N-heterocyclic carbene-catalyzed atroposelective synthesis of N-Aryl phthalimides and maleimides via activation of carboxylic acids

Abstract

Traditionally, N-aryl phthalimides are synthesized by the condensation of phthalic anhydride and aniline derivatives, usually proceeding under harsh conditions. The alternative mild and organocatalytic strategies for their synthesis are underdeveloped. Herein, we demonstrate the organocatalytic atroposelective synthesis of N-aryl phthalimides via the traditional N-C_C=O disconnection under mild conditions. The in-situ acid activation of phthalamic acid and subsequent N-heterocyclic carbene (NHC)-catalyzed atroposelective amidation allowed the synthesis of well-decorated N-aryl phthalimides in excellent yields and enantioselectivities. Mechanistic studies reveal the addition of NHC to the in situ generated isoimides, thus introducing a unique mode of generating acylazoliums. Interestingly, both enantiomers of the product can be accessed from the same phthalic anhydride and aniline using the same NHC pre-catalyst. Moreover, this strategy has been extended to the atroposelective synthesis of N-aryl maleimides.

N-aryl phthalimides are usually synthesized by the condensation of phthalic anhydride and aniline derivatives under harsh conditions. Herein, the authors report the N-heterocyclic carbene- catalyzed atroposelective synthesis of N-aryl phthalimides via the traditional N-C_C = O disconnection under mild conditions.

Introduction

N-Substituted phthalimides are ubiquitous in numerous bioactive compounds^1,2. For instance, thalidomide having phthalimide skeleton is used for the treatment of multiple myeloma³ and tuberculosis⁴, while apremilast is useful for the treatment of psoriatic arthritis⁵. Moreover, N-substituted phthalimide derivatives are widely applicable as catalysts^6,7, dyes⁸, and for the synthesis of polymers⁹. Given the diverse applications of phthalimides, direct and mild synthesis of such molecules is highly desirable.

Among the various strategies known for the construction of N-substituted phthalimides^10,11, the most common approach involves the condensation of phthalic anhydride and primary amine at elevated temperature or in the presence of acidic dehydrating agents (Fig. 1A)^12–14. Moreover, a variety of transition metal-catalyzed synthesis of functionalized N-substituted phthalimides are known, such as the carbonylative cyclization of aromatic amides^15–17 or 1,2-dihaloarenes¹⁸ in the presence of CO at high pressure. N-substituted phthalimides can also be synthesized from diols via Ru catalysis¹⁹. or the cyclization of isocyanates with benzoic acid/amide derivatives^20,21. However, all these methods require pre-functionalized starting materials, high temperatures, and/or expensive metal catalysts to furnish the phthalimides. Additionally, the atroposelective synthesis of N-aryl phthalimides via Pd-catalyzed carbonylation of aryl iodide with CO was disclosed by Li group²². Notably, very few organocatalytic routes for the synthesis of N-aryl phthalimides have been realized. The benzannulation strategy via oxidative [4 + 2] annulation of α,β-unsaturated aldehydes with N-substituted maleimides for the synthesis of achiral²³ and atroposelective²⁴ N-substituted phthalimides has been established. However, these methodologies require pre-functionalized maleimides as starting materials for the [4 + 2] annulation.

Fig. 1. Traditional N–C_C=O disconnection for N-aryl phthalimides synthesis.

A Traditional method for the synthesis of achiral N-substituted phthalimides. B This work: atroposelective synthesis of N-aryl phthalimides/maleimides.

Although the dehydration of phthalamic acid needs drastic reaction conditions (Fig. 1A), this is a straightforward route to phthalimides from the two readily available starting materials. However, the use of this disconnection for the atroposelective synthesis of N-aryl phthalimides has not been hitherto accomplished. We hypothesized that phthalamic acid generated from phthalic anhydride and 2-substituted aniline could be activated by using pivaloyl chloride (PivCl), and this could possibly generate the isoimide intermediate²⁵. Thereafter, the addition of a chiral N-heterocyclic carbene (NHC)^26–41 to the isoimide could generate the acylazolium intermediate, which make the carbonyl carbon electrophilic enough for the atroposelective amidation of less nucleophilic N-aryl acid amides (Fig. 1B)^42–45. Based on this concept, herein, we report the organocatalytic atroposelective synthesis of N-aryl phthalimides via the traditional N–C_C=O disconnection under mild conditions^46–48. Notably, the NHC-catalyzed activation of carboxylic acids for the enantioselective synthesis of heterocycles was independently reported by the Scheidt⁴⁹, and Ye groups⁵⁰. Although this strategy has been subsequently used by many groups for the enantioselective synthesis of a plethora of chiral compounds^51–54, atroposelective synthesis using the acid activation employing NHC catalysis has not been explored to date^55–61. It is worth mentioning that, related atroposelective amidation strategy has been recently employed by our group for the synthesis of N–N axially chiral quinazolinones⁶².

Results

Reaction optimization for C-N axially chiral phthalimides

Our present study commenced with the treatment of substituted benzoic acid 1a with PivCl and the carbene generated from the chiral pre-catalyst 3⁶³ using K₂CO₃ in THF. Under these conditions, we were elated to obtain the enantioenriched C-N axially chiral phthalimide 2a in 99% yield and 98:2 enantiomeric ratio (er; Table 1, entry 1). The screening of NHC catalysts revealed that carbenes derived from the triazolium salts 4 and 6 were equally effective for catalyzing the amidation reaction; however, the carbene precursor 5 was found to be ineffective (entries 2–4). The reaction did not work without the acid activation using PivCl (entry 5), and the use of HATU as the acid activator instead of PivCl resulted in reduced yield and er of 2a (entry 6)⁵³. The use of other (in)organic bases such as Cs₂CO₃, Na₂CO₃, DMAP and DBU furnished reduced er of 2a although the reactivity was good (entries 7-10). Moreover, the reactions performed in other solvents such as MTBE, CHCl₃ and toluene returned inferior results compared to the standard conditions (entries 11-13). The reaction run for 12 h instead of 24 h afforded 2a in reduced yield of 82% maintaining the high selectivity (entry 14). Hence, entry 1 in Table 1 was selected for the substrate scope analysis (For details, please see the Supplementary Sections 1.3, 1.4).

Table 1.

Optimization of the reaction conditions^[a]

entry	variation of the standard conditionsa	yield of 2a (%)b	er of 2ac
1	none	99 (99)d	98:2
2	4 instead of 3	99	97:3
3	5 instead of 3	<5	ND
4	6 instead of 3	96	5:95
5	without PivCl	<5	ND
6	HATU instead of PivCl	92	87:13
7	Cs2CO3 instead of K2CO3	99	89:11
8	Na2CO3 instead of K2CO3	90	92:8
9	DMAP instead of K2CO3	75	81:19
10	DBU instead of K2CO3	90	87:13
11	MTBE instead of THF	91	93:7
12	CHCl3 instead of THF	80	77:23
13	toluene instead of THF	70	96:4
14	12 h instead of 24 h	82	98:2

^aStandard conditions: 1a (0.125 mmol), 3 (10 mol %), K₂CO₃ (1.5 equiv), THF (1.5 mL), 25 °C and 24 h.

^bDetermined by ¹H NMR analysis of crude products using CH₂Br₂ as the internal standard.

^cThe er value was determined by HPLC analysis on a chiral stationary phase.

^dIsolated yield.

Substrate scope and studies on mechanism

With the identified reaction conditions in hand, the generality of the reaction was then investigated. Initially, the phthalamic acids derived from electronically different anilines were tested (Fig. 2). A series of 2-tert butyl aniline-derived phthalamic acids bearing electron-donating substituents, halogens, and aryl group at the para-position underwent smooth atroposelective amidation reaction under the optimized conditions to furnish the corresponding N-aryl phthalimides in excellent yields and high er values (2b-2g). The electron withdrawing ester moiety was also well tolerated affording 2 h in 97% yield and 92:8 er. The presence of thienyl moiety, and carbon-carbon multiple bonds at the 4-position of aniline did not adversely affect the outcome of the reaction (2i-2l). Moreover, phthalamic acid derived from the meta-nitro substituted aniline afforded the desired product 2 m in 99% yield and 97:3 er. We thereafter searched the other substitutions in the ortho position of anilines. Replacing the tert-butyl group at the ortho position of aniline to a cumene group or a methoxy methyl propyl group yielded the corresponding products 2n and 2o without a noticeable change in the er. The incorporation of di-aryl or di-heteroaryl moieties effectively restrains rotation around the C-N bond, thereby facilitating the formation of 2p and 2q with excellent enantioselectivity. The N-aryl moiety, featuring a phenyl sulfonyl group at the ortho position, poses a challenge in rotational control around the C-N axis, owing to the elongated C–S bond compared to that in the C–C (bulky alkyl group) bond. Despite this challenge, the synthesis of 2r was accomplished with 71% yield and 80:20 er. Notably, the reactions performed using di-ortho substituted aniline-derived phthalamic acid substrates although reacted well but with poor er values. Next, we evaluated the variation in the benzoic acid moiety. Electron-donating groups and halo substitution at the ortho-position of phthalamic acids provided the target N-aryl phthalimides with excellent yields and enantioselectivities (2s-2x).

Fig. 2. Substrate scope for the synthesis of C-N axially chiral phthalimides.

General conditions: 1 (0.25 mmol), 3 (10 mol %), PivCl (1.5 equiv), K₂CO₃ (1.5 equiv), THF (3.0 mL), 25 °C, and 24 h. Yields of isolated products are given and the er was established by HPLC analysis on a chiral stationary phase. ^a 15 mol % 3 and 2.5 equiv of K₂CO₃ were used. ^b Reaction performed in 0.125 mmol scale. ^c DABCO was used as base and reaction performed at −20 °C.

The structure and the absolute stereochemistry of the C–N axis was confirmed by the X-ray analysis of compounds 2w and 2x (CCDC 2252546 (2w) and CCDC 2252545 (2x)). The absolute configuration of the C–N axis of other phthalimides was assigned by analogy. Notably, the 2-nitro substituted phthalamic acid provided the product 2 y in 99% yield and 88:12 er. Interestingly, scope of the present methodology could be expanded beyond the atroposelective synthesis of monosubstituted phthalimides. Phthalamic acids generated from di- and tri-substituted phthalic anhydrides also proved to be effective substrates, resulting in the synthesis of di- and tri-substituted phthalimides in excellent yields and enantioselectivities. Unsymmetrical naphthyl phthalic anhydride-derived phthalimide 2z, was formed in 95% yield with 96:4 er. Similarly, disubstituted phthalimides such as 4-methoxy-5-nitro phthalimide 2aa and 4-methoxy-7-nitro phthalimide 2ab were formed in excellent yields with good to excellent enantioselectivities. Moreover, sterically demanding tri-substituted phthalimic acid did not affect the reaction outcome (2ac). Furthermore, performing the reaction with the acid derived from pyridinic anhydride afforded the desired product 2ad in 99% yield and 71:29 er, proving the generality of the present methodology.

The reaction of unsymmetrical phthalic anhydride with aniline derivatives generally produces two regioisomeric phthalamic acids, where one regioisomer forms predominantly due to the selective amine addition to the less sterically hindered carbonyl. For instance, the treatment of 4-methylisobenzo furan-1,3-dione with 2-tert-butyl aniline afforded the phthalamic acids 1a and 1a’ as separable mixture of regioisomers in 56% and 12% isolated yields respectively (Fig. 3A). The structure of regioisomer 1a was confirmed using X-ray analysis of the crystals (CCDC 2261808 (1a)).

Fig. 3. Synthesis of both enantiomers using the same NHC catalyst and studies on C-N rotational barrier.

A Accessing both enantiomers of the product from the same phthalic anhydride and aniline. B studies on rotation barrier: (i) effect of C-N bond rotation on temperature, (ii) plot for calculating the rotational barrier (M and m represent the percentage of major and minor enantiomer).

Interestingly, both enantiomers of the C-N axially chiral N-aryl phthalimide can be accessed using the same phthalic anhydride and aniline employing the same enantiomer of the NHC pre-catalyst. Treatment of 1a under the optimized conditions afforded 2a in 99% yield and 98:2 er. The reaction of the regioisomer 1a’ with the same NHC pre-catalyst 3 furnished the enantiomer of 2a (ent-2a) in 99% yield and 3:97 er. Thus, by tuning the substitution position on benzoic acid moiety, the opposite enantiomer is accessible.

To gain insight into the rotational restriction around the C–N bond, the N-aryl phthalimide 2a was heated for 2 h at different temperatures in toluene and monitored the change in ee values. Notably, up to 70 °C, there was no change in ee value revealing the restricted rotation around the C–N bond (Fig. 3B). Further, an increase in temperature beyond 70 °C allowed the rotation around the C–N axis as observed by the lowering of ee values, and at 140 °C the ee was 0% indicating complete racemization. For compound 2n, the rotational restriction around the C–N bond was observed up to 50 °C, and the complete racemization was observed at 120 °C. We have also determined the C–N rotational barrier for compounds 2a and 2n experimentally and with the aid of density functional theory (DFT) studies. By monitoring the variation of er values at different time intervals while keeping the temperature at 100 °C, the Δ${\mathrm{G}}_{\mathrm{rot}}^{‡}$ for the C–N bond in 2a was determined experimentally to be (30.46 ± 0.03) kcal/mol using the procedure of Curran⁶⁴. The DFT calculated value was 31.9 kcal/mol, which is in good agreement with the experimental value. Similarly, the experimental C–N rotational barrier for 2n was (29.35 ± 0.04) kcal/mol, and the calculated value was 30.2 kcal/mol. Moreover, the t_1/2 of racemization for 2a and 2n were determined to be 35.8 years and 5.5 years respectively at 25 °C.

The present catalytic strategy is scalable in a 1.0 mmol scale without erosion of reactivity and selectivity. The reaction of 1a under the optimized conditions furnished 2a in 99% yield and 98:2 er (Fig. 4A). Moreover, lowering the catalyst loading to 2.0 mol % of 3, the desired products 2a and 2t were synthesized in comparable yields and selectivities (Fig. 4B), thereby demonstrating the practicality of the developed protocol.

Fig. 4. Practicality of the developed protocol.

A Scale-up synthesis of N-aryl phthalimides. B Low catalyst loading synthesis of N-aryl phthalimides.

To gain insight into the mechanism of the reaction, a few experiments were performed. When the phthalamic acid 1a was treated with PivCl under the optimized reaction conditions in the absence of the NHC precursor, the isoimide 7a was formed in 99% yield (Fig. 5A)²⁵. Interestingly, when the isolated isoimide 7a was treated with NHC precursor and base, the desired phthalimide 2a was formed in 98% yield and 96:4 er (Fig. 5B). These experiments clearly indicate that the isoimide 7a, is the intermediate in the present reaction (For details, please see Supplementary Sections 2.1(a), 2.1(b)). These studies revealed that the crucial acylazolium intermediate was formed through the addition of NHC to the isoimide intermediate. To the best of our knowledge, such NHC addition to isoimide has not been documented previously, thus paving a unique pathway for the generation of the acylazolium intermediate^26–41.

Fig. 5. Mechanistic experiments.

A Isoimide intermediate isolation. B Reaction using isoimide intermediate.

Based on the results of the mechanistic experiments, a tentative mechanism of the reaction has been proposed (Fig. 6). The reaction likely proceeds via the in-situ activation of phthalamic acid 1a in the presence of PivCl to form activated anhydride (I), which subsequently undergoes an intramolecular cyclization leading to the formation of the isoimide 7a. The nucleophilic addition of the carbene generated from 3 to the isoimide could generate the acylazolium intermediate (II), which upon atroposelective amidation followed by regeneration of NHC furnished the desired C-N axially chiral phthalimides 2a. Based on the absolute configuration determined using X-ray analysis, a plausible mode of enantioinduction has been proposed. In Int-I, the bulky aminoindanol and tert-butyl groups lie on opposite sides of the plane defined by the phthalamic acid moiety, thereby having lower energy than the sterically congested Int-II. Hence, the reaction is likely proceeding via the Int-I.

Fig. 6.

Tentative mechanism of the reaction and the envisioned mode of enantioinduction.

Moreover, control experiments indicated that stirring the acid 2a in THF in the absence of 3 and K₂CO₃ afforded 2a (racemic) in 52% yield (For details, please see the Supplementary Section 2.1(d)). The in situ generated HCl (during the pivaloyl anhydride formation, intermediate I, Fig. 6) could mediate the isoimide 7a to phthalimide 2a conversion. This was further confirmed by the direct conversion of isoimide 7a to phthalimide 2a in 58% yield by treatment with HCl in dioxane.

Atroposelective synthesis of C-N axially chiral maleimides

The present NHC-catalyzed atroposelective strategy can easily be extended to the atroposelective synthesis of N-aryl maleimides (Fig. 7). The synthesis of axially chiral N-aryl amino maleimides was recently disclosed by our group⁶⁰, but this strategy involved the kinetic resolution of pre-functionalized maleimides via [3 + 3] annulation strategy in the presence of chiral NHC precursor, where a maximum 50% yield of the maleimides was accomplished. Herein, with the slightly modified reaction conditions, the N-aryl maleimide 9a was formed in 65% yield with 93:7 er. Halo substitution (9b, 9c), as well as the electron-withdrawing CO₂Et group (9d) present in the para-position of aniline afforded the atroposelective amidation product in good yields and er values. Moreover, cumene moiety present at the ortho-position of aniline also furnished the desired maleimide 9e without a notable change in er. Moreover, the present methodology also well effective for the atroposelective synthesis of disubstituted maleic anhydrides 9f in good yield with moderate er value.

Fig. 7. Scope for the synthesis of C-N axially chiral maleimides.

General conditions: 8 (0.25 mmol), 3 (10 mol %), PivCl (1.5 equiv), K₂CO₃ (1.5 equiv), THF (1.5 mL), 25 °C, and 36 h. Yields of isolated products are given and the er was established by HPLC analysis on a chiral stationary phase.

Synthetic applications

The C-N axially chiral phthalimide 2a and maleimide 9a can be utilized as synthetically useful precursors for further synthetic elaborations of C-N axially chiral derivatives. Benzylic bromination of 2a in presence of NBS/benzoyl peroxide leads to the synthesis of C-N axially chiral benzyl bromide derivative 10a (Fig. 8). Oxidation of 10a in presence of IBX and DMSO afforded the axially chiral aldehyde 11a in 93% yield and 97:3 er. An alkene moiety was successfully installed via the Wittig reaction using the aldehyde 11a, which afforded the styrene derivative 12a in 66% yield and 94:6 er. The nucleophilic substitution of bromide in 9a by using NaN₃ furnished the benzyl azide 13a in 94% yield and 98:2 er. Interestingly, the (3 + 2) cycloaddition of the azide 13a with the aryne generated from the TMS-triflate precursor under transition-metal-free conditions resulted in the synthesis of the benzotriazole derivative 14a in 95% yield and 98:2 er. Regarding the functionalization of the N-aryl maleimides, monobromination of the double bond in 9a was accomplished using bromine and the desired product 9f was formed in 92% yield and 92:8 er. The bromo derivative 9f underwent aza-Michael reaction with n-butyl amine to form the C-N axially chiral amino maleimide 15a in 41% yield and 95:5 er. Finally, hydrogenation of 9a using Pd/C afforded the C-N axially chiral N-aryl succinimide 16a as a single diastereomer bearing both point and axial chirality.

Fig. 8. Functionalization of C-N axially chiral phthalimides and maleimides.

A Functionalization of C-N axially chiral N-aryl phthalimides. B Functionalization of C-N axially chiral N-aryl maleimides.

Discussion

In conclusion, we have demonstated NHC-catalyzed atroposelective synthesis of N-aryl phthalimides and maleimides by the traditional N-C_C=O disconnection approach employing the acid activation strategy. By using the conventional and simple disconnection approach for N-aryl phthalimides, the reaction proceeds under much milder conditions, excellent yields, and high enantioselectivities, and is suitable for lower catalyst loadings⁶⁵. Interestingly, using the same enantiomer of NHC pre-catalyst, both enantiomers of the product could be accessed starting from same phthalic anhydride and aniline. Preliminary mechanistic investigation unveils that PivCl triggers the activation of phthalamic acid, culminating in the formation of a reactive isoimide intermediate. Subsequent introduction of NHC opens a unique avenue for the generation of the pivotal acylazolium intermediate, marking a crucial step in the process. In addition, the rotational energy barrier for the C-N bond in N-aryl phthalimides were calculated experimentally and using DFT study concluding that the products have configurationally stable C-N chiral axis. The derivatized axially chiral phthalimides/maleimides have proven the synthetic utility of the present study.

Methods

Procedure for the atroposelective synthesis of N-Aryl phthalimides

In a flame-dried screw-capped test tube equipped with a magnetic stir bar was taken K₂CO₃ (52 mg, 0.375 mmol, 1.5 equiv) from the glovebox, then phthalamic acid derivative 1 (0.25 mmol, 1.0 equiv) was added. Then, the screw-capped tube was evacuated and backfilled with argon. To this mixture was added THF (2.0 mL) under argon atmosphere followed by PivCl (43 μL, 0.375 mmol, 1.5 equiv). The resultant reaction mixture was kept stirring at 25 °C until the full conversion of acid to the corresponding anhydride (monitored by TLC; typically, 30 minutes). To this mixture, the triazolium salt 3 (9.2 mg, 0.025 mmol, 10 mol %) and THF (1.0 mL) were successively added and stirred for 24 h. Then the solvent was evaporated to get the crude residue, which was purified by flash column chromatography on silica gel to afford the corresponding C-N axially chiral phthalimides.

Procedure for the atroposelective synthesis of N-Aryl maleimides

In a flame-dried screw-capped test tube equipped with a magnetic stir bar was taken K₂CO₃ (52 mg, 0.375 mmol, 1.5 equiv) from the glovebox, then carboxylic acid derivative 8 (0.25 mmol, 1.0 equiv) was added. Then, the screw-capped tube was evacuated and backfilled with argon. To this mixture was added THF (1.0 mL) under argon atmosphere followed by PivCl (43 μL, 0.375 mmol, 1.5 equiv). The resultant reaction mixture was kept stirring at 25 °C until the full conversion of acid to the corresponding anhydride (monitored by TLC; typically, 30 minutes). To this mixture, the triazolium salt 3 (9.2 mg, 0.025 mmol, 10 mol %) and THF (0.5 mL) were successively added and stirred for 36 h. Then the solvent was evaporated to get the crude residue, which was purified by flash column chromatography on silica gel to afford the corresponding C-N axially chiral maleimides.

Procedure for the racemic synthesis of N-Aryl phthalimides/maleimides

In a flame-dried screw-capped test tube equipped with a magnetic stir bar was taken K₂CO₃ (21.0 mg, 0.15 mmol, 1.5 equiv) from the glovebox, then carboxylic acid derivative 1/9 (0.1 mmol, 1.0 equiv) was added. Then, the screw-capped tube was evacuated and backfilled with argon. To this mixture was added THF (0.8 mL) under argon atmosphere followed by PivCl (18 μL, 0.15 mmol, 1.5 equiv). The resultant reaction mixture was kept stirring at 25 °C until the full conversion of acid to the corresponding anhydride (monitored by TLC; typically, 30 minutes). To this mixture, the triazolium salt 17 (2.7 mg, 0.01 mmol, 10 mol %) was added and stirred for 24 h. Then the solvent was evaporated to get the crude residue, which was purified by column chromatography on silica gel to afford the corresponding racemic phthalimides/maleimides.

Author contributions

S.B. and A.T.B. conceived and designed the project. S.B. performed the optimization studies, substrate scope analysis and mechanistic studies. S.S.R. helped in the substrate scope studies. S.B. wrote the first draft of the manuscript and A.T.B. edited the manuscript. All authors have given approval to the final version of the manuscript.

Peer review

Peer review information

Nature Communications thanks Jian Wang and the other anonymous reviewer(s) for their contribution to the peer review of this work. A peer review file is available.

Data availability

The authors state that the data supporting the findings of this study are available within the article and Supplementary Information file, or from the corresponding author upon request. The data for the coordinates of the optimized structures are present in the Source Data file in the Supplementary Information Section. Crystallographic data for the structures reported in this article have been deposited at the Cambridge Crystallographic Data Center, with the deposition numbers CCDC 2252546 (2w), CCDC 2252545 (2x), CCDC 2261808 (1a) and CCDC 2261807 (1w). These data can be obtained free of charge from Data Center via www.ccdc.cam.ac.uk/data_request/cif Source data are provided with this paper.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at 10.1038/s41467-024-49799-5.