Shailendra Singh Choudharyab,
Ratanamala S. Daroleab,
Gamidi Rama Krishnaa and
Beeran Senthilkumar
*ab
aDivision of Organic Chemistry, CSIR-National Chemical Laboratory, Dr Homi Bhabha Road, Pune-411008, India. E-mail: b.senthilkumar.ncl@csir.res.in
bAcademy of Scientific and Innovative Research (AcSIR), Ghaziabad 201002, India
First published on 9th September 2025
This study introduces a one-pot, transition metal-free strategy for synthesizing deoxybenzoins, overcoming the challenges of conventional methods. The reaction involves dual acylation of γ-aryl β-keto esters using K2CO3 in dioxane at 90 °C, followed by a concerted transformation to form deoxybenzoin. The protocol operates under mild conditions, tolerates a broad range of substrates, and produces minimal by-products, making it a practical, scalable alternative for accessing pharmaceutically relevant deoxybenzoins.
Additionally, deoxybenzoin serves as a key precursor for synthesizing various industrially and biologically significant compounds such as isoflavones, agrochemicals, dyes, herbicides, and fire-resistant polymers.3 It is also a versatile structural moiety for the construction of heterocyclic compounds like pyrimidinones, quinoxalines, and indoles.4 However, their synthetic flexibility and capability to undergo facile transformation to a wide range of functional groups made them valuable building blocks in organic synthesis (Fig. 2).5
The conventional method for the synthesis of deoxybenzoins involves a Friedel–Crafts reaction and requires the prefunctionalization of phenylacetic acid to phenylacetyl chloride using a harsh chlorinating agent, hence affecting process efficiency (Scheme 1a).6 To overcome these challenges, various strategies have been developed, such as employing substrates with internal multifunctional groups (e.g., diazonium, sulfonyl hydrazone, and phosphorus ylide) (Scheme 1b)7 or using metal reagents like RMgBr and R3Al.8 These approaches reduced the multistep operations but did not depend on the external metal catalysts, additives, and prefunctionalized substrates.
Among advanced methodologies, metal-catalyzed hydration of alkynes and Brønsted acid or photocatalyzed alkyne hydration has shown a promising approach (Scheme 1c).9 Similarly, the Wacker process has proven valuable for synthesizing ketones from terminal or 1,2-disubstituted alkenes via olefin oxidation.10 Other modifications, such as electrochemical oxidation,11 metal-catalyzed oxidation,12 and the use of molecular oxygen as an oxidant,13 have further enhanced the synthetic utility of deoxybenzoins.
Recent investigations have introduced several innovative methodologies for deoxybenzoin synthesis. These include alkyne hydration catalyzed by N-heterocyclic carbene-based gold complexes,14 the reaction of non-stabilized diazoalkanes with aldehydes in the presence of Sc(OTf)3,15 and the integration of photoredox catalysis with carbene chemistry to achieve reductive single-electron alkylation of acyl azoliums, enabling ketone synthesis from carboxylic acids (Scheme 1d).16 Additional strategies have involved the benzylic arylation of toluene mediated by LiN(SiMe3)2 with CsF as an additive, as well as transformations involving N-methoxy-N-methylbenzamide and methyl benzoate derivatives (Scheme 1e).17 Furthermore, metal-catalyzed Heck coupling reactions have also been reported for deoxybenzoin synthesis (Scheme 1f).18 Despite these advancements, challenges such as high reaction temperatures, reliance on metal catalysts, functional group incompatibility, and the need for supplementary acid catalysts still persist. Direct, simple, and cost-effective synthetic methods for deoxybenzoin remain scarce in the literature.
To address this gap, we report a new synthesis method for deoxybenzoin derivatives through a base-mediated dual acylation reaction of γ-aryl β-keto ester. This methodology involves the addition under mild conditions, simplifying the synthetic pathway and avoiding the need for metals, additives, and ligands. The reaction proceeds via the nucleophilic addition–elimination mechanism, where the base effectively facilitates the concerted reaction, resulting in the formation of the desired deoxybenzoin derivatives. The mild reaction conditions minimize side reactions and ensure compatibility with a wide range of functional groups, making the process versatile and scalable. Additionally, the cost-effectiveness of this approach stems from its reliance on readily available starting materials and the elimination of expensive catalysts or reagents. This innovative strategy provides a significant advancement in the field by addressing the limitations of existing methods and offering an environment-friendly alternative for the efficient synthesis of deoxybenzoin.
Entry | Base | Solvent | T (°C) | Time (h) | Yieldb % | |
---|---|---|---|---|---|---|
(3a) | (4a) | |||||
a Reaction conditions: For (entries 1–7) 1a (0.4062 mmol, 1 equiv.) and 2a (0.4062 mmol, 1 equiv.), K2CO3 (0.4062 mmol, 1 equiv.) in dioxane solvent (2 ml), for (entries 8–10) 1a (0.4062 mmol, 1 equiv.) and 2a (0.6093 mmol, 1.5 equiv.), K2CO3 (0.6093 mmol, 1.5 equiv.) in dioxane solvent (2 ml), for (entries 11–17) 1a (0.4062 mmol, 1 equiv.) and 2a (0.8124 mmol, 2 equiv.), K2CO3 (0.8124 mmol, 2 equiv.) in dioxane solvent (2 ml).b Isolated yield.c 1,4-diazabicyclo[2.2.2]octane.d 1,4 – dioxane.e Room temperature.f 8-diazabicyclo[5.4.0]undec-7-ene. | ||||||
1 | DABCOc | Dioxaned | Rte | 12 | NR | NR |
2 | DBUf | Dioxane | Rt | 12 | NR | NR |
3 | NEt3 | Dioxane | Rt | 12 | 24 | 18 |
4 | K3PO4 | Dioxane | Rt | 12 | 25 | 22 |
5 | Cs2CO3 | Dioxane | Rt | 12 | 20 | 15 |
6 | NaOH | Dioxane | Rt | 12 | 0 | 0 |
7 | NaOMe | Dioxane | Rt | 12 | 0 | 0 |
8 | K2CO3 | Dioxane | Rt | 12 | 30 | 25 |
9 | K2CO3 | Dioxane | 55 | 12 | 35 | 30 |
10 | K2CO3 | Dioxane | 75 | 12 | 44 | 40 |
11 | K2CO3 | Dioxane | 90 | 6 | 82 | 15 |
12 | K2CO3 | Dioxane | 90 | 3 | 50 | 20 |
13 | K2CO3 | Xylene | 90 | 6 | 50 | 35 |
14 | K2CO3 | Toluene | 90 | 6 | 40 | 50 |
15 | K2CO3 | CH3CN | 90 | 6 | 26 | 28 |
16 | K2CO3 | Acetone | 90 | 6 | 32 | 18 |
17 | K2CO3 | Dioxane | 90 | 24 | 82 | 15 |
18 | NEt3 | Dioxane | 60 | 12 | 28 | 14 |
19 | NEt3 | Dioxane | 80 | 12 | 30 | 15 |
20 | NEt3 | Dioxane | 90 | 12 | 30 | 15 |
21 | NEt3 | Dioxane | 100 | 12 | 31 | 15 |
22 | K3PO4 | Dioxane | 60 | 12 | 34 | 16 |
23 | K3PO4 | Dioxane | 80 | 12 | 55 | 20 |
24 | K3PO4 | Dioxane | 90 | 12 | 60 | 18 |
25 | K3PO4 | Dioxane | 100 | 12 | 65 | 15 |
26 | K2CO3 | Dioxane | 100 | 12 | 82 | 15 |
No reaction was observed when 8-Diazabicyclo[5.4.0]undec-7-ene (DBU) and 1,4-diazabicyclo[2.2.2]octane (DABCO) had been used (Table 1, entries 1 and 2). However, in the presence of triethylamine in dioxane at room temperature for 12 h, the corresponding products 3a and 4a were observed to yield 24% and 18%, respectively (Table 1, entry 3). K3PO4, Cs2CO3 yielded the desired products with low to moderate yields (Table 1, entries 4 to 5). No product formation was observed using strong bases such as NaOH and NaOMe (Table 1, entries 6 to 7). Surprisingly, K2CO3 had proved to be the most effective inorganic base in dioxane solvent compared to the other bases. Moreover, increasing the temperature, the yields of products 3a and 4a have increased (as shown in Table 1, entries 9 to 11). Therefore, the reaction conditions were further optimized at 90 °C. Finally, it was found that using 2 equiv. of benzoyl chloride (2a) and 2 equiv. of K2CO3 is sufficient to generate the desired product 3a in 82% and the minor product 4a in 15% isolated yield at 90 °C for 6 h (Table 1, entry 11), representing our best result. Common organic non-polar and polar aprotic solvents such as toluene, xylene, DMF, DMSO, MeCN, and acetone resulted in lower yields (Table 1, entry 14 to 16) and dioxane was found to be the best solvent for our desired product formation. The stability of product 3a was examined under prolonged reaction times, revealing no impact on its stability under the reaction conditions (Table 1, entry 17).
We evaluated K2CO3, K3PO4, and NEt3 at 60, 80, 90, and 100 °C, with reactions at 100 °C conducted in sealed vials to prevent solvent loss. The results indicate that: (i) K3PO4 showed enhanced conversion at elevated temperatures, achieving a maximum isolated yield of 65% at 100 °C (Table 1, entry 25). (ii) NEt3 remained ineffective above room temperature, likely due to its lower basicity and limited solubility in dioxane (Table 1, entry 18 to 21) and (iii) K2CO3 displayed performance consistent with previous observations (Table 1, entry 26) with an optimum yield obtained at 90 °C. Overall, these results reveal a modest temperature-dependent base effect and support K2CO3 at 90 °C as the optimal reaction condition.
With the optimized reaction conditions established, we investigated the substrate scope, focusing on methyl 3-oxo-4-(2,4,5-trifluorophenyl)butanoate (Table 2). γ-aryl-β-ketoesters reacted with ortho-, meta-, and para-fluoro-substituted benzoyl chlorides (2b–2d), yielded the corresponding products 3b–3d in 74–77% yields. Similarly, 3,5-difluorobenzoyl chloride (2e) produced the product 1-(3,5-difluorophenyl)-2-(2,4,5-trifluorophenyl)ethan-1-one (3e) in 75% yield. Halo-substituted derivatives of 1-phenyl-2-(2,4,5-trifluorophenyl)ethan-1-one (1a), where chloro- and bromo-containing benzoyl chlorides (2f–2i) were employed and delivered the desired deoxybenzoin derivatives 3f–3i in good yields. Furthermore, benzoyl chlorides bearing electron-withdrawing groups, such as 4-CH2Cl-, 4-CF3-, 3,5-diCF3- and 4-NO2- containing benzoyl chlorides (2j–2m) produced our desired products 3j–3m with very good yields. In addition, benzoyl chlorides containing electron-donating groups, including 4-OMe, 4-Me, and 3,5-DiMe (2n–2p), furnished the products 3n–3p in 68–71% yields, respectively. Interestingly, the electronic effect of the substituents did not significantly impact the reaction outcomes (Table 2, entries 3b–3p). Additionally, furan-2-carbonyl chloride (2q) provided an excellent yield of 80%, while using cyclohexanecarbonyl chloride (2r) and cyclopropanecarbonyl chloride (2s) resulted in moderate yields of 66 and 55%, respectively.
During our investigation, we observed that certain benzoyl chlorides showed no reactivity toward the formation of the desired deoxybenzoin (Fig. 3). For instance, electron-rich derivatives such as 2u and 2v, as well as the electron-deficient 4-cyanobenzoyl chloride (2w), and heteroaromatic acyl chlorides including isonicotinoyl chloride hydrochloride (2x) and nicotinoyl chloride hydrochloride (2y), failed to yield the target products. Similarly, aryl substrates bearing electron-donating groups, such as 1j and 1k, were also unproductive under the optimized conditions (Fig. 3).
Further, we investigated the reaction of various γ-aryl-β-ketoesters with differently substituted benzoyl chlorides under optimized conditions (Table 3). When methyl 3-oxo-4-phenylbutanoate (1b) was treated with various benzoyl chlorides (2a, 2e, 2i, 2p), yielding the corresponding products 3t–3w in 65 to 67%, respectively. These results demonstrated the broad compatibility of the reaction with benzoyl chlorides bearing diverse substituents. Similarly, methyl 4-(4-fluorophenyl)-3-oxo butanoate (1c) was reacted with 3,4-dimethoxybenzoyl chloride (2t), furan-2-carbonyl chloride (2q), and cyclohexanecarbonyl chloride (2r), affording products 3x, 3y, and 3z in yields of 67, 73, and 55%, respectively. These results highlighted the ability of the reaction to tolerate both aromatic and aliphatic acylating agents. Additionally, methyl 4-(2-chlorophenyl)-3-oxo butanoate (1d) was tested with 3,5-dimethylbenzoyl chloride (2p), yielded the corresponding product 3aa′ in 60%, showcasing compatibility with electron-donating substituents on the benzoyl chloride (Table 3). However, methyl 4-(4-chlorophenyl)-3-oxobutanoate (1e) exhibited excellent reactivity with a range of benzoyl chlorides (2a) including 4-Me (2o), 2,3-dichloro (2g), 3,5-dichloro (2h), and 4-CH2Cl (2j) substituents, producing the corresponding products 3ab′ to 3af′ with very good yields. This range of reactivity emphasized the adaptability of the substrate to various functional groups.
Further to our delight, various benzoyl chlorides (2a, 2d, 2h, 2k, 2l, 2m) also participated with methyl 4-(3,4-dichlorophenyl)-3-oxobutanoate (1f) in the reactions and provided the desired deoxybenzoins 3ag′–3al′ in 59–68% yield (Table 3). These reactions demonstrated the method's efficiency in accommodating electron-withdrawing groups on the benzoyl chloride. Moreover, methyl 4-(3-bromophenyl)-3-oxo butanoate (1g) reacted with cyclopropanecarbonyl chloride (2s) to yield product 3am′ in 56%, indicating moderate compatibility with alicyclic carbonyl chloride. Similarly, Methyl 4-(4-bromophenyl)-3-oxobutanoate (1h) successfully reacted with 2a, 2c, and 2d to afford products 3an′ to 3ap′ with moderate to good yields. Finally, methyl 4-(4-nitrophenyl)-3-oxobutanoate (1i) reacted with benzoyl chloride (2a) to produce 2-(4-nitrophenyl)-1-phenylethan-1-one (3aq′) with the yield of 81% (Table 3). These results further underscored the reaction's tolerance to halogen-substituted benzoyl chlorides and its potential for generating high yields with substrates bearing diverse functional groups.
To evaluate the synthetic practicality of this one-pot reaction, we performed a gram–scale reaction (Scheme 2) using 1a and 2a on an 8.12 mmol scale (Scheme 2). The reaction successfully yielded 3a with an isolated yield of 78%. This result confirms the method's reproducibility on a larger scale, offering a feasible protocol for potential industrial applications.
To gain a deeper understanding of the reaction mechanism, a series of control experiments were performed using substrates 1a and 2a under different temperatures and reagent stoichiometries. When 1a was treated with an equimolar amount of 2a and K2CO3 at room temperature, intermediate II was formed in 40% yield and intermediate III in 20% yield within 10 minutes, along with unreacted 1a, as confirmed by NMR and mass spectrometry (see SI). Extending the reaction time to 2 hours under the same conditions resulted in a decrease in intermediate II and a corresponding increase in intermediate III, suggesting the transformation of II into III (Scheme 3). Further continuation of the reaction for 6 hours using 2 equivalents each of 2a and K2CO3 resulted in the formation of the desired products 3a and 4a, along with decreased amounts of intermediates II and III, and trace amount of the intermediate VI, as confirmed by HRMS. When the same reaction was carried out at 50 °C for 6 h, an increased yield of 3a and 4a was observed (Scheme 3).
The proposed reaction mechanism as shown in Scheme 4, was supported by our control experiments and relevant literature precedents. Initially, K2CO3 deprotonates the α-carbon of the β-keto ester, generating enolate I. This enolate undergoes a base-promoted nucleophilic acyl substitution with the benzoyl chloride (2a), resulting in the formation of intermediate II. Intermediate II was successfully isolated and characterized by NMR and HRMS (see SI). In the presence of excess K2CO3, a second deprotonation at the α-position occurs, enabling a subsequent nucleophilic attack on a second equivalent of benzoyl chloride to form intermediate III. Another deprotonation of intermediate III generates a carbanion, which undergoes an intramolecular nucleophilic attack on a neighbouring carbonyl group, leading to the formation of intermediate IV. This species then undergoes a concerted transformation to afford the thermodynamically stable compound V, along with a trace amount of species VI. Protonation of compound V furnishes the major product, deoxybenzoin derivative 3. In contrast, intermediate VI undergoes in situ hydrolysis19 followed by thermal decarboxylation,20 giving rise to the minor β-ketoester product 4.
CCDC 2284931–2284933 contain the supplementary crystallographic data for this paper.23
Supplementary information: Full experimental details, copies of spectral data, structures, X-ray crystal structures, and crystallographic data (CIF). See DOI: https://doi.org/10.1039/d5ra05242d.
This journal is © The Royal Society of Chemistry 2025 |