Density functional theory calculations on the cyclization of di-t-butyl 2-(2-aminophenyl)-2-methyl malonate (1) to t-butyl 3-methyloxindole-3-carboxylate (2) reveal that acetic acid-assisted protonation of the carbonyl oxygen atom reduces the activation Gibbs free energy significantly lower than methanol-assisted pathways. Experimental data confirm that reaction concentration plays a pivotal role in oxindole formation. Experimental results also indicate distinct reaction mechanisms at low and high concentrations. Achieving high enantioselectivity for chiral compound 2 in high-concentration reactions requires discovering a novel chiral acid.

The construction of a quaternary carbon atom at the 3-position of oxindole is a valuable target in organic synthesis.¹⁾ Many 3,3-disubstituted oxindoles serve as important intermediates in the synthesis of bioactive compounds, prompting the development of numerous synthetic methodologies for their preparation.^2,3) Among these, the tandem reaction of reduction-cyclization reaction of di-alkyl 2-nitrophenyl-2-alkyl malonates has emerged as an efficient and concise method for oxindole formation. In fact, we reported a quantitative Brønsted acid-catalyzed cyclization reaction of compound 3 to oxindole 2⁴⁾ (Fig. 1A).

Fig. 1. (A) Tandem Reaction of Reduction-Cyclization Reaction of Compound 3; (B) Brønsted Acid-Catalyzed Cyclization Reaction of Compound 1⁴⁾

Detailed experimental studies revealed that, despite the steric hindrance introduced by the bulky t-butyl groups, the anilinic nitrogen atom of compound 1, a relatively weak nucleophile, effectively attacked the carbonyl carbon atom of the t-butyl ester (Fig. 1B). Density functional theory (DFT) calculations supported this finding, suggesting that nucleophilic attack of the anilinic nitrogen on the activated carbonyl group of the t-butyl ester is a key plausible mechanism for oxindole formation.⁴⁾ Building on this work, we applied the reaction to a chiral Brønsted acid-catalyzed asymmetric cyclization reaction of compound 1 to oxindole 2a.⁵⁾ Although the addition of (S)-TRIP to the reaction mixture yielded chiral compound 2a, the enantioselectivity achieved was moderate (Fig. 2). These results underscore the need for a deeper understanding of the reaction mechanism to improve this asymmetric process.

Fig. 2. Asymmetric Oxindole Cyclization of Compound 1 Using (S)-TRIP, Chiral Brønsted Acid⁵⁾

In this study, we present a detailed mechanism of the oxindole cyclization reaction of compound 1, supported by DFT calculations and experimental observations. This pathway, which is concentration-dependent, provides new insights into the formation of oxindole 2 from compound 1.

Our previous DFT calculations indicated that the transition state TS1, involving the association of compound 1 with two MeOH molecules, could lower the activation Gibbs free energy via proton transfer among compound 1 and the two MeOH molecules⁴⁾ (pathway A, Fig. 3). To further explore the role of Brønsted acids, we examined an acetic acid (AcOH)-assisted transition state TS2 in DFT calculations (pathway B, Fig. 3). In the previous report, the proton transfer step (deprotonation of the amino group and protonation of the carbonyl oxygen atom) is a rate-determining step in the cyclization reaction.

Fig. 3. Hypothesized Pathway A through Transition State TS1 and Pathway B through the AcOH-Assisted Transition State TS2 for Oxindole Cyclization

To evaluate the difference between MeOH-mediated proton transfer and AcOH-mediated proton transfer, we performed DFT calculations to assess the contribution of pathways A and B to the cyclization reaction of compound 1 to oxindole 2. Activation energies were calculated using the Gaussian 16 software package.⁶⁾ The M06⁷⁾ functional was applied in combination with the 6-311 + G** basis set, and geometries of the ground-state molecules were fully optimized with the keyword Int = ultrafine. Transition-state geometries were confirmed through frequency analysis and intrinsic reaction coordinate calculations.

As calculated previously, the rate-determining step in these pathways is the nucleophilic addition of the anilinic nitrogen atom of compound 1 to the carbonyl carbon atom forming the intermediate IM (Fig. 3). This step is facilitated by protonation of the carbonyl oxygen atom and deprotonation from the amino group. Updated DFT calculations indicate that the intermolecular proton transfer among one molecule of compound 1 and two MeOH molecules (TS1 in Fig. 4, 34.81 kcal/mol) results in a relatively high activation energy, rendering this pathway insufficient to promote oxindole cyclization. Reducing the number of MeOH molecules significantly increased the Gibbs free energy for intermolecular proton transfer from the amino group to the carbonyl oxygen atom (TS3 in Fig. 4, 41.53 kcal/mol). The high Gibbs free energy of TS3 is nearly identical to that of TS4 (TS4 in Fig. 4, 42.74 kcal/mol). By contrast, intermolecular interaction between compound 1 and AcOH markedly decreased the Gibbs free energy (TS2 in Fig. 4, 21.51 kcal/mol). These results suggest that AcOH-mediated activation of the carbonyl group is a key factor in promoting oxindole cyclization.

Fig. 4. Transition States of the Cyclization Reaction from DFT Calculations

Second, we investigated the effects of the concentration of compound 1. We initially hypothesized that the reaction concentration might influence the oxindole cyclization reaction of compound 1. Observing that compound 1, when stored at room temperature, partially cyclized to compound 2, we speculated that intermolecular interactions among compound 1 molecules could facilitate the cyclization process. This hypothesis was based on our previous experience that the acyl transfer reaction is dependent on substrate concentration.⁸⁾ To directly observe the occurrence of the oxindole cyclization reaction, we monitored the reaction using ¹H-NMR analysis. Reaction mixtures were stirred in MeOH (for 7 d) or CH₂Cl₂ (for 3 d) at either room temperature or 60°C. The results are summarized in Table 1.

Table 1. Effects of Reaction Temperature and Concentration on the Oxindole Cyclization of Compound 1

Entry	Solvent	Concentration (M)	Temp. (°C)	Time (h)	Ratioa) 1:2
1	MeOH	0.3	rt	165	57:434)
2	MeOH	0.01	rt	165	99:1
3	MeOH	0.01	60	165	25:75
4	CH2Cl2	0.01	rt	74	N.D.
5	CH2Cl2	0.3	rt	74	95:54)

a)Ratio determined by ¹H-NMR analysis. N.D. The Peak of compound 2 was not detected.

We hypothesized that intermolecular proton transfer among compound 1 molecules alone would not occur at the low reaction concentration. As expected, lowering the reaction concentration from 0.3 to 0.01 M significantly decreased the production of cyclized compound 2 (entries 1 and 2). At 0.01 M, increasing the reaction temperature to 60°C slightly enhanced the reactivity (entry 3). To confirm the role of MeOH, we evaluated the conversion ratio between compounds 1 and 2 in the oxindole cyclization reaction in CH₂Cl₂ (entries 4 and 5). In CH₂Cl₂, the cyclization reaction did not proceed independently of the reaction concentration. These results indicate that the combination of proton transfer involving compound 1 molecules and MeOH molecules, along with concentration-dependent activation, promotes the oxindole cyclization reaction of compound 1.

Third, we tested the combination of various solvents with AcOH or Et₃N. The results are summarized in Table 2. The cyclization reaction produced a small amount of compound 2 in both polar and non-polar aprotic solvents at a concentration of 0.3 M in the absence of AcOH (entries 1‒4). We speculate that oxindole cyclization occurred due to condensation during solvent removal and silica gel column chromatography. However, the addition of AcOH significantly accelerated the reaction in tetrahydrofuran (THF), 1,2-dimethoxyethane (DME), toluene, and CH₃CN (entries 5‒8). In contrast, adding triethylamine (Et3N) to the reaction mixture in CH₂Cl₂ did not promote the production of compound 2 (entry 9). These results suggest that the key factor in accelerating the reaction is the protonation of the carbonyl oxygen atom, which activates the carbonyl group of the t-butyl ester. Conversely, deprotonation of the anilinic nitrogen atom by a Brønsted base is not effective in driving the reaction forward.

Table 2. Oxindole Cyclization of Compound 1 in Aprotic Solvents

Entry	Solvent	Additive	Yield (%)	SM recovery (%)
1	THF	None	12	84
2	DME		15	81
3	Toluene		16	78
4	CH3CN		12	82
5	THF	AcOH	92	8
6	DME		91	8
7	Toluene		99	N.D.
8	CH3CN		92	6
9	CH2Cl2	Et3N	<1a)

a)Determined by ¹H-NMR analysis. N.D. not detected.

Finally, we assessed the effect of the reaction concentration in the AcOH-assisted reaction mixture. We found that increasing the reaction concentration improved the conversion yield of compound 2 (entries 1‒4, Table 3). The reaction was smoothly completed within 24 h at a concentration of 0.3 M (entry 4). Interestingly, we previously reported that increasing the reaction concentration decreased the reaction time and enantioselectivity in the asymmetric cyclization reaction of compound 1 using (S)-TRIP (entries 2‒4, Table 4). These observations suggest that the reaction pathways differ depending on the reaction concentration, leading to moderate enantioselectivity at lower concentrations and low or racemic enantioselectivity at higher concentrations.

Table 3. Effects of Reaction Concentration on the AcOH-Assisted Oxindole Cyclization of Compound 1

Entry	Concentration (M)	Conversion yield (%)a)
1	0.01	58
2	0.03	80
3	0.1	93
4	0.3	100

a)Determined by ¹H-NMR analysis.

Table 4. Effects of the Concentration of Compound 1 and Catalyst Loading on the Oxindole Cyclization Reaction⁵⁾

Entry	Concentration (M)	Time	Isolated yield (%)	ee(%)a)
1	0.01	2h	87	49
2	0.03	1h	96	39
3	0.1	0.5h	95	34
4	0.3	10 min	99	21

a)enantiomeric excess.

In summary, we identified a possible mechanism for the oxindole cyclization reaction of compound 1 to produce oxindole 2, which is influenced by the reaction concentration of compound 1. DFT calculations revealed that the activation Gibbs free energy of TS2 is significantly lower than that of TS1, TS3, and TS4, indicating that the AcOH-mediated protonation of the carbonyl oxygen atom of compound 1 is a key factor in driving the cyclization reaction. By contrast, the effects of proton transfer among compound 1 molecules and MeOH molecules (TS1) are relatively limited. In fact, increasing the concentration of compound 1 to 0.3 M enhanced the reactivity; the contribution of concentration alone is restricted in the absence of AcOH. Notably, AcOH-assisted carbonyl activation plays a pivotal role in promoting the reaction, regardless of reaction concentration. Furthermore, our findings suggest that the reaction mechanism for Brønsted acid-assisted cyclization differs between low (0.01 M) and high (0.3 M) concentrations. These results underscore the importance of exploring alternative chiral acids that can achieve higher enantioselectivity, particularly under conditions of increased concentration, to improve the efficiency and enantioselectivity of the cyclization reaction of compound 1.

Part of this work was supported by the Kansai University Fund for the Promotion and Enhancement of Education and Research, 2023 “Drug discovery of compounds for the treatment of COVID-19,” Japan Science and Technology Agency (JST) Support for Pioneering Research Initiated by the Next Generation (SPRING) (Grant Number: JPMJSP2150), JSPS KAKENHI (Grant Numbers: 19K07008 and 23K06061), and AMED (Grant Numbers: JP23wm0325062 and JP24ek0109692).

The authors declare no conflict of interest.

This article contains supple-mentary materials.

• 1)  Cao Z. Y., Zhou F., Zhou J., Acc. Chem. Res., 51, 1443–1454 (2018).
• 2)  Nichinde C. B., Patil B. R., Chaudhari S. S., Mali B. P., Gonnande R. G., Kinage A. K., RSC Adv., 13, 13206–13212 (2023).
• 3)  Lv J., Ma M., Geng Z., Huang Y., Wei H., Yang R., Fang X., Bi Y., New J. Chem., 48, 18356–18359 (2024).
• 4)  Yamai Y., Tanaka A., Yajima T., Ishida K., Natsutani I., Uesato S., Nagaoka Y., Sumiyoshi T., Heterocycles, 97, 192–210 (2018).
• 5)  Ishida K., Shimizu M., Yamai Y., Natsutani I., Uesato S., Nagaoka Y., Sumiyoshi T., Heterocycles, 99, 1398–1411 (2019).
• 6)  Frisch M. J., Trucks G. W., Schlegel H. B., et al., Gaussian 16, Revision C. 01, Gaussian, Inc., Wallingford CT, 2016.
• 7)  Zhao Y., Truhlar D. G., Theor. Chem. Account., 120, 215–241 (2008).
• 8)  Uruno Y., Tanaka A., Hashimoto K., Usui S., Inoue Y., Konishi Y., Suwa A., Takai K., Katoda W., Fujiwara N., Sumiyoshi T., Tetrahedron, 69, 9675–9681 (2013).