Abstract

Heterocyclic compounds are vital in rational drug design by providing convenient means for the optimization of drugs or drug candidates. Isoxazolidine (a heterocyclic compound) derived from the (3 + 2) cycloaddition reaction is reported to possess anticancer, antiviral, antibacterial, and anti-inflammatory properties. The mechanistic study of the chemo-, regio-, and stereo-selectivities of the (3 + 2) cycloaddition reaction (32CA) of diarylnitrones (A1) to 1-(4-nitrophenyl)-5H-pyrrolin-2-one (A2) has been conducted using the density functional theory (DFT) method at the M06/6-311G (d, ) computational level. The 32CA reaction of A1 and A2 proceeds through a chemo-selective addition of A1 across the C=C olefinic bond of A2 to furnish both kinetically and thermodynamically favored reaction routes. The effect of substitution on the 32CA reaction has been studied using a representative set of electron-acceptor and electron-releasing groups on both A1 and A2. The chemo-, regio-, and stereo-selectivities observed in the 32CA reaction remained unchanged irrespective of the substituents used on both reactants, but a change in kinetics and thermodynamics was observed based on the substituents. In the 32CA reaction of A1 and A2, the “exo” cycloadduct formation was favored over the “endo” cycloadducts in all instances.

1. Introduction

Metal-mediated cycloadditions became unpopular compared to their metal-free processes because metals serve as a potential toxic waste source. However, metal-involving syntheses (particularly metal-catalysed reaction) are still preferred over relevant metal-free routes because metal centres could strongly activate reactants and/or substantially reduce the number of steps leading to target compounds [16].

For example, isoxazolidines (a heterocyclic compound) can be generated from the (3 + 2) cycloaddition (32CA) reaction, which involves the coupling of a three-atom component (TAC) and an ethylene derivative [7]. The 32CA reaction of nitrones to ethylene derivatives leads to the generation of the isoxazolidine heterocyclic compounds, which possess anticancer, antiviral, antibacterial, and anti-inflammatory properties [8]. 5H-pyrrolin-2-ones, also known as 3-pyrrolin-2-ones, are nitrogen-containing heterocycles that can be derived from the oxidation of pyrroles [9]. Due to the reactive nature of this moiety, it is capable of undergoing reactions such as epoxidation [10, 11], cyclopropanation [12], and cycloaddition [13] reactions.

Langlois et al. [14] reported initial findings for the 32CA reaction of 5H-pyrrolin-2-ones with nitrones and stated that the reaction gave a mixture of stereoisomers and regioisomers which were separable by chromatography.

Muzychenko and coworkers [15] examined the 32CA reaction of diarylnitrone (A1) with 1-(4-nitrophenyl)-5H-pyrrolin-2-one (A2) to produce 2-oxa-6-oxo-3-phenyl-4-(R-phenyl)-7-(4-nitrophenyl)-3,7-diazabicyclo (3,3.0) octane 1 as the major product (Scheme 1). It was reported that A1 chemo-selectively added across the C=C olefinic bond of A2 over the C=O functionality. Also, in the addition of A1 across the C=C olefinic bond, it added regio-selectively across A2 in an “exo” manner cis to the N-nitrophenyl group of A2 to give a single isomeric product. Muzychenko and coworkers [15] reported that forming an “endo” approach is also plausible to be isolated in the reaction. From their experiment, it was concluded that different substituents on the nitrone gave different reaction times and product yields.

In this study, the regio-, chemo-, and stereo-selectivities and reaction mechanism of the 32CA of diarylnitrone derivatives (A1) with 1-(4-nitrophenyl)-5H-pyrrolin-2-one(A2) are elucidated. Path A observed experimentally by Muzychenko and coworkers [15] and plausible path B were considered in this study. We herein seek to determine the effect of electron-releasing groups (ERGs) and electron-acceptor groups (EAGs) by employing a wide range of substituents based on how strong, weak, or moderate the substituents are. Also, the effect of small and bulky groups on the reaction was explored. Understanding the factors controlling the reaction will help enhance chemical insights into the reaction. The study is based on Scheme 1 to depict the various regio- and stereo-selectivities involved in the reaction. The computed selectivities based on activation energies trends are then rationalized using local and global reactivity indices based on the conceptual DFT.

2. Computational Details and Methodology

Here, we present only a brief statement of the method because a fuller description is available in [16, 17]. The M06 hybrid functional [18] as implemented in Gaussian 09 [19] has been employed together with the split valence triple-ξ (TZ) basis set 6-311G (d, ) [2022] in this study.

3. Results and Discussion

From Scheme 1, the mechanism of the 32CA reaction of diarylnitrone derivatives (A1) with 1-(4-nitrophenyl)-5H-pyrrolin-2-one (A2) is proposed to go through two different reaction pathways. 1-(4-nitrophenyl)-5H-pyrrolin-2-one (A2) has multiple reactive centres; therefore, diarylnitrone can add across its C=C olefinic bond in a (3 + 2) manner and similarly across the C=O bond. The zero-point corrected absolute energies and coordinates of all the computed structures reported herein are shown in Tables S1S7 in the Supplementary file. The reaction path A from Scheme 1 emerges from the addition of A1 across the C=C double bond of A2 through transition states TS1A-EXO, TS2A-ENDO, TS3A-EXO, and TS4A-ENDO to furnish products P1A-EXO, P2A-ENDO, P3A-EXO, and P4A-ENDO, respectively. P1A-EXO and P3A-EXO result from the “exo” approach of A1 with cis-stereospecific addition to the 1-(4-nitrophenyl)-5H-pyrrolin-2-one C=C olefinic bond. Products P1A-EXO and P2A-ENDO are stereoisomeric pairs, likewise P3A-EXO and P4A-ENDO. Also, the pairs P1A-EXO and P3A-EXO and P2A-ENDO and P4A-ENDO are regioisomers.

Path B originates from the addition of A1 across the C=O double bond of A2 to afford P1B and P2B through their respective transition states, TS1B and TS2B. Product pairs P1B and P2B are regioisomers.

3.1. Analysis of the 32CA Parent Reaction of 1-(4-Nitrophenyl)-5H-Pyrrolin-2-One (A2) and Diarylnitrone (A1, R=H)

This section provides mechanistic insights into the 32CA reaction of the diarylnitrone (A1, R=H) with 1-(4-nitrophenyl)-5H-pyrrolin-2-one (A2). The reported imaginary frequencies of the optimized transition states in the reaction of 1-(4-nitrophenyl)-5H-pyrrolin-2-one (A2) with diarylnitrone (A1, R=H) in gas phase at 25°C are presented in Tables S8. Figure 1 represents the Gibbs free energy profile for the 32CA reaction of A2 and A1 (R=H) in kcal/mol. Table 1 presents the calculated rate constants for the 32CA reaction of A2 and A1 (R=H) to furnish the cycloadducts. From Figure 1, the most preferred pathway is path A, which is associated with the chemo-selective addition of A1 (R=H) across the C=C olefinic bond of A2 to furnish the cycloadduct. Within path A, the formation of the “exo” product P1A-EXO (reaction energy of −22.2 kcal/mol) through TS1A-EXO (activation barrier of 4.5 kcal/mol) is the most kinetically favored reaction route with P1A-EXO being thermodynamically stable. The nearest competing pathway is the formation of P4A-ENDO (reaction energy of −20.3 kcal/mol) via the transition state TS4A-ENDO (activation barrier of 7.6 kcal/mol), where TS1A-EXO is kinetically favored by a difference of 3.1 kcal/mol. The cycloadduct P1A-EXO is irreversible since the reaction is highly exergonic due to its reaction energy. The lower activation barrier observed for TS1A-EXO and the thermodynamic stability of P1A-EXO correspond to the higher yield observed in the experiment. The endergonic cycloadducts found along path B, as a result of the addition of A1 across the C=O functionality of the A2 molecule via the transition states TS1B and TS2B with activation barriers of 13.0 kcal/mol and 51.5 kcal/mol, respectively, more than the most favorable transition state TS1A-EXO, were not isolated experimentally. The favored reaction route established in this study is in complete agreement with the experimental observation reported by Muzychenko et al. [15]. This indicates that the formation of P1A-EXO was a function of the step leading to it being kinetically favored over all other possible reaction channels or stages.

Table 1 shows the calculated rate constants for the formation of the isomeric cycloadducts considered in Scheme 2. As seen in Table 1, P1A-EXO has the highest rate constant of 3.1 × 109 s−1; this accounts for the greater yield observed experimentally. The closest competing pathway has a calculated rate constant of 1.7 × 107 s−1, making the formation of P1A-EXO about 182 times faster than the formation of P4A-ENDO, again confirming why the product P1A-EXO was isolated in a higher yield by Muzychenko et al. [15].

3.2. Effect of Substituents on the 32CA Reaction of 1-(4-Nitrophenyl)-5H-Pyrrolin-2-One (A2) and Diarylnitrone (A1, R = ERGs, and EAGs)

We report on the effect of electron-releasing groups (ERGs) and electron-acceptor groups (EAGs) at the meta and para positions of the aryl group of the A1 molecule on the reaction. The activation and reaction energies of the transition states and products in the 32CA reaction of A1 (R = EAGs and ERGs) are displayed in Tables 2 and 3, respectively. Table 2 shows the substitution of the illustrative set of ERGs and EAGs at the para position, whereas Table 3 shows substitution at the meta position. It ought to be stated that, after the exhaustive search to locate certain transition states, it proved futile. However, we believe it will not substantially affect the pattern we see based on the accumulated data. Tables 2 and 3 show that both ERGs and EAGs on A1 favors the formation of the product where A1 adds across the C=C olefin bond of the A2. The formation of P1A-EXO through TS1A-EXO is seen to be the most kinetically preferred pathway and the product is thermodynamically stable in all instances when both ERGs and EAGs are placed at both para and meta positions of A1. Due to the higher activation barriers of the transition states (TS1B and TS2B) and the thermodynamically unstable products (P1B and P2B) that are provided as seen in path B, the addition of A1 across the C=O olefinic bond of A2 can be ruled out. Hence, no substituents effect was explored for path B.

Table 2 shows that CH3 and (CH3)2N on the para position of A1 reduced the activation barriers by a margin of about 1.4 and 0.6 kcal/mol, respectively. Also, the products formed are thermodynamically stable. The substituents OCH3 and NO2 at the para positions of A1 tend to increase the activation barriers by 4.7 and 0.6 kcal/mol, respectively, with the formation of stable products. However, when Br is placed in the para position, all the activation energy increases, but that of TS1A-EXO, which leads to P1A-EXO, reduces the activation barrier by 0.3 kcal/mol. Similar trends were observed for the substitution at the meta position of A1 with the EAGs and ERGs, as shown in Table 3. The negative activation barrier obtained for TS1A-EXO (R=(CH3)2N) can be attributed to the inability of the level of the theory to describe the system very well, leading to the underestimation of the activation energy.

3.3. Effect of the Substituents on the Ethylene Derivative (A2) in the 32CA Reaction of 1-(4-Nitrophenyl)-5H-Pyrrolin-2-One (A2) and Diarylnitrone

This section explores the effect of the substituents on the 1-(4-nitrophenyl)-5H-pyrrolin-2-one. Herein, the nitrophenyl group of A2 was replaced by H, CH3, and COOEt. This study investigated how various small and bulky groups on A2 also affect the reaction. Scheme 3 shows an updated scheme for the reaction studied. Table 4 shows the activation and reaction energies of the transition states and products of the 32CA reaction of A1(R=H) and A2 (X=H, CH3, and COOEt) explored.

As seen in Table 4, the formation of the “exo” product P1A-EXO through TS1A-EXO is still the favored reaction route. Activation barriers increased to 4.1 kcal/mol when a smaller group (A2, X=H) was used to form stable products. When X=CH3, the activation barrier was reduced to 1.2 kcal/mol relative to when X=H. In the 32CA reaction of A2 (X=COOEt) and A1 (R=H), a bulkier group, on the other hand, reduced the activation barriers to about 0.1 kcal/mol of which the values obtained are close to the parent reaction. Therefore, bulkier groups on A2 reduce the activation barrier of the transition state, which leads to the formation of the favored product.

3.4. Analysis of the 32CA Reaction of 1-(4-Nitrophenyl)-5H-Pyrrolin-2-One (A2) Derivatives and Diarylnitrone Derivatives with the Global Reactivity Indices

This section discusses the inherent reactivity and selectivity of the 32CA reaction of the diarylnitrone derivatives and 1-(4-nitrophenyl)-5H-pyrrolin-2-one derivatives. The computed global reactivity indices for the reaction are displayed in Tables 5 and 6. The electronic chemical potential (μ), chemical hardness (η), electrophilicity (ω), and the maximum charge transfer (ΔNmax) [23] results are displayed in Tables 5 and 6. From Table 5 (para), the ω values of TAC derivatives are also in the order (CH3)2N < CH3 < OCH3 < Br < NO2.

Table 5 shows the nucleophilic indices (N) [23, 24] designating the nucleophilicity ability of the various reactants. It is evident that A1 (R=NO2) has the lowest N value of 2.66 eV, making A1 (R=NO2) the poorest nucleophile, while A1 (R=(CH3)2N) has the highest N value of 4.12 eV, making A1 (R=(CH3)2N) the best nucleophile.

From Table 6 (meta), the ω values of TAC derivatives are also in the order (CH3)2N) < OCH3 < CH3 < Br < NO2. ERGs on the meta position of the diarylnitrone reduce the ω values, implying why A1 acts as a nucleophilic specie. Table 6 also displays the computed nucleophilic indices (N) indicating the nucleophilic abilities of the various reactants. It is observed that A1 (R=NO2) has the lowest nucleophilicity value of 2.32 eV, making A1 (R=NO2) the poorest nucleophile, whereas A1 (R=(CH3)2N) has the highest N value of 4.11 eV, making A1 (R=(CH3)2N) the best nucleophile.

4. Conclusion

The diarylnitrone moiety (A1) adds chemo- and regio-selectively to the C=C olefinic bond over the C=O olefinic bond of 1-(4-nitrophenyl)-5H-pyrrolin-2-one (A2) to form a stereoselective cycloadduct, P1A-EXO, which is shown from the results obtained for this study. The reaction, therefore, proceeds through the proposed path A; thus, path A is kinetically favored over path B. The most kinetically favored reaction channel is the formation of P1A-EXO through TS1A-EXO in support of the findings made by Muzychenko et al. Also, the formation of P2A-ENDO (−20.4 kcal/mol), P3A-EXO (−19.7 kcal/mol), and P4A-ENDO (−20.3 kcal/mol) are possible since their formation energies are thermodynamically stable and the activation barriers for the transition states, TS2A-ENDO (10.6 kcal/mol), TS3A-EXO (9.6 kcal/mol), and TS4A-ENDO (7.6 kcal/mol), are appreciably low; this goes on to favor the report of an “endo” approach of A1 being possible by Muzychenko et al. However, P1A-EXO was isolated as a major product since the rate constant for the formation of the preferred product P1A-EXO in the 32CA of diarylnitrone A1 and 1-(4-nitrophenyl)-5H-pyrrolin-2-one A2 is 3.1 × 109 s−1, which is about 182 times faster than the next competing pathway with a rate constant of 1.7 × 107 s−1. Irrespective of the electronic nature of the substituents on the para and meta positions of A1, the chemo-, regio-, and stereo-selectivities remained unchanged. Smaller and bulkier groups on A2 maintain the trends of the chemo-, regio-, and stereo-selectivities of the reaction. However, the bulky groups tend to reduce the activation energies of the transition states that lead to the formation of the products. Future research should explore solvent effects on the mechanism of the (3 + 2) cycloaddition reaction of 1-(4-nitrophenyl)-5H-pyrrolin-2-one with other dipoles such as azides and nitrile oxides that may provide novel reaction routes to the formation of azoles and isoxazolines which possess antifungal, anti-inflammatory, and antibacterial properties.

Data Availability

The data used to support the findings of this study are available on request from the corresponding author.

Conflicts of Interest

The authors declare that there are no conflicts of interest.

Authors’ Contributions

SA, JAK, and CHB conceived the idea and planned the study. SA ran computations and wrote the manuscript. JAK, RT, EA, and AA analyzed data. All the authors approved the manuscript before submission.

Acknowledgments

The authors are very grateful to the National Council for Tertiary Education, Republic of Ghana, for a research grant under the Teaching and Learning Innovation Fund (TALIF/KNUST/3/0008/2005) and South Africa’s Centre for High-Performance Computing for access to additional computing resources on the lengau cluster.

Supplementary Materials

The zero-point corrected absolute energies and coordinates of all the computed structures reported herein are shown in Tables S1–S7 in the Supplementary file. The reported imaginary frequencies of the optimized transition states in the reaction of 1-(4-nitrophenyl)-5H-pyrrolin-2-one (A2) with diarylnitrone (A1, R=H) at the M06/6-311G (d, ) level of the theory in the gas phase at 25°C are presented in Tables S8. (Supplementary Materials)