Halogen-Guided Reconstructive Transformation of Hydroxycoumarin to Coumaranone

Abstract

Halogenation-driven skeletal transformations play an important role in the biosynthesis of natural products, where halogenation can promote C–C bond cleavage under mild conditions. Drawing inspiration from these biosynthetic processes, the halogenative skeletal reconstruction of hydroxycoumarins is reported, which unexpectedly facilitates carbonyl deletion to yield valuable heterocyclic scaffolds. Notably, cleavage of both the C–C and C–O bonds occurs under the mildest nonenzymatic conditions reported to date (low reaction temperature and near-neutral reaction media), while also demonstrating broad substrate scope. This transformation converts readily accessible hydroxycoumarins into biologically relevant coumaranone scaffolds via a halogen-identity-dependent decarbonylative reconstruction. This work establishes halogenation-enabled carbonyl deletion as a useful strategy for skeletal transformation and provides an efficient approach to molecular scaffold editing relevant to medicinal chemistry.

Ketone and ester functionalities represent chemically and thermodynamically stable structural motifs that are commonly found in pharmaceuticals, , organic electronics, and polymeric materials, , including the widely used polyethylene terephthalate (PET) and polyether ether ketone (PEEK). Owing to their high intrinsic stability, cleavage of these carbonyl bonds has traditionally required activation under strongly basic , or acidic , conditions. More recently, transition-metal catalysis has emerged as a viable approach for promoting such bond scission.

Recent advancements have been reported in methodologies that enable decarbonylative bond cleavage followed by functionalization. − These approaches generally involve an initial decarbonylation step, which generates organometallic, radical, or anionic intermediates that subsequently undergo nucleophilic functionalization (Figure a). Consequently, these transformations tend to depend heavily on transition-metal (TM) catalysis to facilitate bond activation and intermediate formation. Notably, no reports describe these processes occurring under both mild, near-neutral conditions at ambient temperature and in the absence of transition-metal catalysis. Specifically, ester decarbonylation, which involves the cleavage of both C–C and C–O bonds, remains particularly challenging (Figure b). For instance, Yamaguchi et al. reported that Ni- or Pd-catalyzed decarbonylative ether synthesis required elevated temperatures (150 °C) and extended reaction times (18 h). Subsequently, Luu et al. demonstrated a Ni-catalyzed decarbonylative coupling reaction of cyclic lactones, which required heating at 150 °C for 24 h. They also highlighted the need for systematic catalyst screening for individual substrates.

1.

Decarbonylation-triggered functionalization strategy.

More recently, Pu et al. reported a Cu-catalyzed conversion of coumarins into benzofuran (120 °C, 20 h), which proceeded via oxidative ring opening followed by decarboxylative O-arylation.

In contrast, enzymes typically enable decarbonylation under notably mild conditions, generally at ambient temperature and near-neutral pH. − Specifically, halogenases facilitate the introduction of halogen atoms into substrates, which can potentially induce decarbonylation through spontaneous bond cleavage, distinct from decarbonylation-triggered functionalization strategies (Figure a). This halogenative bond cleavage strategy is fundamental in the sophisticated biosynthesis of structurally diverse bioactive natural products. For instance, the biosynthesis of cytotoxic dichlorodiaporthin and the antibiotic chloramphenicol involve gem-dichlorination catalyzed by flavin-dependent halogenases (FDHs), constituting a fundamental step that enables subsequent decarbonylation cleavage (Figure b). , Specifically, the produced gem-dichlorinated 1,3-diketones undergo hydrolytic carbonyl C–C bond cleavage, leading to acetyl group elimination. FDHs also play a role in the biosynthesis of tetrabromopyrrole and polybrominated diphenyl ethers (PBDEs), as well as pentabromopseudilin, a marine antibiotic (Figure c). Bromination at the ipso position of the aromatic carboxylic acid has been proposed to trigger decarboxylation. These examples highlight halogen incorporation as a strategic promoter of carbonyl bond cleavage under extremely mild conditions, typically at ambient temperature and near-neutral pH. Consequently, halogenation-initiated bond cleavage has emerged as a promising strategy; however, despite its conceptual appeal and synthetic potential, this approach remains underdeveloped. −

2.

Synthetic pathways toward bioactive compounds featuring halogenation-triggered C–C bond cleavage.

One of the main research focuses of our group is the development of a new heterocyclic synthetic strategy termed “cut-to-fuse”. This innovative approach involves halogenation-triggered bond cleavage of a homocyclic compound (the “cut” step) to generate the corresponding chain, followed by reconstructive cyclization (the “fuse” step) to yield the desired heterocyclic compound (Figure a). It was hypothesized that the incorporation of halogens could facilitate dual bond scission within cyclic ester structures, subsequently leading to intramolecular cyclization and thereby achieving carbonyl deletion to form new heterocycles.

3.

Skeletal editing: carbonyl deletion from β-keto-esters.

Skeletal editing has garnered significant attention as a powerful strategy for modifying molecular frameworks and exploring underexplored chemical spaces; , however, achieving precise control over the transformation of skeletal structures within heterocyclic systems is both highly desirable and challenging. − Among the skeletal editing methodologies, carbonyl deletiondefined as the net removal of a CO unit from a cyclic carbonyl frameworkhas emerged as a highly effective strategy for the ring contraction of ketones, amides, esters, and thioesters. However, ester-to-ether carbonyl deletion remains challenging and a practical method has never been reported.

This paper describes a serendipitously discovered skeletal reconstruction of hydroxycoumarins into coumaranones (Figure b). Although previously reported ester bond cleavage in coumarin frameworks has conventionally relied on strongly basic conditions and elevated temperatures (e.g., 10% NaOH aq., 140 °C and 20% NaOH aq., 160 °C), , halogen incorporation enables dual bond scission within ester functionalities under physiological temperature and near-neutral conditions, representing an exceptionally mild nonenzymatic platform. This transformation provides rapid access to coumaranonesprivileged scaffolds in bioactive and medicinally relevant compounds − from readily available hydroxycoumarins.

Building on previous research by our group in 2024, this work focused on the development of a method for the fluorinative C–C bond cleavage of hydroxycoumarin (Figure a). A preliminary study revealed that hydroxycoumarin 1aa′ produced methyl salicylate 2aa′ and fluorinated methyl phenylacetate 2ab′ in moderate yields, suggesting that the reaction proceeded via a known fragmentation pathway, as documented previously. , The initial strategy involved cleaving the C–C bond to enable reconstruction into heterocycles; however, the substrate underwent fragmentation into two separate compounds, rendering the molecular transformation challenging. Subsequently, the effects of employing a less electron-withdrawing chlorine atom and a less nucleophilic solvent were examined. Upon reaction of hydroxycoumarin 1aa with N-chlorosuccinimide (NCS) in a mixed solvent of acetone and water in the presence of K₃PO₄, a unique transformation specific to chlorine was observed, with no fragmentation pathway detected; instead, an unexpected decarbonylated coumaranone 3aa was obtained. Recognizing this unforeseen decarbonylative process as a distinct and potentially general molecular transformation, the reaction parameters were investigated in further detail. Accordingly, hydroxycoumarin 1aa was subjected to reaction with NCS under basic conditions, followed by treatment with an additional base. Using 2 equiv of K₂CO₃ or K₃PO₄, the desired product 3aa was obtained in low yield (Figure b, entries 1 and 2). Upon increasing the concentration of hydroxycoumarin 1aa from 0.1 to 0.2 M, coumaranone 3aa was produced in quantitative yield (Figure b, entries 3 and 4). The influence of reaction concentration and the quantity of base are summarized in Table S3. This realization of bond scission at room temperature under near-neutral, nonenzymatic conditions is particularly noteworthy.

4.

Reaction findings and optimization.

β-Keto esters are prevalent motifs in numerous natural products, with 4-hydroxycoumarins being particularly abundant and easily synthesized. , However, the molecular transformations of hydroxycoumarins remain underexplored, likely due to the necessity of cleaving robust bonds within the coumarin structure, and ultimately rendering such transformations challenging. ,,− The structural reconstruction of various 4-hydroxycoumarin derivatives 1 was therefore attempted using a chlorination-induced transformation under the optimized mild conditions, affording the corresponding products in good yield (Figure a). Specifically, the reaction of 4-hydroxycoumarin derivatives containing methoxy, halogen, and azide substituents on their benzylaryl rings proceeded smoothly, producing the corresponding 3-coumaranones 3ab–3ae in high yields. A pinacol ester was also well-tolerated in this reaction, effectively generating 3af without the requirement for subsequent purification by column chromatography. Importantly, these compounds could be subjected to further transformations, including transition-metal-catalyzed bond-forming reactions, if desired. Notably, 4-hydroxycoumarins bearing protic moietiesincluding free phenol and carboxylic acid groupswere transformed into their corresponding products (3ag, 3ah), although the latter was obtained in a modest yield. Hydroxycoumarin derivatives bearing meta-substituted benzyl groups were also tolerated (3ai, 3aj), and the reaction proceeded effectively in the presence of ortho-substituents on the benzyl ringincluding halogen (bromo), electron-withdrawing (cyano, nitro), and electron-donating (methyl, methoxy) groupsaffording the expected products 3ak–3ao in good yield. Furthermore, derivatives bearing naphthalene, heteroarene (pyridine, thiophene, furan, and 2-substituted aromatic rings), and aliphatic substituents, also afforded the corresponding 3-coumaranone derivatives (3ap–3au). However, the isopropyl-substituted substrate did not undergo the anticipated transformation, likely due to increased steric hindrance. The phenyl-substituted substrate 1aa′ produced the corresponding diketone compound, 1-(2-hydroxyphenyl)-2-phenylethane-1,2-dione, rather than the expected coumaranone (see Supporting Information). Subsequently, the effect of different substituents on the 4-hydroxycoumarin core was investigated, revealing that bromo and methoxy substituents at the 6-position (3ba, 3bb), methoxy substituents at the 7-position, and 2-substituted derivatives, were well-tolerated (3bc, 3bd). Thus, to further explore the synthetic utility of this method, several structurally diverse substrates were examined (Figure b). Using dicoumarol 1av as the substrate, benzofuroisocoumarin 3av was obtained in quantitative yield. Notably, substituted isocoumarins exhibit various biological properties, including anticancer, antiallergic, and antimicrobial activities. , The transformation was also performed using a conjugation-extended substrate, affording tricyclic compound 3be. When applied to pyran derivative 1bf, the reaction proceeded to yield furanone 3bf.

5.

Substrate scope of the decarbonylation. ^aReaction conditions: 1 (0.20 mmol), AcONa (0.20 mmol), NCS (0.20 mmol), and H₂O (1.0 equiv) in AcOEt (1.0 mL) at room temperature for 3 h. K₃PO₄ (0.60 mmol) was then added and the reaction continued for 4 h. ^bNMR Yield. ^c 1 Isolated as the methyl ester. ^dThe carboxylic acid was isolated as methyl ester 3ah′ with TMS-diazomethane in 37% yield (see Supporting Information). ^e1 (0.20 mmol), AcONa (0.20 mmol), NCS (0.20 mmol), and H₂O (1.0 equiv) in AcOEt (2.0 mL) at room temperature for 3 h. K₃PO₄ (0.80 mmol) was then added and the reaction continued for 4 h.

A control experiment, which was performed in the absence of a chlorinating reagent, resulted in complete recovery of the starting material, confirming the critical role of chlorination in this transformation (Figure a). Additionally, in an attempt to achieve a one-pot decarbonylative transformation, a mixture of mono- and dichlorinated products was observed (Table S1). Specifically, although the monochlorinated product 2aa was successfully converted to coumaranone 3aa, the dichlorinated product 2aa-Cl was not transformed into the desired product (Figure b). This demonstrates that 2aa-Cl is not a reaction intermediate and confirms that selective monochlorination is essential for cyclization. Subsequently, this molecular transformation was examined in a stepwise reaction sequence (Figure c).

6.

Mechanistic studies.

Initially, hydroxycoumarin 1aa was chlorinated with NCS to yield the corresponding chlorinated product 1aa-Cl in quantitative yield. Subsequent treatment of 1aa-Cl with sodium acetate afforded decarboxylated compound 2aa, which was further converted into coumaranone 3aa in the presence of K₂CO₃. To gain mechanistic insight into this transformation, proton nuclear magnetic resonance (¹H NMR) spectroscopy was employed, comparing the spectra recorded during this stepwise transformation with those of decarboxylative product 2aa and coumaranone 3aa (Figure S1). Upon changing the solvent to methanol, decarboxylation did not occur; instead, ester-containing coumaranone 3aa-COOCH₃ was obtained in 55% yield via chlorinated intermediate 2aa-COOCH₃ , further supporting the involvement of hydrolytic decarboxylation in the presence of water (Figure d). Radical trapping experiments employing TEMPO and dibutylhydroxytoluene did not indicate the involvement of a radical mechanism in the current reaction (Scheme S5).

Although the exact details of the mechanism have yet to be fully elucidated, it is proposed that the chlorination of hydroxycoumarin 1 generates intermediate A, facilitating decarbonylation to yield coumaranone 3; this step may occur through two conceivable reaction pathways (Figure ). In pathway A, chlorinated intermediate A undergoes hydrolysis, which induces intramolecular cyclization and decarboxylation to produce the coumaranone structure. The feasibility of this pathway cannot be dismissed, as the cyclization of chlorinated intermediate 2aa-COOCH₃ to coumaranone 3aa-COOCH₃ occurs smoothly under mildly basic conditions (control experiment, Figure d). Concurrently, this direct transformation may involve decarboxylation and chloride extrusion, generating a transient carbene species that facilitates ring closure. − Alternatively, in pathway B, nucleophilic attack of water on chlorinated intermediate A could promote hydrolysis, followed by concomitant decarboxylation to produce intermediate 2, which may then undergo intramolecular cyclization. Based on the current evidence, either of these routes could account for, two parallel routes, the direct pathway from intermediate A (pathway A) and decarboxylation followed by intramolecular cyclization (pathway B), are possible for the formation of coumaranone 3.

7.

Proposed reaction mechanism and intermediates.

To illustrate the potential application of this protocol in organic synthesis, the coumaranone products were subjected to various synthetic transformations. Initially, the reaction of 1aa was conducted on the gram scale, yielding the desired product 3aa in 91% yield (Figure a). In a subsequent derivatization, the treatment of coumaranone 3aa with NaBH₄, followed by acidification, afforded benzofuran 4 in 80% yield (Figure b). It is also known that coumaranones serve as valuable substrates in recent skeletal editing techniques, providing significant scope for structural modification. − Additionally, the α-carbon atom was converted to a quaternary carbon atom via methylation to generate 5 in 63% yield. Furthermore, a click reaction was performed to afford triazole-containing 3-coumaranone 3ae′ (Figure c). Moreover, palladium-catalyzed coupling gave the corresponding product 3af′ in moderate yield.

8.

Synthetic applications of the developed protocol.

Compared with existing methods, − this protocol is advantageous as it employs easily accessible coumaranone bearing boron pinacol esters (3af), which can be directly used in Pd-catalyzed coupling reactions without isolation. This enables rapid construction of diverse coumaranone frameworks through reaction with various nucleophiles.

An unprecedented molecular transformation was serendipitously identified, which was inspired by a unique biotransformation, in which halogen incorporation induces C–C bond cleavage adjacent to a carbonyl group. This method achieved carbonyl deletion through sequential C–C and C–O bond cleavage as a form of skeletal editing. The incorporation of halogen atoms overcame the longstanding challenge of dual C–C and C–O bond scission in esters under the mildest nonenzymatic conditions reported to date. Overall, this transformation offers an efficient synthetic route from 4-hydroxycoumarins to 3-coumaranones by exploiting the halogen effect. More broadly, the success of this reaction design offers potential applications in drug design, agrochemicals, and enzymology. In medicinal chemistry, there is a current emphasis on substituting traditional multistep sequences with single-step transformations. This challenge has been addressed through skeletal editing strategies, and the reaction discussed here conceptually aligns with this objective.

Glossary

Abbreviations

PET

polyethylene terephthalate

PEEK

polyether ether ketone

TM

transition-metal

FDH

flavin-dependent halogenase

PBDE

polybrominated diphenyl ether

NCS

N-chlorosuccinimide

¹H NMR

proton nuclear magnetic resonance.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/jacsau.6c00801.

• Experimental procedures, characterization data and spectroscopic data (PDF)

H.C. and T.D. conceived and designed the experiments and directed the project. Y.Y. and Y.M. performed the experiments. Y.Y., H.C. and M.F. analyzed the data and checked the experimental details. Y.Y., K.K., H.C. and T.D. wrote the paper. All authors have read and agreed to the published version of the manuscript.

The authors declare no competing financial interest.