Facile Chemodivergent Synthesis of Fused Indolo[2,3‑b]quinoxalines and Benzo[4,5]imidazo[1,2‑c]quinazolin-6(5H)‑ones: An Organocatalytic Trifluoromethanesulfonic Acid-Controlled Mechanistic Switch between Condensation and Ring Expansion of the Isatin Core

Abstract

A mild metal-free chemodivergent synthesis leading to products of two different chemotypes, such as indolo­[2,3-b]­quinoxalines and benzo­[4,5]­imidazo­[1,2-c]­quinazolin-6­(5H)-ones, has been described from common starting materials, isatin and 1,2-diaminobenzene, at room temperature in an environmentally benign EtOH solvent. The EtOH mediated condensation between isatin and 1,2-diaminobenzene delivers exclusively fused indolo­[2,3-b]­quinoxalines in good/excellent yields (65–89%) in 8 h, and catalytic protic acid (CF₃SO₃H) in EtOH furnishes N-fused benzo­[4,5]­imidazo­[1,2-c]­quinazolin-6­(5H)-ones selectively in good/excellent yields (66–88%) in 2 h via rearrangement of the isatin core following C2–C3 bond cleavage and N-atom insertion with 1,2-diaminobenzene. Thus, the switch selectivity between benzo­[4,5]­imidazo­[1,2-c]­quinazolin-6­(5H)-ones and indolo­[2,3-b]­quinoxalines could be finely programmed with the presence/absence of catalytic CF₃SO₃H induced by an ionic vs non-ionic pathway. A wide array of starting materials housing structural and electronic variations (e.g., electronically neutral, moderate/strong electron-donating, moderate electron-attracting functionalities, and heteroarene) allows the scope of diversity creation and provides facile access to molecular libraries with elaborate, well-decorated structures. The operational simplicity, ease of product purification (in the case of indolo­[2,3-b]­quinoxalines) by precipitation/filtration, recyclability of catalytic CF₃SO₃H, and a high degree of chemoselectivity through decisive cleavage of the C2–C3 bond over other labile bonds under protic acid catalysis signify distinct advantages of developed protocols. The late-stage synthetic manipulation at the free-NH site of final products and a “gram-scale” preparation further demonstrate the promising usefulness of chemodivergent synthesis.

Introduction

The N-heterocycles are profusely present as primary pharmacophoric units in many approved, prescribed, purchasable drugs and several marketed pharmaceutical agents. Numerous dietary components, nutraceuticals, and agrochemicals are enriched with the N-heterocycle as a key constituent. The early stage of a drug discovery program through the selection of hit-to-lead candidates, lead optimization, and late lead derivatization is majorly centered on small molecule N-heterocyclic scaffolds collected from large compound libraries of elaborated structural diversity of N-heterocycle backbones, which in turn are identified from an efficient evaluation of structure–activity relationship (SAR) and structure–property relationship (SPR) studies. The smooth biological assessment of N-heterocycle-based scaffolds at the discovery medicinal chemistry level to announce new lead candidates demands an uninterrupted supply chain of compounds rapidly synthesizing a large molecular library for an efficient SAR/SPR survey. The synthetic organic and medicinal chemistry enterprises are constantly engaged in developing fit-for-purpose synthetic routes of small-molecule N-heterocyclic scaffolds for easy and quick access to highly decorated molecular libraries, enabling timely SAR/SPR investigations. Moreover, a well-defined synthetic strategy of N-heterocycles allows rapid and facile scale-up of an identified new chemical entity (NCE) or new molecular entity (NME) at the drug development phase.

The chemodivergent transformation has been recognized as an efficient synthetic tool for facile diversity generation in molecular structures from common starting materials by manipulating the difference in reactivity among functional groups through judicial selection of catalyst, co-catalyst, ligand, additive, solvent, reaction temperature, reaction time, pH, and even a small change in the substrate. The application of chemodivergent synthesis has seen impetus in recent years, offering easy access to products of different chemotypes from the same starting materials to expand the chemical, molecular, and high-quality drug-like space for a quick and straightforward exploration of SAR/SPR analysis.

The isatin moiety is a privileged, medicinally important fused N-heterocycle that houses two electronically distinguishable carbonyl groups, such as amide and ketone carbonyl. The isatin scaffold has been regarded as a key building block responsible for numerous biological events/functions and further has been studied popularly to access diversified small-molecule heterocyclic entities. It was envisioned that a diversity-oriented synthesis around chemically different carbonyl groups of the isatin core with a common starting material, e.g., 1,2-diaminobenzene, through judicial engineering of reaction parameters, would result in fused binuclear polycyclic products of different chemotypes, such as indolo­[2,3-b]­quinoxaline and benzo­[4,5]­imidazo­[1,2-c]­quinazolin-6­(5H)-one, and could further expedite the assembly of a focused large molecular library of N-heterocycles possessing structural diversity. Several therapeutic new lead molecules are enriched with a (6,5,6)-membered fused indolo­[2,3-b]­quinoxaline unit as an essential core and display a broad spectrum of biological and pharmacological functions, such as antibacterial activity, anticancer activity, antiviral activity, inhibition of β-catenin/T-cell factor protein–protein interaction, etc., to name a few (Figure A). The other analogous (6,6,5)-membered N-fused imidazo­[1,2-c]­quinazolin-6­(5H)-one is an important polynuclear hybrid scaffold; however, it has been relatively under­explored for biological activities. For example, the imidazo­[1,2-c]­quinazolin-6­(5H)-one serves as a key nucleus in identified new lead molecules to exhibit potent hypertensive activity by selectively inhibiting the α1-adrenoceptor and a weak anxiolytic profile as a potent inhibitor of the benzodiazepine receptor (Figure B). Even the biological and pharmacological studies of tetracyclic N-fused congener benzo­[4,5]­imidazo­[1,2-c]­quinazolin-6­(5H)-one have not been explored yet. Perhaps, the definitive C2–C3 bond cleavage, N-atom insertion, and subsequent ring expansion of the isatin core to the imidazo­[1,2-c]­quinazolin-6­(5H)-one ring in the presence of multiple reactive sites and labile bonds, such as the C3-site and N1–C2 bond, remain as a formidable challenge and require deliberate reaction conditions.

1.

Selected examples of new lead molecules housing (A) indolo­[2,3-b]­quinoxaline and (B) imidazo­[1,2-c]­quinazolin-6­(5H)-one core.

The condensation reaction of isatin (1) and the 1,2-diaminobenzene derivative (2) allows a direct synthesis of the tetracyclic indolo­[2,3-b]­quinoxaline (3) scaffold using bentonite clay in an EtOH/H₂O solvent system under microwave irradiation or under conventional heating (70 °C), under microwave irradiation in a glacial AcOH solvent , and in neat conditions using catalytic para-toluenesulfonic acid, in a refluxing AcOH solvent or biphasic AcOH/EtOH solvent system, and using catalytic (20 mol %) sulfamic acid in EtOH at room temperature. The reports, however, work with a very limited substrate scope (only one/two/three/four/seven examples) without any attempt of generalization of electronic/steric influences of the substituents ,− and require a specialized reaction set-up/apparatus and environmentally hazardous solvents (e.g., N,N-dimethylformamide (DMF) and CH₂Cl₂) for product purification, ,, and the isolated yields of final products are very poor. The Pd­(II)/Pd(0)-catalyzed 2-fold C–N coupling strategy of 2,3-dibromoquinoxalines with secondary aryl amines or 2-bromo-3-(2-bromophenyl)­quinoxaline with primary aryl/aliphatic amines has been reported to synthesize 3 using super-stoichiometric base, such as NaO ^t Bu (3 equiv), expensive specialized phosphine-based ligands at high reaction temperature (100–105 °C), and super-stoichiometric 2,3-dibromoquinoxalines (1.5 equiv) or primary aryl/aliphatic amines (3 equiv). The Pd­(II)-catalyzed Buchwald–Hartwig coupling followed by an annulation by an intramolecular oxidative cyclodehydrogenation strategy has been applied to synthesize structurally similar 5H-[1,2,5]­oxadiazolo­[3′,4′:5,6]­pyrazino­[2,3-b]­indole entities and relies on super-stoichiometric (2.5 equiv) base K₃PO₄ in refluxing toluene and further requires a prolonged reaction period of 40 h. However, the residual heavy metal impurities and contamination in pharmaceutical intermediates and active pharmaceutical ingredients are a severe concern for effective biological evaluation of drug-like substances.

The transformation of 1 to the benzo­[4,5]­imidazo­[1,2-c]­quinazolin-6­(5H)-one (4) motif typically relies on relatively harsh reaction conditions. It follows a cascade reaction sequence of C2–C3 bond cleavage, a new N-atom insertion, and subsequent isatin ring re-organization. A metal-free oxidative environment (O₂ atmosphere) in DMSO solvent requires 130 °C temperature for successful conversion of 1 to 4 in a reaction with 2. A combination of stoichiometric molecular iodine (1.2 equiv) plus super-stoichiometric (5 equiv) tert-butyl hydroperoxide (TBHP) is needed in DMSO solvent at 90 °C for the synthesis of 4 when the starting materials are indole and 2. Very recently, organic Brønsted acid-catalyzed skeletal editing technology has been disclosed to synthesize structurally relevant scaffold dibenzo­[a,c]­phenazines from 2-arylindoles with 1,2-diaminoarenes, which relies on super-stoichiometric oxidant, tert-butyl nitrite (TBN) (4 equiv), using a hazardous CH₂Cl₂ solvent at high temperature (120 °C). Catalyst-free reaction of 1 and 2 in aq. EtOH (1:1 v/v) at 80 °C has been reported synthesizing 4. The stoichiometric (1 equiv) and super-stoichiometric (2 equiv) Mo­(CO)₆ has been applied as a reductant and promoter of cyclization between 1 and 2-nitroaniline/1,2-dinitrobenzene in DMF at 90 °C, forming 4, which requires a prolonged reaction time (55 h). Even the stoichiometric/super-stoichiometric metal raises the issue of metal contamination in final desired products. The reported protocols indeed use nondesirable hazardous solvents (DMSO, CH₂Cl₂) − and stoichiometric/super-stoichiometric strong oxidants (TBHP, TBN), which are toxic and corrosive in nature, , and require high temperatures (120 °C, 130 °C) , and a longer reaction period (55 h). Thus, there is a need to develop a better, cleaner, milder, and more robust technology to synthesize 3 and 4 from commonly available starting materials.

At this outset, we planned to design an approach with an organo-catalyst-controlled selectivity switch between products of two different chemotypes 3 and 4 from common starting materials 1 and 2 in a mild, eco-responsible condition. The sole report describes the selectivity switch between 3 and 4 from starting materials 1 and 2, governed by the electronic characteristics of the benzenoid unit in 1 and not by other reaction parameters. The report has a scarcity of substrate scope and exemplifies only four derivatives. Herein, we demonstrate facile synthesis of 3 and 4 from 1 and 2 and judicial manipulation of 3 vs 4 selectivity in the absence/presence of catalytic trifluoromethanesulfonic acid (CF₃SO₃H) at room temperature, enabling straightforward access to a large compound library comprising structural diversity to enrich the medicinal chemist’s synthetic toolbox.

Results and Discussion

The facile cleavage of the labile N1–C2 bond of 1 under an alkaline condition during the reaction with 2 is known. The popularity of Brønsted acid-catalyzed imine formation via condensation of an aldehyde/ketone with amines caught our attention. In a model study, the interrogation of condensation between N-phenylisatin (1a) (1 equiv) and 1,2-diaminobenzene (2a) (1 equiv) was started in the presence of different catalytic Brønsted acids in EtOH to identify the best optimized reaction condition (Table ). To our surprise, organic protic acids, such as para-toluenesulfonic acid (TsOH·H₂O) and methanesulfonic acid (CH₃SO₃H) in catalytic quantity (each 10 mol %), afforded 3a in 29% and ring expansion product 4a in 48% isolated yield (entry 1) and 3a in 38% and 4a in 59% isolated yield (entry 2) after 5 h at room temperature (28 °C, oil bath). The stronger catalytic protic acid CF₃SO₃H (10 mol %) allowed a faster reaction rate and selectively delivered 4a in 88% isolated yield (major product) after 2 h (entry 3). The results are in contrast to the prior report, where catalytic sulfamic acid (20 mol %) gave exclusively condensation product 3. The reduction of the catalytic load of CF₃SO₃H to 5 mol % provided an inferior result, albeit 4a was the major product (entry 4). However, an increase in the quantity of CF₃SO₃H to 20 mol % did not improve the reaction outcome (entry 5 vs entry 3). Replacing CF₃SO₃H with catalytic AcOH (10 mol %), however, switched the selectivity to condensation product 3a (isolated in 44% yield) along with 4a in 16% isolated yield after 5 h at room temperature (28 °C, oil bath) (entry 6). Compared to AcOH, a stronger protic acid, trifluoroacetic acid, in catalytic quantity (10 mol %) delivered 4a as the major one (57% isolated yield) and 3a as the minor product (27% isolated yield) after 5 h (entry 7). The weaker pivalic acid in catalytic amounts (10 mol %) proved to be inferior, although 3a was the major product (isolated yield 35%) and 4a was the minor product (isolated yield 16%) (entry 8). Thus, it is clear that the selectivity switch between 3a and 4a could be easily adjusted by applying appropriate catalytic protic acids with varying acid strength, which also determines the overall product conversion and speed of the reaction. In an attempt to push the selectivity exclusively to the condensation product, the reaction between 1a and 2a was conducted without any catalyst at room temperature (28 °C, oil bath) in EtOH, and 3a was obtained exclusively in decent yield (68%); however, the reaction was sluggish (8 h) (entry 9). Refluxing EtOH without any catalyst improved the rate of reaction, but the yield of 3a was reduced to 54%, and 4a was isolated as a minor product (22% isolated yield). A systematic screening of several reaction parameters was performed (Tables S1–S3, electronic Supporting Information for a detailed description). Finally, the best optimized reaction condition to access exclusively 3a has been selected as entry 9 in Table , and the condition for the selectivity switch to 4a has been identified as entry 3 in Table .

1. Representative Data of Reaction Condition Optimization for Condensation/Ring Expansion of 1a with 2a .

				yield (%)
entry	catalyst (X)	temp. (°C)	time (h)	3a	4a
1	TsOH.H2O (10)	rt	5	29	48
2	CH3SO3H (10)	rt	5	38	59
3	CF3SO3H (10)	rt	2	08	88
4	CF3SO3H (5)	rt	2	28	61
5	CF3SO3H (20)	rt	2	11	86
6	AcOH (10)	rt	5	44	26
7	CF3CO2H (10)	rt	5	27	57
8	PivOH (10)	rt	5	35	16
9	none	rt	8	68	0
10	none	reflux	4	54	22

^aReaction condition: 1a (1 mmol), 2a (1 mmol, 1 equiv), catalyst (X mol %), EtOH (2 mL), rt (28 °C, oil bath) or reflux.

^bIsolated yield of 3a and 4a.

Subsequently, the scope of the reaction was studied with structurally and electronically divergent 1 and 2 synthesizing various functionalized 3 (Table ). The electronic and steric influences of the substituents on the aromatic units and N1-atom of 1 and 2 were evaluated. The electronically neutral N1-phenyl ring containing 1 reacted with electron-neutral 2a and electron-rich 4,5-dimethyl-1,2-phenylenediamine (2b) and afforded the corresponding fused polycyclic products 3a/3b in good yields. Moderate and strong electron-releasing functionality, such as –Me, –OMe, and –F on the benzenoid ring, the N1-aryl unit, and 2, performed the reaction well and delivered the products 3c, 3d, 3e, and 3f in synthetically useful yields. The moderately electron-withdrawing –Cl atoms on 2 provided the corresponding condensation product 3g in 69% isolated yield. The –Me functionality at the N1-site instead of the bulky N1-aryl ring smoothly participated in the reaction with electronically neutral 2a, electron-rich 2 housing F-atoms, and electron-deficient 2 possessing Cl-atoms to deliver the respective products 3h, 3i, and 3j in decent yields. Unsymmetrical 4-methyl-1,2-phenylenediamine (2e) furnished a regio-isomeric mixture of condensation products (3k + 3k′) in good yield (65%) with 3k/3k′ ≈ 1:1. Satisfyingly, the free-NH isatin moiety reacted well with 2a and afforded the desired condensation product 3l in 89% isolated yield. The isatin unit housing an N1-heteroarene, such as the 3-thiophene ring, performed the condensation reaction well with 2a and delivered the respective product 3m in synthetically useful yield. In none of the cases was the ring expansion N-fused polycyclic product 4 observed, which mirrors the specific reactivity of nucleophilic amines with carbonyl groups of 1 under mild conditions. All of the final products (3a–3m) were precipitated out of EtOH solution after completion of the reaction. The precipitates were filtered out, and the residues were washed with cold EtOH and further re-crystallized in EtOH to collect the desired products in pure form, except the regio-isomeric mixture of (3k + 3k′), and thus avoid column purification. The filtrate (EtOH solution) could further be used for the next batch of reactions. Thus, the in-built separation of products from EtOH solution after completion of reaction typically permits a continuous mode of operation for multiple batches.

2. Substrate Scope Study of Condensation of the Isatin Scaffold (1) with 1,2-Diaminobenzene Derivatives (2) Forming 3 ^,

^aReaction condition: 1 (1 mmol), 2 (1 mmol, 1 equiv), EtOH (2 mL), rt (28 °C, oil bath), 8 h.

^bIsolated yield of 3.

^cReaction was performed at 40 °C.

Next, we turned our attention to shift product selectivity to 4, synthesizing various functionalized analogues and further generalizing the substrate scope with structurally and electronically diversified 1 and 2 under the standard condition stated in entry 3 of Table . Subsequently, the electronic and steric properties of the substituents on the N1-atom of 1 and aromatic rings and of both 1 and 2 were assessed (Table ). Electronically neutral 1a and 2a reacted well and gave the corresponding ring expansion product 4a in excellent yield (88%) along with a minor amount of condensation product 3a (8% isolated yield). Moderate and strong electron-donating groups, such as –Me, –OMe, and –F attached to the N1-aryl ring and the benzenoid unit of 1 and on 2 at various locations, performed the transformation easily to afford the desired ring-expansion products 4b, 4c, 4d, 4e, 4f, and 4g in good yields (in the range of 66%–80%), and the respective condensation products 3c, 3c′, 3d′, 3e′, 3f′, and 3g′ were isolated as minor isomers (in the range of 11%–18%). The combination of moderate/strong electron-releasing groups, such as –Me, –OMe, and –F on the N1-aryl unit and benzenoid ring of 1, and moderate electron-attracting functionality, such as –Cl on 2, effectuated the transformation, readily delivering expected ring expansion products 4h, 4i, and 4j in excellent yields (83%–88%) along with minor condensation products 3h′, 3i′, and 3j′. The substitution of the N1-aryl ring with an electron-deficient N1-benzyl ring possessing a para–CF₃ group and a –F functionality on the benzenoid ring provided the desired ring expansion product 4k in synthetically useful yield in a reaction with 2b, and the condensation product 3k″ was isolated as a minor isomer (18% yield). Pleasingly, free-NH isatin performed the reaction smoothly with 2a and afforded the desired ring expansion product 4l in an excellent yield (85%) as well as the corresponding condensation product 3l, which was isolated as a minor isomer (12% isolated yield), allowing scope for further synthetic elaboration at the free-NH site. The unsymmetrical 2e on reaction with N-methylisatin furnished a regio-isomeric mixture of ring expansion products (4m + 4m′) in excellent yield (83%) in an equimolar ratio (≈1:1) along with a minor regio-isomeric mixture of condensation products (3m′ + 3m″) in an equimolar ratio (≈1:1). A heteroarene, such as a 3-thiophene ring corded to the N1-site of isatin, on reaction with 2a afforded the desired ring expansion product 4n as a major isomer in an excellent yield (86%) along with the minor condensation product 3m (10% yield). In all examples mentioned in Table , the ring expansion N-fused polycyclic products (4) have been isolated as the major isomer plus a minor quantity of condensation products (3). The results are in sharp contrast to the exclusive formation of 3 from a condensation of 1 and 2 under the catalytic influence of sulfamic acid. The present standard catalytic condition with CF₃SO₃H (10 mol %) exhibits a high degree of chemoselectivity through a preferential cleavage of the C2–C3 bond over the labile N1–C2 bond of 1 and is primarily responsible for the products of two different chemotypes, 4 as a major isomer and 3 as a minor isomer in Table . Moreover, the developed technology for ring expansion product 4 operates at room temperature with a faster rate of reaction using environmentally responsible solvent EtOH, unlike the prior reports, ,, and offers a mild, robust, efficient catalytic process for straightforward access to ring-rearranged N-fused polycyclic product benzo­[4,5]­imidazo­[1,2-c]­quinazolin-6­(5H)-one. In all cases of Table , the crude mixture of 3 and 4 was precipitated out of EtOH after completion, and the mixture was further separated by column chromatography to isolate 3 and 4 in pure form, except for regio-isomeric mixtures (4m + 4m′) and (3m′ + 3m″).

3. Substrate Scope Study of Ring Expansion of the Isatin Scaffold (1) with 1,2-Diaminobenzene Derivatives (2) Forming 4 ^,

^aReaction condition: 1 (1 mmol), 2 (1 mmol, 1 equiv), CF₃SO₃H (10 mol %), EtOH (2 mL), rt (28 °C, oil bath), 2 h.

^bIsolated yield of 4 and 3.

^cReaction was performed at 40 °C.

In representative examples, the structures of compound 4a and compound 3c′ were unambiguously confirmed through single-crystal X-ray diffraction analysis (Tables S5 and S6, electronic Supporting Information). The oak ridge thermal ellipsoid plot (ORTEP) diagrams of compound 4a (CCDC 2490350) and compound 3c′ (CCDC 2501082) are presented in Figure .

2.

ORTEP representations of (A) compound 4a (CCDC 2490350) and (B) compound 3c′ (CCDC 2501082).

The compounds 4l and 3l were successfully synthetically modified at the free-NH site by attaching a propargyl group following base (NaH)-mediated N-propargylation with propargyl bromide (5a) in DMF, which resulted in the respective products 4o and 3n′ in quantitative yields (Schemes A and B). The results demonstrate the possible utility of the developed chemodivergent protocols, which indeed offer late-stage diversification at the appropriate position. In fact, the easy synthetic manipulation forming structurally decorated compounds (4o and 3n′) could reasonably endow diversification of a given molecular library. Furthermore, the reproducibility, scalability, and potential usefulness of the chemodivergent synthesis have been proved through a “gram-scale” preparation of compound 4a (Scheme C). A thorough comparison of product yields and different reaction parameters for the synthesis of condensation product 3l and ring-expansion product 4l with prior reports ,− ,,,, (Tables S7 and S8, electronic Supporting Information) as representative examples establishes the superiority of the presently developed protocols in terms of using metal-free organocatalytic conditions, employing an environmentally preferred solvent, being devoid of toxic corrosive oxidizing agents and acidic solvents, remaining free from expensive ligands and (super)­stoichiometric transition metals, not requiring a specialized reaction set-up/apparatus, operating at room temperature with a shorter reaction period under mild conditions, providing good/excellent product yields, and offering generalized scope of substrates with electronic and structural variations resulting in diversity generation.

1. (A) Late-Stage Synthetic Modification of Compound 4l with 5a; (B) Late-Stage Synthetic Elaboration of Compound 3l with 5a; (C) “Gram-Scale” Production of Compound 4a .

The recoverability and reusability of a homogeneous catalyst from the reaction mixture are often considered difficult and pose a challenge. Yet, the recoverability/reusability of a homogeneous catalyst could have a notable impact on the sustainability evaluation of a chemical transformation from small laboratory to multi-kg/pilot plant scale through lowering the environmental factor (E-factor) of the process as far as possible. The catalytic CF₃SO₃H (10 mol %) in EtOH has been reused for three consecutive cycles (Table ). A minor deviation of the isolated yield of 4a and 3a in the first recycle was observed from the fresh run (entry 1 vs entry 2). The second recycle provided a larger deviation of the isolated yield of both 4a and 3a (entry 3). However, at the third recycle, the yield of 4a reduced drastically to 51%, and that of 3a improved to 33% (entry 4). The results could be due to the loss of homogeneous catalytic CF₃SO₃H at each successive recycling stage. The stability of reused catalytic CF₃SO₃H has further been confirmed by comparing its ¹⁹F NMR spectrum after each cycle with an authentic pure sample of CF₃SO₃H, which is in good agreement. The in-built precipitation of a crude mixture of products (3 and 4) from the reaction medium and the parallel recoverability of homogeneous catalytic CF₃SO₃H in EtOH are expected to allow a continuous mode of operation compared to different batchwise phases applicable to nonrecovered catalysts.

4. Recyclability Study of Catalytic CF₃SO₃H Synthesizing 4a and 3a from 1a and 2a .

		yield (%)
entry	run	4a	3a
1	fresh	88	8
2	first recycle	81	9
3	second recycle	70	14
4	third recycle	51	33

^aReaction condition: 1a (1 mmol), 2a (1 mmol, 1 equiv), CF₃SO₃H (10 mol %), EtOH (2 mL), rt (28 °C, oil bath), 2 h.

^bIsolated yield of 4a and 3a.

A logical mechanistic pathway for CF₃SO₃H-catalyzed formation of N-fused polycyclic 4 is depicted in Scheme . Initially, the H⁺ ion protonates the O atom at the C3-site of 1a (species I) and electrophilically activates the C3-position, which is nucleophilically attacked by lone pair electrons from the N-atom of 2a and forms hemiaminal-type species II. The loss of a water molecule and subsequent intramolecular N-cyclization by another N-atom of 2a in species II generates spiro-cyclic species III and releases a H⁺ ion. The H⁺ ion further electrophilically activates the C2-center through protonation of the O atom (species IV). Second, intramolecular N-cyclization via a nucleophilic attack by the adjacent N-atom to the C2-center forms a strained [5,3,5]-membered fused aziridine-type ring system (species V) and releases a H⁺ ion. The strained aziridine ring in species V opens up following a nucleophilic back push by the appended –OH group, and relocation of the H⁺ ion from the O atom to the C3 atom forms ring elongation species VI, which, further on, upon aerial oxidation, delivers long π-conjugated desired product 4a. The regenerated H⁺ ion during aziridine ring formation (species IV → V) could start another round of the catalytic process.

2. Postulated Mechanism for CF₃SO₃H-Catalyzed Synthesis of 4a .

Based on the prior literature report on imine formation in EtOH at room temperature from a condensation of a carbonyl unit with the primary amine group, we have proposed a reasonable mechanistic route (Scheme ) for the synthesis of 3a via condensation of 1a with 2a in EtOH. A dual “electrophilic-nucleophilic” activation pathway , involving hydrogen-bond (H-bond) donor ability via the O–H group and H-bond acceptor ability via the O atom in EtOH could be anticipated. Mechanistically, the EtOH molecule electrophilically activates C3-carbonyl of 1a through H-bond [CO···H–O] formation and nucleophilically activates 2a by another H-bond [O···H–N] formation, akin to water-mediated H-bond-promoted dual activation of carbonyl and amine groups creating an imine functionality. The H-bond network brings 1a and 2a in proximity, facilitating intermolecular nucleophilic attack by the N-atom lone pair electron from 2a to the C3-carbonyl of 1a and forming hemiaminal species II. The loss of a water molecule from II, aided by the H-bond network of EtOH, generates the first imine species VII. An intramolecular N-cyclization by nucleophilic attack from lone pair electrons of the second amino unit to electrophilically activated C2-carbonyl in intermittently formed VII, guided by H-bond assembly of the EtOH molecule, forms the second hemiaminal-type species VIII, which, on loss of another water molecule, assisted by H-bonding assembly from EtOH, delivers 3a. Unlike the spiro-cyclic species III generated from species II under CF₃SO₃H-catalyzed conditions in Scheme , resulting in 4a as the major downstream isomer, the neutral condition in the EtOH medium generated intermittent imine species VII in Scheme and signifies a difference in the mechanistic route, ionic in Scheme vs the non-ionic pathway in Scheme . Most likely, acid-catalyzed dehydration in species II (Scheme ) generates a distinct localized positive charge at the C3-center, which is quickly nucleophilically attacked by a second amine group in an intramolecular fashion, forming species III. However, under the non-ionic H-bond-assisted pathway in only EtOH solvent (Scheme ), distinct localized positive charge creation on the C3-center during dehydration is highly unlikely, which eventually generates imine species VII from species II and further results in exclusively condensation product 3a in a downstream channel.

3. Proposed Mechanism for EtOH-Mediated Synthesis of 3a .

Conclusion

In conclusion, we have developed eco-friendly, metal-free, chemodivergent, facile synthetic routes for fused polycyclic indolo­[2,3-b]­quinoxaline and N-fused polycyclic benzo­[4,5]­imidazo­[1,2-c]­quinazolin-6­(5H)-one scaffolds from common starting materials isatin and 1,2-diaminobenzene. The simple condensation between free-NH/N-substituted isatin with 1,2-diaminobenzene in EtOH at room temperature delivers exclusively indolo­[2,3-b]­quinoxalines, and catalytic protic acid (CF₃SO₃H) in EtOH at room temperature furnishes the ring expansion product benzo­[4,5]­imidazo­[1,2-c]­quinazolin-6­(5H)-one as a major isomer, along with a minor amount of indolo­[2,3-b]­quinoxaline. Thus, the selectivity switch between products of two different chemotypes could easily be modulated through the absence/presence of catalytic Brønsted acid, which typically follows two independent reaction pathways: (i) the non-ionic H-bond-assisted route for indolo­[2,3-b]­quinoxalines and (ii) the ionic mechanism for benzo­[4,5]­imidazo­[1,2-c]­quinazolin-6­(5H)-ones. While the indolo­[2,3-b]­quinoxaline formation proceeds by condensation between electrophilic carbonyl with nucleophilic amine units, a cascade reaction sequence of C3-spirocyclization/fused aziridine ring formation through C2–C3 bond/cleavage of the C2–C3 bond resulting in ring expansion with N-atom insertion and concomitant oxidation could reasonably be anticipated for the synthesis of benzo­[4,5]­imidazo­[1,2-c]­quinazolin-6­(5H)-one. The notable advantages of the protocols are as follows: (i) mild operation conditions at room temperature, (ii) use of environmentally benign EtOH as a solvent, (iii) the metal-free air/moisture stable organocatalytic (CF₃SO₃H) approach eliminating metal contamination of end products, (iv) a broad range of substrate scopes with structural plus electronic variations allowing easy diversification, (v) good to excellent product yields, (vi) simple experimental set-up, (vii) easy product purification by precipitation/filtration in the case of the condensation approach, (viii) reusability of catalytic CF₃SO₃H, and (ix) a high degree of chemoselectivity through definitive cleavage of the C2–C3 bond over other labile bonds under protic acid catalysis. The chemical modifications of the free-NH unit containing indolo­[2,3-b]­quinoxaline and benzo­[4,5]­imidazo­[1,2-c]­quinazolin-6­(5H)-one further create structural complexity of final products to value-added molecules, and a “gram-scale” production of ring expansion product establishes possible practicality of the developed chemodivergent synthesis. The organocatalytic ring expansion and rearrangement presented herein could promote skeletal editing of common, readily available molecules to highly decorated and sophisticated materials under mild and efficient conditions. Moreover, the present catalytic chemodivergent synthesis would inspire the synthetic organic/medicinal chemistry fraternity to develop mild, robust, and efficient strategies for easy access to products of different chemotypes from common starting materials, offering the synthesis of small-molecule-focused libraries and expediting the timely evaluation of SAR/SPR.

Experimental Section

Typical Procedure for the Preparation of 6-Phenyl-6H-indolo­[2,3-b]­quinoxaline (3a)

A mixture of 1-phenylindoline-2,3-dione (1a) (224 mg, 1 mmol, 1 equiv) and 1,2-diaminobenzene (2a) (109 mg, 1 mmol, 1 equiv) in EtOH (2 mL) was kept for stirring at room temperature (28 °C). The reaction progress was monitored using thin-layer chromatography (TLC). After the reaction was completed (8 h, TLC), a solid precipitate formed, which was filtered off from the EtOH solution, and the residue was washed with cold EtOH (2 × 1 mL). The crude product was re-crystallized in EtOH and dried in a vacuum oven at 50 °C overnight to afford 3a as a yellow solid (200 mg, 68%); mp: 236–238 °C (reported: 238–239 °C); ¹H NMR (600 MHz, CDCl₃): δ (ppm): 8.55 (dd, J = 7.7, 1.7 Hz, 1H), 8.33 (dd, J = 8.0, 1.6 Hz, 1H), 8.10 (dd, J = 8.1, 1.6 Hz, 1H), 7.76–7.70 (m, 4H), 7.69–7.63 (m, 3H), 7.55–7.50 (m, 2H), 7.46–7.43 (m, 1H); ¹³C NMR (101 MHz, CDCl₃): δ (ppm): 146.0, 144.9, 140.7, 140.3, 140.0, 135.5, 131.2, 129.9, 129.4, 129.0, 128.4, 128.1, 127.3, 126.7, 122.8, 122.0, 120.0, 110.8; IR (ATR) ν_max: 2907, 1574, 1241, 1111 cm^–1.

Typical Procedure for the Preparation of 5-Phenylbenzo­[4,5]­imidazo­[1,2-c]­quinazolin-6­(5H)-one (4a)

To a mixture of 1-phenylindoline-2,3-dione (1a) (224 mg, 1 mmol, 1 equiv) and 1,2-diaminobenzene (2a) (109 mg, 1 mmol, 1 equiv) in EtOH (2 mL), CF₃SO₃H (8.8 μL, 0.1 mmol, 10 mol %) was added and kept for stirring at room temperature (28 °C). The reaction progress was monitored using TLC. After the reaction was completed (2 h, TLC), a solid precipitate formed (a crude mixture of 3a and 4a), which was filtered off from the EtOH solution. The residue was washed with cold EtOH (1 × 1 mL). An attempt to separately recrystallize 3a and 4a from the crude mixture in EtOH failed, which directed us to pass the mixture through a chromatography column (silica gel, 100–200 mesh) and elute with hexane/EtOAc to afford 4a as a white solid (273 mg, 88%) along with 3a as a yellow solid (23 mg, 8%). Compound 4a, mp: 292–294 °C (reported: 295–296 °C); TLC (hexane/EtOAc, 80:20 v/v): R _f ≈ 0.5; ¹H NMR (400 MHz, CDCl₃): δ (ppm): 8.60 (dd, J = 7.5, 1.2 Hz, 1H), 8.44 (dd, J = 8.1, 1.2 Hz, 1H), 7.95 (dd, J = 8.0, 1.1 Hz, 1H), 7.69–7.59 (m, 3H), 7.54–7.50 (m, 1H), 7.47–7.42 (m, 4H), 7.40–7.36 (m, 1H), 6.72 (dd, J = 8.1, 1.2 Hz, 1H); ¹³C NMR (101 MHz, CDCl₃): δ (ppm): 147.1, 147.0, 144.1, 139.4, 136.3, 132.0, 131.4, 130.7, 129.8, 129.4, 125.8, 125.6, 124.6, 124.2, 119.6, 116.5, 115.6, 113.2; IR (ATR) ν_max: 2914, 1678, 1581, 1247, 1111 cm^–1.

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

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsomega.5c10483.

• Various optimization data of several reaction parameters, general experimental procedure, thorough characterization of synthesized compounds, and copies of ¹H, ¹³C, ¹⁹F NMR, and HRMS (PDF)

• Compound 4a data (CIF)

• CheckCIF data of compound 4a (PDF)

• Compound 3c′ data (CIF)

• CheckCIF data of compound 3c′ (PDF)

The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript.

The authors declare no competing financial interest.