings, precluding the syn stereochemistry. However, two diastereomeric products will form due to the creation of a second spirocenter (C_2′/3) during the process. Notably, the ¹H NMR signals of C_4′-H, C_5′-H_a and C₄-H, which incidentally have been previously used in related systems^13e to assign the relative stereochemistry of all chiral centers, were most informative and served also herein as key in assigning the relative stereochemistry. Pleasingly, the presence of cross peaks between C_4′-H and C₄-H as well as between C_5′-H_a and C₄-H indicates proximity and a syn-relationship of the C₄-H with both, C_5′-H_a and C_4′-H.

While the 5g and 5′g diastereomers exhibited comparable ¹³C chemical shifts except for C₂ (δ 160.5 for 5g vs δ 165.7 for 5′g) and C_7′′′ (δ 174.5 for 5g vs δ 176.4 for 5′g), the magnetic anisotropy resulting from the syn stereochemistry of the neighboring benzothiazole and oxindoline aromatic rings due to the inversion of the C₃/C_2′ center impacted the chemical shifts of the oxindoline protons to a very significant extent in the 5g isomer. The C₄-H, C₅-H, and C₆-H of the 5g isomer experienced extreme upfield shifting from δ 8.00 ppm to 6.52 ppm, δ 7.26 ppm to 6.57 ppm, and δ 7.40 ppm to 7.09 ppm, respectively. On the contrary, the C₇-H experienced downfield shift from δ 6.73 ppm to 6.83 ppm. The significant upfield shift in the resonance of C₄-H, C₅-H, and C₆-H may be attributed to the shielding cone of the syn benzothiazole aromatic system, providing additional evidence to the assigned syn stereochemistry.

Reaction Mechanism

As proposed in Scheme 3, the regio- and stereochemistry for the formation of spiro[indoline-3,2′-pyrrolidine]-3′-carbonitriles 5/5′a rather than the corresponding regioisomers spiro[indoline-3,2′-pyrrolidine]-4′-carbonitriles 6/6′a are suggested based on electronic and steric factors. The cycloaddition reaction proceeds via “C1-attack” where the nucleophilic carbon of azomethine ylide is with the less hindered/more electrophilic carbon (C1) of the α,β,-alkenenitriles 4a–g. As indicated in Table 1, the exo-cycloadduct was consistently obtained as the major diastereomer in all instances.

Scheme 3. Regio- and Stereoisomeric Pathways for the 32CA Reaction between Azomethine Ylide 3 and Alkene 4a.

Computational Approach

Initially, two possible E and Z configurations of dipolarophiles 2-(benzo[d]thiazol-2-yl)-3-(aryl)acrylonitriles 4a–g were optimized where the E-configurations were ∼6.30–4.40 kcal/mol more stable than the Z-configurations at B3LYP/cc-pVTZ (Tables S98–S99). Since isomerization of the geometric isomers is not possible under the reaction conditions, calculations and discussion have been limited to the E-isomers.

Also, we have investigated the formation of dipole 3in situ by reaction of isatin (1) and N-methylglycine (2). The negative formation energy value confirms the formation of dipole 3. The optimization of dipole 3 for both the E- and Z-configuration at B3LYP/cc-pVTZ are shown in Table S1.

Analysis of Frontier Molecular Orbital (FMO) and Conceptual Density Functional Theory (CDFT) Reactivity Indices

Frontier molecular orbital (FMO) which was developed by Fukui¹⁴ is one of the best methods to analyze reactions. 1,3-Dipolar cycloaddition reactions are classified into three types on the basis of the relative FMO energies (HOMO/LUMO interaction) between the dipole and dipolarophile.¹⁵

On the basis of the frontier molecular orbital (FMO) theory, Sustmann classified cycloaddition processes into three distinct types. Type I involves the interaction between the HOMO of the dipole (HOMO_dipole) and the LUMO of the dipolarophile (LUMO_{dipolarophile}), also known as normal-electron demand (NED) 1,3-dipolar cycloaddition (DC) reactions. This classification encompasses a significant portion of 1,3-DC reactions. Type II is characterized by the interaction between the LUMO of the dipole and the HOMO of the dipolarophile, termed as inverse-electron demand (IED) 1,3-DC reactions. Lastly, type III involves dipole/dipolarophile pairs with similar HOMO and LUMO energies. In such cases, both LUMO dipolarophile, HOMO dipole and LUMO dipole, HOMO dipolarophile interactions may influence reactivity and regiochemistry, allowing for the occurrence of both NED and IED processes.

In this study, the HOMO–LUMO energy gaps propose that the LUMO_{(dipolarophile)}–HOMO_(dipole) interaction controls the cycloaddition reaction (normal electron demand reactions).¹³ The calculated energies of the frontier orbitals for dipole 3 and dipolarophiles 4a–g are presented in Table 2. It is clear that the energy gaps ΔE = LUMO_{(dipolarophiles 4a–g)} – HOMO_{(E-dipole 3)} lower than those of the corresponding Z-dipole (3) prove the reactions are cycloaddition of E-form of dipole (3) with the E-form of dipolarophiles 4a–g.

Table 2. Frontier Orbital Energies (eV) for Dipole 3 and Dipolarophiles 4a–g at the B3LYP/cc-pVTZ Level of Theory.

^aΔE = LUMO_{(dipolarophile 4a–g)} – HOMO_(dipole 3) (normal electron demand).

^b(ΔE) = LUMO_(dipole 3) – HOMO_{(dipolarophile 4a–g)} (inverse electron demand).

In recent decades, conceptual density functional theory (CDFT) has provided a strong framework for the establishment and development of chemical reactivity theory.^16,17 The analysis of global and local CDFT reactivity indices in the ground state of reagents in the framework of DFT descriptors is a powerful tool for the description of molecular reactivity in polar processes,^17,18 especially for a previous study on the polar character of cycloadditions.¹⁹ Likewise, the understanding of electrophilicity/nucleophilicity indices in the context of DFT are powerful tools to explain the behavior of 1,3-dipolar cycloaddition reactions, and it has been successfully used to classify dipole and dipolarophile pairs used in such reactions.²⁰

The ability of a molecule to transfer charge in its ground state, approximated by Koopmans theory, can be described by the electronic chemical potential μ, which is defined as the arithmetic mean of one-electron energies of the frontier molecular orbitals HOMO and LUMO, as μ = −(I + A)/2.²¹ The chemical hardness η, which is a measure of the stability of a system, can be considered as η = (I – A) where I (ionization potential) = –E_HOMO and A (electron affinity) = –E_LUMO. η can be viewed as a measure of the difficulty of changing the number of electrons in a system, which is conceptually similar to nonpolarizability or hardness. Since softness, S, is the opposite of hardness, it has been defined as the inverse of hardness as S = 1/η.²² The global electrophilicity index (ω), which measures the stabilization in energy when the system acquires an additional electronic charge from the environment, was given by the following simple expression, ω = μ²/2η, in terms of the electronic chemical potential μ and the chemical hardness η.²³ The global indices μ, η, S, and ω are calculated and are listed in Table 3.

Table 3. Global Properties and Global Electrophilicity/Nucleophilicity Indices Values for Dipole 3 and Dipolarophiles 4a–g Involved in the Cycloaddition Reactions^a.

^aAll computations were carried out with the Gaussian 09 suite of programs. Calculations based on the method of DFT were performed at the B3LYP/cc-pVTZ level of theory.

^bHOMO energy of tetracyanoethylene (TCE) is −0.34586 (in a.u.) at the same level of theory.

The electronic chemical potential, μ, of dipole 3 is higher than the α,β,-alkenenitriles 4a–g (−5.16 < μ < −4.29 eV); also dipole 3 with configuration “E” has an electronic chemical potential (μ = −3.35 eV) higher than its stereoisomer with configuration “Z” (μ = −3.45 eV).

Additionally, α,β,-alkenenitriles 4a–g act as electrophiles due to the larger value of their electrophilicity ω (2.67 < ω < 4.15) relative to the electrophilicity value of E-dipole 3 (ω = 1.71 eV), and therefore the charge transfer takes place from the E-dipole (3) to (E)-2-(benzo[d]thiazol-2-yl)-3-(aryl)acrylonitriles 4a–g as the dipolarophile.

Recently, Domingo et al. presented a unique electrophilicity scale to classify reagents involved in CA reactions.²⁴ Analysis of the developed electrophilicity indices ω presented in Table 3 indicates that α,β,-alkenenitriles 4a–g are among the strong electrophiles, while dipole 3 is mostly among the marginal electrophiles (nucleophiles), suggesting that in a polar 1,3-dipolar cycloaddition, α,β,-alkenenitriles 4a–g and dipole 3 act as electrophile and nucleophile, respectively. In accordance with this model, the polar character of a dipole-dipolarophile interaction can be evaluated by the difference Δω in the global electrophilicity of the two reagents.^24a It is clear that the 32CA of (3) as dipole with (4a–g) as dipolarophile (Δω are in the range of 0.96 and 2.44, ≪ 4.50 eV) are reactions controlled by a one-step mechanism.

Recently, Gázquez et al.²⁵ proposed new reactivity indices, defined as the electroaccepting, ω⁺ [ω⁺ = A∧2/(2(I – A))], and electrodonating, ω^– (ω^– = I∧2/(2(I – A)), powers. Where ω⁺ represents the measure of the tendency of a given system to accept charge, ω^– is the tendency to release charge. It is noteworthy that a greater ω⁺ value of a system reflects a better ability to accept charge, while a smaller value of ω⁺ corresponds to a better electron donor. Both quantities are calculated by employing the vertical ionization energy I and electron affinity A and collected in Table 3.

It is important to note that according to these definitions and as mentioned in Table 3, the α,β,-alkenenitriles 4f (ω⁺ = 1.67) and 4g (ω⁺ = 1.97) are the molecules with the greater ability to accept charge (strong electrophiles), while the (E)-dipole 3 (ω⁺ = 0.44) azomethine ylide is the molecule with the greater ability to donate charge (good nucleophile).

On the other hand, the low values of the ω index cannot be directly associated with the nucleophilic character. To obtain a working model descriptor for nucleophilicity, the empirical nucleophilicity index N proposed by the Kohn and Sham based on the HOMO energies can be defined as N = E_HOMO (eV) – E_HOMO(TCE) (eV).²⁶

The nucleophilicity values N for the azomethine ylide dipole 3 and our series of dipolarophiles 4a–g are presented in Table 3. The nucleophilicity of (E)-dipole 3 (Ν = 4.42 eV) shows that it is most nucleophilic toward dipolarophiles 4a–g.

In addition, we calculated the nucleophilicity based on the assumption of Chattaraj et al. that electrophilicity and nucleophilicity are inversely related.^27,28 The proposed nucleophilicity parameter was described as the multiplicative inverse of the electrophilicity index (ω) and denoted as N′ = 1/ω.^24,29

In this sense, Roy et al. proposed the nucleophilicity index, N″, as the reciprocal of the electrodonating ω^– potency.³⁰ Since the nucleophilicity index determined as 1/ω^– was less than 1, the nucleophilicity index N″ has recently been defined as N″ = ((1/ω^–) × 10).^30,31 As shown in Table 3, the calculated N′ and N″ follow the same order as found for the corresponding nucleophilicity N descriptor. Thus, a comparison of the values obtained with the three nucleophilicity models, the nucleophilicity N values with the reciprocal of electrophilicity, N′, and Roy’s N″ values, shows reasonable agreement.

Moreover, the regiochemistry of the polar cycloaddition reactions can be investigated using local parameters of reactivity, including the condensed Fukui (f_k^±), Parr functions analysis (P_k^±), local electrophilicity ω_k = ωP⁺_k, and local nucleophilicity N_k = NP^-_k indices.³² The local electrophilic and nucleophilic functions for azomethine ylide dipole 3 and dipolarophiles 4a–g are summarized in Table 4. Figure 3 shows the atomic spin density (ASD) maps of the radical cation of azomethine ylide 3 and radical anions of alkenes 4a–g at B3LYP/cc-pVTZ.

Table 4. Nucleophilic and Electrophilic Fukui f_k^±, Local Softness s_k^±, Parr Indices P_k^±, Local Electrophilicity ω_k, and Local Nucleophilicity N_k for the Most Relevant Heavy Atoms of Azomethine Ylide 3 and α,β,-Alkenenitriles 4a–g^a.

^aCalculations based on the method of DFT were performed at the B3LYP/cc-pVTZ level of theory.

Figure 3.

Atomic spin density (ASD) maps of the radical cation of azomethine ylide 3 and radical anions of alkenes 4a–g at B3LYP/cc-pVTZ.

The nucleophilic Parr functions, P_k^–, of the azomethine ylide 3 are 0.36 (C3) and 0.32 (C4) and the electrophilic Parr functions P_k⁺ of alkene 4a–g are 0.23–0.40 (C1) and 0.05–0.10 (C2). Consequently, the C3 carbon is the most nucleophilic center of azomethine ylide 3 (N_k = 1.59) and C1 is the most electrophilic site of alkene 4 (ω_k = 0.95–1.16). Therefore, the most favorable regioisomeric pathway will be associated with the initial C^AY₃-C^alkene₁ bond formation. These results are consistent with experimental observations demonstrating that the 32CA process between azomethine ylides 3 and alkenes 4a–g occurs through interaction between C^AY₃-C^alkene₁ and C^AY₄-C^alkene₂, leading to the formation of the pyrrolidine derivatives 5a–g that facilitate pathway A (C1-attack) (Scheme 2).

To gain a deeper insight into the electrophilic and/or nucleophilic activation at the different sites of a molecule, Chattaraj et al. proposed the local reactivity difference index R_k, which can predict the local electrophilic and/or nucleophilic activation within an organic molecule.³³ The R_k index is defined as

a

b

From the data (Table 4) for dipolarophiles 4a–g, the C1 site is more electrophilic than C2 (C1, R_k = +0.69 – +1.07; C2, R_k = −0.45 – ±0.34).

On the other hand, the hard and soft acids and bases (HSAB) principle³⁴ and the local softness s_k^± can be used to predict the regioselectivity of 32CA reactions.³⁵ Accordingly, the interacting atoms for the preferred regioisomeric pathway of a 32CA reaction have approximately equal softness values. The local softness values s_k^± are calculated by s_k^± = S.f_k^±,³⁶ where S is the global softness and f_k^± are the respective Fukui functions. The softness matching index ΔS^cd_ab calculated by ΔS^cd_ab = (s^–_a – s⁺_c)² + (s^–_b – s⁺_d)². s^–_aAnd s^–_b are the softness values for the electrophilic attack for the two atoms of the dipole involved in the cycloaddition, while s⁺_c and s⁺_d are the softness for the nucleophilic attack for the two atoms of the dipolarophile involved in the 32CA reaction, with the lower value of ΔS^cd_ab indicating the preferred pathway. For 32CA of azomethine ylide 3 and alkene 4, the ΔS^cd_ab value for the formation of regioisomer 5 (ΔS¹²₃₄ = 0.95–1.36) is smaller than that for 6 (ΔS²¹₃₄ = 1.27–1.47), supporting pathway A (Table 5).

Table 5. Softness Matching Index ΔS^cd_ab of Possible Pathways for the Cycloaddition Reaction of Azomethine Ylide 3 and α,β-Alkenenitriles 4a–g.

Interestingly, a plot of electrophilicity difference (Δω = ω_(alkene) – ω_(E-dipole)) against softness matching index (ΔS¹²₃₄) shows a linear graph with a strong correlation (R² = 0.93) as shown in Figure 4, while no correlation occurs against a (ΔS²¹₃₄) facilitating pathway A (C1 attack).

Figure 4.

A plot of electrophilicity difference (Δω) against softness matching index (ΔS¹²₃₄) of CA addition reactions of azomethine ylide 3 and alkenes 4a–g.

Electrostatic Potential Distribution (ESP)

The molecular surface is a useful simulation technique for the model of quantitative molecular surface analysis, which allows us to distinguish the location of the electron density.³⁷ The electrostatic potential map helps to analyze the distribution of charges in several types of molecules, including organic substances. Lately, the molecular electrostatic potential (MESP) has been used to predict the probable site of molecules for cycloaddition reactions. While it has been used extensively to interpret and predict the reaction behavior of molecules, electrostatic effects can also contribute to regioselectivities.³⁸

To explore the local reactivity indices mentioned earlier, the electrostatic potential distribution (ESP) pattern of the starting materials azomethine ylide 3 and α,β,-alkenenitriles 4a–g has been generated and is presented in Figure 5. The surface area of the individual molecules has been established as the isosurface of density ρ = 0.001, and the color values are given in kcal/mol. The red color represents the zone with the negative potential of MEP, associated with reactive electrophilic sites, and the blue color is adapted to the zone with the positive potential and represents the suitable center of the nucleophilic attacks.

Figure 5.

Electrostatic potential (ESP) maps of azomethine ylide 3 and alkenes 4a–g (B3LYP/cc-pVTZ, kcal/mol).

From Figure 5, it is evident that the C1 center in alkenes 4a–g are located in the zone with the positive potential, confirming their high electrophilicity indices to be the more proper center of nucleophilic attack compared to the C2 center, supporting formation of regioisomeric adduct 5 as a major product (pathway A).

These results were found to be in exact agreement with the experimental observations, thus confirming the accuracy of the calculated descriptors for the local electrophilicity and nucleophilicity of the reactants. These descriptors successfully explained the regioselective behavior of the dipoles and dipolarophiles in the 32CA reaction and provided insights into the underlying mechanism of regioselectivity.

Energies of Transition-State Structures

Although the FMO theory, based on electronic factors without consideration of steric factors, provides a good basis for understanding the regioselectivity of 32CA reactions, in many cases steric factors control the regiochemistry of the reaction.³⁹

Experimentally, it was observed that the cycloaddition reactions of isatin (1), N-methylglycine (2), and dipolarophiles 4a–g are regiospecific. Only a single regioisomer of 5a–g was obtained through pathway A (C1-attack), while the formation of the other regioisomer 6a–g was not observed at all.

Hypothetically, the reaction could proceed via pathway A (C1-attack) or pathway B (C2-attack), and in both cases the orientation of the reactants could be endo or exo. Therefore, a computational study was carried out to determine the energetics of reaction pathways A and B through the exo and endo transition states for the formation of possible regioisomers 5a/5′a and 6a/6′a, respectively, starting from azomethine ylide 3 and alkenes 4a (Figure 6). The stereoselectivity of the reaction of azomethine ylide 3 and alkene 4a can be rationalized by considering the activation and reaction energies, enthalpies, and Gibbs free energies of the possible pathways A and B for the formation of the possible stereoisomeric products 5/5′ and 6/6′ via the exo- and endo-TS, respectively (Table 6). In the gas phase, the calculated relative Gibbs free energies of activation associated with possible pathways of this 32CA reaction at B3LYP/cc-pVTZ are 22.27 (5(TS-I)), 26.77 (5(TS-II)), 33.16 (6(TSI)), and 33.33 kcal/mol (6(TS-II)). The TSs associated with regioisomeric pathway 5 show lower activation energies than pathway 6. This finding indicates that regioisomeric pathway 5 is the exclusive product of the cycloaddition reaction, demonstrating high regioselectivity in all cases. However, for both the regioisomeric pathway, the exo approaches have smaller activation free energies than endo approaches. In the gas phase, the free energy of 5(TS-I) is 4.50 kcal/mol lower than that of 5(TS-II), favoring the formation of the stereoisomer 5a as a major product which is in agreement with experimental outcomes (Scheme 3). The geometries of the products and TSs for the possible reaction pathways of azomethine ylide 3 and alkene 4a are given in Figure 7.

Figure 6.

Relative energies (kcal/mol) for the reactants (3 + 4a), TSs, and products for the four possible reaction pathways.

Table 6. Calculated Electronic Activation Energies E_a, Reaction Enthalpies ΔH, Reaction Gibbs Free Energies ΔG, Reaction Energies ΔE_rxn, Activation Enthalpies ΔH^#, Activation Gibbs Free Energies ΔG^# (in kcal/mol) and Reaction Entropies ΔS (in cal/mol K), and Bond Length Differences ΔR (in Å) for the Cycloaddition Reaction of Azomethine Ylide 3 and Alkene 4a at the B3LYP/cc-pVTZ Level.

Structure	Ea	ΔH	ΔG	ΔErxn	ΔH#	ΔG#	ΔS	ΔR
5(TS-I)	7.36	–9.93	5.14	–12.05	8.03	22.27	–50.52	0.93
5(TS-II)	11.99	–9.53	6.18	–11.75	12.74	26.77	–52.70	0.92
6(TS-I)	18.65	–13.29	1.37	–15.61	19.12	33.16	–49.19	0.37
6(TS-II)	18.45	–10.78	4.29	–13.17	18.98	33.33	–50.57	0.60

Figure 7.

Optimized products and transition state geometries and global electron density transfer (GEDT) at exo-TS and endo-TS in the cycloaddition reaction of dipole 3 with alkene 4a in possible pathways A and B at the B3LYP/cc-pVTZ level. The C3······C1/C4·······C2 (pathway A) and C3·······C2/C4·······C1 (pathway B) bond distance (A°) are shown in dashed lines. The blue arrow indicates the direction of GEDT.

Table 6 also shows bond length differences (ΔR) of the two forming bonds of the possible TSs. Accordingly, regioisomeric pathway 5 is more asynchronous than the other pathway (pathway 6).

On the other hand, the polarity of the reaction seems to be an important factor that determines the feasibility of the reaction. The global electron density transfer (GEDT) was computed by the sum of the natural atomic charges (q), which were obtained by a natural population analysis (NPA), and in which the atoms belonged to each framework (f) at the TSs,⁴⁰ GEDT (f) = ∑_qϵfq, which fluxes from dipole 3 toward the alkene 4 framework. Along the possible reaction pathways, the GEDT values at the TSs given in Figure 7 are 0.32e at 5(TS-I), 0.28e at 5(TS-II), 0.25e at 6(TS-I), and 0.23e at 6(TS-II). These values highlight the highly polar nature of the reaction. A good correlation between GEDT values at TS and the computed relative free energy of activation (ΔG^#) can be established (R² = 0.95). Hence, this is consistent with the fact that the higher GEDT, the easier bonding changes, lower free energy of activation, and the faster the reaction,⁴⁰ which is in agreement with the experimental outcomes (Figure 8).

Figure 8.

Plot of the activation free energies vs GEDT at TSs for the cycloaddition reaction of dipole 3 with alkene 4a in possible pathways A and B at the B3LYP/cc-pVTZ level.

Furthermore, the observed stereochemistry can be explained by analyzing the molecular electrostatic potential (MESP) map and studying the electrostatic interactions. Figure 9 shows that in the energetically more favorable transition state 5(TS-I), approach of the reactants superimposes the oppositely charged regions, which leads to attractive forces between two fragments. On the other hand, in the energetically less favorable transition states 5(TS-II), 6(TS-I), and 6(TS-II), regions with the same charge are forced to lie over each other, which leads to repulsive forces between two fragments. Therefore, due to the electrostatic attractive forces between two interacting fragments, the formation of 5(TS-I) is more favorable than the formation of 5(TS-II), 6(TS-I), or 6(TS-II), which is fully consistent with the experimental results.

Figure 9.

Molecular electrostatic potential MESP maps of the transition states associated with the possible regioisomeric attacks. The blue and red colors display low and high electron density regions, respectively (ρ = 0.001 at B3LYP/cc-pVTZ, kcal/mol).

Conclusion

In summary, we have developed an efficient method for the synthesis of novel spirooxindole-pyrrolidine-benzothiazoles via a one-pot three-component regio- and stereoselective 1,3-dipolar cycloaddition reaction. This reaction involves the in situ generation of nonstabilized azomethine ylides by decarboxylative condensation of isatin and N-methylglycine, followed by their reaction with various (E)-2-(benzo[d]thiazol-2-yl)-3-(aryl)acrylonitriles. The obtained products were characterized using 1D and 2D-NMR spectroscopy, confirming their regio- and stereochemistry. The regiospecific cycloaddition reactions proceed via pathway A (C1-attack), resulting in a single regioisomer. The pure exo- and endo-products were isolated using chromatographic techniques. The computational analysis revealed that the exo-transition state displays the lowest activation energy among the potential pathways, which is in accordance with the experimental observations.

The investigation of the electrophilic/nucleophilic properties of azomethine ylide 3 and the α,β,-alkenenitriles 4a–g involved in the polar cycloaddition reactions was conducted. The mechanism, direction of charge transfer, and regioselectivity of these cycloaddition reactions were investigated using global and local electrophilicity and nucleophilicity indices as well as HSAB principle and agreed closely with the experimental results.

Specifically, the analysis of local reactivity indices indicated that the C3 carbon is the most nucleophilic center of azomethine ylide 3, while the C1 site of alkene 4 is the most electrophilic. These findings were consistent with the experimental observations, which demonstrated that the 1,3-dipolar cycloaddition process between azomethine ylide 3 and alkenes 4a–g occurs through interaction between C^AY₃-C^alkene₁ and C^AY₄-C^alkene₂, resulting in the formation of the pyrrolidine derivatives 5a–g.

The global electron density transfer (GEDT) analysis indicates that the reactions are polar and electron density fluxes from azomethine ylide 3 toward alkenes 4a–g. The MESP map showed that at the more favorable transition state 5(TS-I), approach of reactants locates the oppositely charged regions over each other, resulting in attractive forces between both reactants.

Experimental Section

All solvents purchased from Sigma-Aldrich are spectroscopic grade and used without further purification. Melting points were determined on a Stuart SMP3 melting point apparatus and are uncorrected. NMR spectra were recorded on a Bruker Avance III HD NMR spectrometer (600 MHz for ¹H, 150 MHz for ¹³C) in CDCl₃ solutions, with residual solvent signals as internal standard. Elemental analyses were performed on a Vario EL v2.3 elemental analyzer; the results were found to be in good agreement with the calculated values (±0.3%). Structural assignments were made with additional information from gCOSY, gHSQC, and gHMBC experiments. The dipolarophiles (E)-2-(benzo[d]thiazol-2-yl)-3-(aryl)acrylonitriles 4a–g were prepared according to literature.⁴¹

Computational Details

Geometry optimizations were computed on 1, 2, 3, 4a–g, 5a–g, and 6a–g in vacuum with density functional theory (DFT) using B3LYP exchange correlation functional⁴² combined with Dunning’s correlation consistent triple-ζ basis set (cc-pVTZ). Regional Fukui functions for electrophilic (f^-_k) and nucleophilic (f⁺_k) attacks were calculated employing the natural population analysis (NPA) at the B3LYP/cc-pVTZ level of theory.⁴³

The electrophilic, (P⁺_k), and nucleophilic, (P^-_k), Parr functions were obtained through the analysis of the Mulliken ASD of the radical anion and the radical cation by single-point energy calculations over the optimized neutral geometries using the unrestricted UB3LYP/cc-pVTZ formalism for radical species (Figure 2).^33d

Synthetic Procedure

General Procedure for the Synthesis of 3′-(Benzo[d]thiazol-2-yl)-1′-methyl-2-oxo-4′-(aryl)spiro[indoline-3,2′-pyrrolidine]-3′-carbonitriles

A mixture of isatin 1 (220 mg, 1.50 mmol), sarcosine 2 (133 mg, 1.50 mmol), and alkene 4 (1.25 mmol) in absolute ethanol (15 mL) was stirred at reflux in an oil bath for 2–5 h. After completion of the reaction (TLC), it was cooled to room temperature and the solvent was removed invacuo. The crude product was chromatographed on silica gel (high-purity grade, pore size 60 Å, 230–400 mesh particle size, 40–63 μm particle size) using dichloromethane as eluent to obtain the pure cycloadducts 5a–g, 5′c, 5′f, and 5′g.

(2′S,3′R,4′R)-3′-(Benzo[d]thiazol-2-yl)-1′-methyl-2-oxo-4′-phenylspiro[indoline-3,2′-pyrrolidine]-3′-carbonitrile (5a)

Yield: 302 mg (55.4%) from 327 mg of 4a; white crystals, mp 127–129 °C. ¹H NMR (CDCl₃, 600 MHz): δ 8.01 (d, J = 7.8 Hz, 1H, C₄-H), 7.99 (d, J = 7.8 Hz, 1H, C_5′′′-H), 7.71 (s, 1H, N₁-H), 7.69 (d, J = 7.2 Hz, 2H, 2 C-H_ph), 7.65 (d, J = 7.8 Hz, 1H, C_2′′′-H), 7.40 (t, J = 7.8 Hz, 1H, C_4′′′-H), 7.36–7.30 (m, 4H, 3C-H_ph, C₆-H), 7.26 (t, J = 7.8 Hz, 1H, C₅-H), 7.23 (t, J = 7.8 Hz, 1H, C_3′′′-H), 6.71 (d, J = 7.8 Hz, 1H, C₇-H), 5.71 (dd, J = 10.2, 7.2 Hz, 1H, C_4′-H), 3.87 (app t, J = 10.2 Hz, 1H, C_5′-H_a), 3.83 (dd, J = 9.6, 7.2 Hz, 1H, C_5′-H_b), 2.22 (s, 3H, N_1′-CH₃) ppm; ¹³C NMR (150 MHz, CDCl₃): δ 175.3 (q, C₂=O), 161.3 (q, C_7′′′), 152.1 (q, C_6′′′), 141.9 (q, C₈), 136.7 (q, C_1′′′), 135.5 (q, C_1″), 130.8 (C₆H), 129.9 (2C_2″-H_ph), 128.3 (2C_3″-H_ph), 128.0 (C_4″H_ph), 126.9 (C₄-H), 126.2 (C_4′′′-H), 125.8 (C_3′′′-H), 124.0 (C_5′′′-H), 123.5 (q, C₉), 123.3 (C₅-H), 121.4 (C_2′′′-H), 117.9 (CN), 109.8 (C₇H), 76.8 (C₃/C_2′), 62.0 (C_3′), 56.9 (C_5′H₂), 47.3 (C_4′H), 35.5 (N_1′-CH₃) ppm. Anal. Calcd for C₂₆H₂₀N₄OS: C, 71.54; H, 4.62; N, 12.83. Found: C, 71.31; H, 4.97; N, 12.97.

(2′S,3′R,4′R)-3′-(Benzo[d]thiazol-2-yl)-4′-(4-methoxyphenyl)-1′-methyl-2-oxospiro[indoline-3,2′-pyrrolidine]-3′-carbonitrile (5b)

Yield: 485 mg (83.2%) from 365 mg of 4b; paige crystals, mp 137–139 °C. ¹H NMR (CDCl₃, 600 MHz): δ 8.01 (d, J = 7.8 Hz, 1H, C₄-H), 8.00 (d, J = 7.8 Hz, 1H, C_5′′′-H), 8.01 (d, J = 9.0 Hz, 2H, Ar–C_2″-H_{p-methoxyphenyl}), 7.67 (d, J = 7.8 Hz, 1H, C_2′′′-H), 7.61 (d, J = 9.0 Hz, 2H, Ar–C_3″-H_{p-methoxyphenyl}), 7.50 (s, 1H, N₁-H), 7.41 (t, J = 7.8 Hz, 1H, C_4′′′-H), 7.34 (t, J = 7.8 Hz, 1H, C₆-H), 7.33 (d, J = 7.8 Hz, 1H, C_3′′′-H), 7.22 (t, J = 7.8 Hz, 1H, C₅-H), 6.73 (d, J = 7.8 Hz, 1H, C₇-H), 5.65 (dd, J = 10.2, 7.2 Hz, 1H, C_4′-H), 3.85 (app t, J = 10.2 Hz, 1H, C_5′-H_a), 3.82 (dd, J = 9.6, 7.2 Hz, 1H, C_5′-H_b), 3.76 (s, 3H, OCH₃), 2.24 (s, 3H, N_1′-CH₃) ppm; ¹³C NMR (150 MHz, CDCl₃): δ 175.2 (q, C₂=O), 161.4 (q, C_7′′′), 159.3 (C_4″-O), 152.2 (q, C_6′′′), 141.9 (q, C₈), 135.5 (q, C_1′′′), 131.0 (2CH_{p-methoxyphenyl}), 130.7 (C₆H), 128.4 (q, C_1″), 126.9 (C₄-H), 126.2 (C_4′′′-H), 125.8 (C_3′′′-H), 124.0 (C_5′′′-H), 123.6 (q, C₉), 123.3 (C₅-H), 121.4 (C_2′′′-H), 118.0 (CN), 113.7 (2CH_{p-methoxyphenyl}), 109.7 (C₇H), 78.7 (C₃/C_2′), 62.2 (C_3′), 57.2 (C_5′H₂), 55.2 (OCH₃), 46.8 (C_4′H), 35.5 (N_1′-CH₃) ppm. Anal. Calcd for C₂₇H₂₂N₄O₂S: C, 69.51; H, 4.75; N, 12.01. Found: C, 69.32; H, 4.89; N, 12.04.

(2′S,3′R,4′R)-3′-(Benzo[d]thiazol-2-yl)-1′-methyl-2-oxo-4′-p-tolylspiro[indoline-3,2′-pyrrolidine]-3′-carbonitrile (5c)

Yield: 298 mg (53%) from 345 mg of 4c; pale yellow crystals, mp 132–133 °C. ¹H NMR (CDCl₃, 600 MHz): δ 8.37 (s, 1H, N₁-H), 7.99 (d, J = 7.2 Hz, 1H, C₄-H), 7.89 (d, J = 8.4 Hz, 1H, C_5′′′-H), 7.58 (d, J = 7.8 Hz, 1H, C_2′′′-H), 7.56 (d, J = 7.8 Hz, 2H, Ar–C_2″-H_p-tolyl), 7.32 (t, J = 7.8 Hz, 2H, C₆-H, C_4′′′-H), 7.26 (t, J = 7.8 Hz, 1H, C_3′′′-H), 7.20 (t, J = 7.8 Hz, 1H, C₅-H), 7.12 (d, J = 9.0 Hz, 2H, Ar–C_3″-H_p-tolyl), 6.73 (d, J = 7.8 Hz, 1H, C₇-H), 5.65 (dd, J = 9.6, 7.2 Hz, 1H, C_4′-H), 3.84–3.75 (m, 2H, C_5′-H), 2.27 (s, 3H, Ar–CH₃), 2.14 (s, 3H, N_1′-CH₃) ppm; ¹³C NMR (150 MHz, CDCl₃): δ 175.8 (q, C₂=O), 161.3 (q, C_7′′′), 152.0 (q, C_6′′′), 142.1 (q, C₈), 137.6 (Ar–C_4″), 135.3 (q, C_1′′′), 133.4 (q, C_1″), 130.7 (C₆H), 129.7 (2C_2″-H_p-tolyl), 129.0 (2C_3″-H_p-tolyl), 126.7 (C₄-H), 126.1 (C_4′′′-H), 125.7 (C_3′′′-H), 123.8 (C_5′′′-H), 123.5 (q, C₉), 123.2 (C₅-H), 121.3 (C_2′′′-H), 117.9 (CN), 110.0 (C₇H), 78.8 (C₃/C_2′), 62.0 (C_3′), 56.9 (C_5′H₂), 47.1 (C_4′H), 35.3 (N_1′-CH₃), 21.0 (ArCH₃) ppm. Anal. Calcd for C₂₇H₂₂N₄OS: C, 71.98; H, 4.92; N, 12.44. Found: C, 71.69; H, 5.07; N, 12.67.

(2′R,3′R,4′R)-3′-(Benzo[d]thiazol-2-yl)-1′-methyl-2-oxo-4′-p-tolylspiro[indoline-3,2′-pyrrolidine]-3′-carbonitrile (5′c)

Yield: 224 mg (39.8%) from 345 mg of 4c; yellow crystals, mp 125–126 °C. ¹H NMR (CDCl₃, 600 MHz): δ 9.21 (s, 1H, N₁-H), 8.09 (d, J = 8.4 Hz, 1H, C_5′′′-H), 7.67 (d, J = 8.4 Hz, 1H, C_2′′′-H), 7.53 (d, J = 7.8 Hz, 2H, Ar–C_2″-H_p-tolyl), 7.46 (t, J = 7.2 Hz, 1H, C_4′′′-H), 7.35 (t, J = 7.2 Hz, 1H, C_3′′′-H), 7.14 (d, J = 7.8 Hz, 2H, Ar–C_3″-H_p-tolyl), 7.03 (t, J = 7.8 Hz, 1H, C₆-H), 6.81 (d, J = 7.8 Hz, 1H, C₇-H), 6.65 (d, J = 7.8 Hz, 1H, C₄-H), 6.58 (t, J = 7.8 Hz, 1H, C₅-H), 5.42 (app t, J = 9.0 Hz, 1H, C_4′-H), 4.23 (app t, J = 9.6 Hz, 1H, C_5′-H_b), 3.60 (app t, J = 9.0 Hz, 1H, C_5′-H_a), 2.29 (s, 3H, Ar–CH₃), 2.27 (s, 3H, N_1′-CH₃) ppm; ¹³C NMR (150 MHz, CDCl₃): δ 176.6 (q, C₂=O), 166.1 (q, C_7′′′), 152.0 (q, C_6′′′), 141.4 (q, C₈), 137.9 (Ar–C_4″), 135.4 (q, C_1′′′), 133.0 (q, C_1″), 130.1 (C₆H), 129.6 (2C_2″-H_p-tolyl), 129.2 (2C_3″-H_p-tolyl), 126.3 (C₄-H), 126.1 (C_4′′′-H), 125.8 (C_3′′′-H), 123.7 (q, C₉), 123.6 (C_5′′′-H), 122.3 (C₅-H), 121.4 (C_2′′′-H), 118.2 (CN), 110.6 (C₇H), 80.1 (C₃/C_2′), 61.3 (C_3′), 57.5 (C_5′H₂), 51.5 (C_4′H), 35.5 (N_1′-CH₃), 21.1 (ArCH₃) ppm. Anal. Calcd for C₂₇H₂₂N₄OS: C, 71.98; H, 4.92; N, 12.44. Found: C, 71.60; H, 5.11; N, 12.71.

(2′S,3′R,4′R)-3′-(Benzo[d]thiazol-2-yl)-4′-(4-chlorophenyl)-1′-methyl-2-oxospiro[indoline-3,2′-pyrrolidine]-3′-carbonitrile (5d)

Yield: 310 mg (52.7%) from 371 mg of 4d; pale yellow crystals, mp 141–143 °C. ¹H NMR (CDCl₃, 600 MHz): δ 7.99 (d, J = 7.2 Hz, 1H, C₄-H), 7.97 (d, J = 8.4 Hz, 1H, C_5′′′-H), 7.87 (s, 1H, N₁-H), 7.65 (d, J = 8.4 Hz, 1H, C_2′′′-H), 7.63 (d, J = 8.4 Hz, 2H, Ar–C_2″-H_{p-chlorophenyl}), 7.40 (t, J = 7.8 Hz, 1H, C_4′′′-H), 7.36 (t, J = 7.8 Hz, 1H, C₆-H), 7.32 (t, J = 7.8 Hz, 1H, C_3′′′-H), 7.29 (d, J = 8.4 Hz, 2H, Ar–C_3″-H_{p-chlorophenyl}), 7.23 (t, J = 7.2 Hz, 1H, C₅-H), 6.72 (d, J = 7.8 Hz, 1H, C₇-H), 5.72 (dd, J = 10.8, 7.2 Hz, 1H, C_4′-H), 3.85 (app t, J = 10.2 Hz, 1H, C_5′-H_a), 3.78 (dd, J = 9.6, 7.2 Hz, 1H, C_5′-H_b), 2.19 (s, 3H, N_1′-CH₃) ppm; ¹³C NMR (150 MHz, CDCl₃): δ 175.2 (q, C₂=O), 161.0 (q, C_7′′′), 152.0 (q, C_6′′′), 142.0 (q, C₈), 135.5 (q, C_1″), 135.2 (q, C_1′′′), 133.9 (ArC_4″-Cl), 131.3 (2C_2″-H_{p-chlorophenyl}), 130.9 (C₆H), 128.5 (2C_3″-H_{p-chlorophenyl}), 126.8 (C₄-H), 126.3 (C_4′′′-H), 125.9 (C_3′′′-H), 124.0 (C_5′′′-H), 123.3 (C₅-H), 123.2 (q, C₉), 121.4 (C_2′′′-H), 117.7 (CN), 109.9 (C₇H), 78.7 (C₃/C_2′), 61.9 (C_3′), 56.7 (C_5′H₂), 46.5 (C_4′H), 35.4 (N_1′-CH₃) ppm. Anal. Calcd for C₂₆H₁₉ClN₄OS: C, 66.31; H, 4.07; N, 11.90. Found: C, 66.59; H, 3.88; N, 11.77.

(2′S,3′R,4′R)-3′-(Benzo[d]thiazol-2-yl)-4′-(4-fluorophenyl)-1′-methyl-2-oxospiro[indoline-3,2′-pyrrolidine]-3′-carbonitrile (5e)

Yield: 548 mg (96.5%) from 350 mg of 4e; pale yellow cystals, mp 183–185 °C. ¹H NMR (CDCl₃, 600 MHz): δ 8.09 (s, 1H, N₁-H), 7.99 (d, J = 7.2 Hz, 1H, C₄-H), 7.95 (d, J = 8.4 Hz, 1H, C_5′′′-H), 7.66 (dd, J = 8.4, 5.4 Hz, 2H, Ar–C_2″-H_{p-fluorophenyl}), 7.63 (d, J = 7.8 Hz, 1H, C_2′′′-H), 7.37 (t, J = 8.4 Hz, 1H, C_4′′′-H), 7.35 (t, J = 8.4 Hz, 1H, C₆-H), 7.30 (t, J = 7.2 Hz, 1H, C_3′′′-H), 7.22 (t, J = 7.8 Hz, 1H, C₅-H), 7.00 (d, J = 8.4 Hz, 2H, Ar–C_3″-H_{p-fluorophenyl}), 6.74 (d, J = 7.8 Hz, 1H, C₇-H), 5.72 (dd, J = 10.2, 6.6 Hz, 1H, C_4′-H), 3.84 (app t, J = 10.2 Hz, 1H, C_5′-H_a), 3.78 (dd, J = 9.6, 7.2 Hz, 1H, C_5′-H_b), 2.17 (s, 3H, N_1′-CH₃) ppm; ¹³C NMR (150 MHz, CDCl₃): δ 175.4 (q, C₂=O), 162.4 (d, J = 244.5 Hz, ArC_4″-F), 161.0 (q, C_7′′′), 152.0 (q, C_6′′′), 142.0 (q, C₈), 135.4 (q, C_1′′′), 132.3 (d, J = 3.2 Hz, C_1″), 131.6 (d, J = 8.1 Hz, 2C_2″-H_{p-fluorophenyl}), 130.8 (C₆-H), 126.8 (C₄-H), 126.2 (C_4′′′-H), 125.9 (C_3′′′-H), 123.9 (C_5′′′-H), 123.3 (C₅-H, C₉), 121.4 (C_2′′′-H), 117.8 (CN), 115.2 (d, J = 21.3 Hz, 2C_3″-H_{p-fluorophenyl}), 110.0 (C₇H), 78.7 (C₃/C_2′), 62.0 (C_3′), 56.9 (C_5′H₂), 46.5 (C_4′-H), 35.4 (N_1′-CH₃) ppm. Anal. Calcd for C₂₆H₁₉FN₄OS: C, 68.71; H, 4.21; N, 12.33. Found: C, 68.50; H, 4.35; N, 12.16.

(2′S,3′R,4′R)-3′-(Benzo[d]thiazol-2-yl)-4′-(4-cyanophenyl)-1′-methyl-2-oxospiro[indoline-3,2′-pyrrolidine]-3′-carbonitrile (5f)

Yield: 380 mg (65.9%) from 359 mg of 4f; white cystals, mp 159–161 °C. ¹H NMR (CDCl₃, 600 MHz): δ 8.03 (d, J = 8.4 Hz, 1H, C_5′′′-H), 7.99 (d, J = 7.2 Hz, 1H, C₄-H), 7.84 (d, J = 8.4 Hz, 2H, Ar–C_2″-H_{p-cyanophenyl}), 7.68 (d, J = 7.8 Hz, 1H, C_2′′′-H), 7.63 (d, J = 8.4 Hz, 2H, Ar–C_3″-H_{p-cyanophenyl}), 7.53 (s, 1H, N₁-H), 7.45 (t, J = 7.2 Hz, 1H, C_4′′′-H), 7.38 (t, J = 7.8 Hz, 1H, C₆-H), 7.36 (t, J = 7.2 Hz, 1H, C_3′′′-H), 7.25 (t, J = 7.8 Hz, 1H, C₅-H), 6.73 (d, J = 7.8 Hz, 1H, C₇-H), 5.85 (dd, J = 10.8, 6.6 Hz, 1H, C_4′-H), 3.91 (app t, J = 10.2 Hz, 1H, C_5′-H_a), 3.78 (dd, J = 9.6, 6.6 Hz, 1H, C_5′-H_b), 2.25 (s, 3H, N_1′-CH₃) ppm; ¹³C NMR (150 MHz, CDCl₃): δ 174.7 (q, C₂=O), 160.6 (q, C_7′′′), 151.9 (q, C_6′′′), 142.5 (q, C_1″), 142.0 (q, C₈), 135.5 (q, C_1′′′), 132.1 (2C_3″-H_{p-cyanophenyl}), 131.1 (C₆H), 130.8 (2C_2″-H_{p-cyanophenyl}), 126.9 (C₄-H), 126.5 (C_4′′′-H), 126.1 (C_3′′′-H), 124.1 (C_5′′′-H), 123.5 (C₅-H), 122.9 (q, C₉), 121.5 (C_2′′′-H), 118.8 (Ar–C_5″N), 117.5 (CN), 111.8 (ArC_4″-CN), 109.9 (C₇H), 78.6 (C₃/C_2′), 61.6 (C_3′), 56.5 (C_5′H₂), 46.7 (C_4′H), 35.5 (N_1′-CH₃) ppm. Anal. Calcd for C₂₇H₁₉N₅OS: C, 70.26; H, 4.15; N, 15.17. Found: C, 70.59; H, 4.45; N, 15.01.

(2′R,3′R,4′R)-3′-(Benzo[d]thiazol-2-yl)-4′-(4-cyanophenyl)-1′-methyl-2-oxospiro[indoline-3,2′-pyrrolidine]-3′-carbonitrile (5′f)

Yield: 175 mg (30.3%) from 359 mg of 4f; pale yellow crystals, mp 150–152 °C. ¹H NMR (CDCl₃, 600 MHz): δ 8.75 (s, 1H, N₁-H), 8.14 (d, J = 7.8 Hz, 1H, C_5′′′-H), 7.78 (d, J = 8.4 Hz, 2H, Ar–C_2″-H_{p-cyanophenyl}), 7.72 (d, J = 8.4 Hz, 1H, C_2′′′-H), 7.64 (d, J = 8.4 Hz, 2H, Ar–C_3″-H_{p-cyanophenyl}), 7.53 (t, J = 7.8 Hz, 1H, C_4′′′-H), 7.40 (t, J = 8.4 Hz, 1H, C_3′′′-H), 7.09 (t, J = 7.2 Hz, 1H, C₆-H), 6.83 (d, J = 7.8 Hz, 1H, C₇-H), 6.57 (t, J = 7.8 Hz, 1H, C₅-H), 6.52 (d, J = 7.8 Hz, 1H, C₄-H), 5.56 (app t, J = 8.4 Hz, 1H, C_4′-H), 4.25 (app t, J = 9.6 Hz, 1H, C_5′-H_b), 3.68 (app t, J = 9.0 Hz, 1H, C_5′-H_a), 2.32 (s, 3H, N_1′-CH₃) ppm; ¹³C NMR (150 MHz, CDCl₃): δ 176.3 (q, C₂=O), 165.8 (q, C_7′′′), 151.9 (q, C_6′′′), 141.8 (q, C_1″), 141.3 (q, C₈), 135.5 (q, C_1′′′), 132.3 (2C_3″-H_{p-cyanophenyl}), 130.7 (2C_2″-H_{p-cyanophenyl}), 130.4 (C₆-H), 126.7 (C_4′′′-H), 126.1 (C_3′′′-H), 126.0 (C₄-H), 123.8 (C_5′′′-H), 123.4 (q, C₉), 122.7 (C₅-H), 121.6 (C_2′′′-H), 118.6 (Ar–C_5″N), 118.0 (CN), 112.1 (ArC_4″-CN), 110.6 (C₇-H), 80.1 (C₃/C_2′), 60.9 (C_3′), 57.2 (C_5′H₂), 51.4 (C_4′H), 35.5 (N_1′-CH₃) ppm. Anal. Calcd for C₂₇H₁₉N₅OS: C, 70.26; H, 4.15; N, 15.17. Found: C, 70.43; H, 4.47; N, 15.09.

(2′S,3′R,4′R)-3′-(Benzo[d]thiazol-2-yl)-1′-methyl-4′-(4-nitrophenyl)-2-oxospiro[indoline-3,2′-pyrrolidine]-3′-carbonitrile (5g)

Yield: 460 mg (76.4%) from 384 mg of 4g; orange crystals, mp 195–197 °C. ¹H NMR (CDCl₃, 600 MHz): δ 8.19 (d, J = 9.0 Hz, 2H, Ar–C_3″-H_{p-nitrophenyl}), 8.06 (d, J = 7.8 Hz, 1H, C_5′′′-H), 8.00 (d, J = 7.8 Hz, 1H, C₄-H), 7.91 (d, J = 9.0 Hz, 2H, Ar–C_2″-H_{p-nitrophenyl}), 7.69 (d, J = 7.8 Hz, 1H, C_2′′′-H), 7.47 (t, J = 7.8 Hz, 1H, C_4′′′-H), 7.40 (t, J = 7.8 Hz, 1H, C₆-H), 7.38 (s, 1H, N₁-H), 7.37 (t, J = 7.8 Hz, 1H, C_3′′′-H), 7.26 (t, J = 7.5 Hz, 1H, C₅-H), 6.73 (d, J = 7.8 Hz, 1H, C₇-H), 5.93 (dd, J = 10.2, 6.6 Hz, 1H, C_4′-H), 3.94 (app t, J = 10.2 Hz, 1H, C_5′-H_a), 3.82 (dd, J = 9.6, 6.6 Hz, 1H, C_5′-H_b), 2.27 (s, 3H, N_1′-CH₃) ppm; ¹³C NMR (150 MHz, CDCl₃): δ 174.5 (q, C₂=O), 160.5 (q, C_7′′′), 151.9 (q, C_6′′′), 147.6 (C_4″-NO₂), 144.6 (q, C_1″), 141.9 (q, C₈), 135.6 (q, C_1′′′), 131.1 (C₆-H), 131.0 (2C_2″H_{p-nitrophenyl}), 127.0 (C_4-H), 126.5 (C_4′′′-H), 126.2 (C_3′′′-H), 124.1 (C_5′′′-H), 123.5 (C₅-H), 123.4 (2C_3″H_{p-nitrophenyl}), 122.8 (q, C₉), 121.5 (C_2′′′-H), 117.5 (CN), 109.9 (C₇-H), 78.6 (C₃/C_2′), 61.7 (C_3′), 56.6 (C_5′H₂), 46.4 (C_4′-H), 35.5 (N_1′-CH₃) ppm. Anal. Calcd for C₂₆H₁₉N₅O₃S: C, 64.85; H, 3.98; N, 14.54. Found: C, 65.09; H, 4.17; N, 14.43.

(2′R,3′R,4′R)-3′-(Benzo[d]thiazol-2-yl)-1′-methyl-4′-(4-nitrophenyl)-2-oxospiro[indoline-3,2′-pyrrolidine]-3′-carbonitrile (5′g)

Yield: 60 mg (10%) from 384 mg of 4g; pale brown crystals, mp 185–187 °C. ¹H NMR (CDCl₃, 600 MHz): δ 8.21 (d, J = 8.8 Hz, 2H, Ar–C_3″-H_{p-nitrophenyl}), 8.17 (d, J = 7.8 Hz, 1H, C_5′′′-H), 8.07 (s, 1H, N₁-H), 7.82 (d, J = 8.4 Hz, 2H, Ar–C_2″-H_{p-nitrophenyl}), 7.74 (d, J = 7.8 Hz, 1H, C_2′′′-H), 7.55 (t, J = 8.4 Hz, 1H, C_4′′′-H), 7.42 (t, J = 8.4 Hz, 1H, C_3′′′-H), 7.10 (t, J = 7.8 Hz, 1H, C₆-H), 6.81 (d, J = 7.8 Hz, 1H, C₇-H), 6.55 (t, J = 7.8 Hz, 1H, C₅-H), 6.45 (d, J = 7.8 Hz, 1H, C₄-H), 5.60 (dd, J = 9.6, 8.4 Hz, 1H, C_4′-H), 4.28 (app t, J = 10.2 Hz, 1H, C_5′-H_b), 3.70 (dd, J = 9.6, 8.4 Hz, 1H, C_5′-H_a), 2.32 (s, 3H, N_1′-CH₃) ppm; ¹³C NMR (150 MHz, CDCl₃): δ 176.4 (q, C₂=O), 165.7 (q, C_7′′′), 151.8 (q, C_6′′′), 147.7 (C_4″-NO₂), 143.9 (q, C_1″), 141.4 (q, C₈), 135.5 (q, C_1′′′), 130.9 (2C_2″H_{p-nitrophenyl}), 130.5 (C₆-H), 126.7 (C_4′′′-H), 126.1 (C_3′′′-H), 126.0 (C_4-H), 123.8 (C_5′′′-H), 123.7 (C₅-H), 123.3 (q, C₉), 122.7 (2C_3″H_{p-nitrophenyl}), 121.6 (C_2′′′-H), 118.0 (CN), 110.7 (C₇-H), 80.1 (C₃/C_2′), 60.8 (C_3′), 57.3 (C_5′H₂), 51.1 (C_4′-H), 35.4 (N_1′-CH₃) ppm. Anal. Calcd for C₂₆H₁₉N₅O₃S: C, 64.85; H, 3.98; N, 14.54. Found: C, 65.03; H, 4.27; N, 14.39.

Data Availability Statement

The data underlying this study are available in the published article and its Supporting Information.

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsomega.4c01552.

• Copies of ¹H, ¹³C{¹H}, ¹⁹F, ¹³C-DEPT 90/135, ¹H–¹H-COSY, ¹H–¹³C-HSQC, and ¹H–¹³C-HMBC NMR spectra for compounds 5a–c, 5′c, 5d–f, 5′f, 5g, 5′g, regio- and stereochemical assignement of exo-cycloadduct (5g), regio- and stereochemical assignment of endo-cycloadduct (5′g), optimized ground state energies of E-configurations and Z-configurations of dipole 3 and dipolarophiles 4a–g, HOMO and LUMO orbitals at 5a and 6a transitions states, optimized coordinates for 5a, 5(TS-I), 5′a, 5(TS-II), 6a, 6(TS-I), 6′a, 6(TS-II) (PDF)

The authors declare no competing financial interest.