A Copper Halide Promoted Regioselective Halogenation of Coumarins Using N-Halosuccinimide as Halide Source

Abstract

A safe, convenient, and regioselective synthesis of 3-halo coumarins using a metal halide (CuX₂ alone or with ZnX₂) promoted halogenation with N-halosuccinimide (NXS) as halide source is reported. The synthesis involved the steady in situ generation of highly reactive positive halogen (X⁺) by the coordination of copper or zinc with the N-halosuccinimide and subsequent electrophilic aromatic substitution of the electron-deficient coumarins. This procedure works well also for the halogenation of less electron-rich naphthoquinones, flavones, and methoxypsoralen in moderate to quantitative yields. This protocol features simple experimental conditions using readily available inexpensive reagents and provides a convenient approach to the chlorination or bromination of some useful heteroaromatic compounds.

Key words

coumarin - halogenation - aryl halide - electrophilic aromatic substitution - N-halosuccinimide

The halogenation of aromatic and heteroaromatic compounds is one of the most valuable synthetic transformations. The halogenated derivatives can be used as bioactive compounds and lead to new pharmaceuticals entities.[1] They also serve as important synthetic precursors for transition-metal-catalyzed cross-couplings.[2] Traditional halogenation protocols rely on the usage of toxic halogen X₂(Cl₂, Br₂) which could cause serious environmental problems. The replacement of X₂ by other safe reagents is ubiquitous to solve this problem,[3] among which, N-halosuccinimide (NXS, X = Cl, Br, I) is quite popular and practical, which, depending on the reaction conditions, can steadily release the halide ‘cation’ (X⁺) or radical (X ^• ) for the halogenation reaction. The N-bromosuccinimide (NBS) works well with some electron-rich arenes.[4] However, it is difficult to brominate the electron-poor arenes, and the chlorination with the less reactive N-chlorosuccinimide (NCS) is even more difficult. Therefore, the halogenations of rather electron-deficient heteroaromatics remain a formidable synthetic task. Herein, we wish to report the copper activation of NXS for both chlorination and bromination of coumarin derivatives.

Coumarin (2H-chromen-2-one) can act as a privileged medicinal scaffold, which exhibits a broad spectrum of biological activities[5] such as anticancer,[6] antioxidant,[7] anti-inflammatory,[8] and so on. Besides these pharmaceutical applications, some coumarin derivatives can be used as fluorescent labels for bioimaging[9] and solar cell sensitizers.[10] These useful chemicals can often be synthesized via the transformations of halogenated coumarins. Although halogenated coumarins can be accessed through the cyclization from halogen-containing starting materials,[11] the direct halogenation of coumarin itself is evidently a more straightforward and economic pathway that benefits from the abundance of commercial availability of the starting materials.

However, it is known that coumarins are usually difficult to be halogenated for their electron-deficient character. In the literature, halogenation of coumarins is often accomplished by some highly reactive reagents (Scheme [1], path a, b). For example, the chlorination of coumarins usually needs to use such reagents as HCl/mCPBA in DMF (path a)[12] or HCl with oxone (path b).[13] The bromination can also be accomplished using HBr/Oxone (path b).[8] [13] However, large excess of oxidant is needed to generate reactive X⁺ from HX, which limited the reaction scope to substrates carrying oxidant-tolerable functional groups, and multichlorinated side products are often formed under the reaction conditions.

Scheme 1 The halogenation of coumarins from the literature and this work

There are some reports on the bromination of coumarins with electron-donating groups such as OH or NH₂, using NBS together with Brønsted acid activators or in a highly polar solvent, such as NBS in the presence of sulfonic acid functionalized silica gel,[14] NH₄OAc,[15] in molten TBABr,[16] or in polyethylene glycol (Scheme [1], path c),[17] etc. The successful bromination can be attributed to both the activation of NBS and the electron-rich nature of the substrates. The lack of more efficient activation method excluded the further application of these reactions to inactive coumarins, not to mention the chlorination by the less active NCS.

Recently, there are a few reports on the Lewis acid activation of N-iodosuccinimide (NIS) for the iodination of electron-rich arenes. Sutherland and coworkers[18] reported a silver (AgNTf₂) catalyzed iodination; Romo’s group[19] and Frontier’s group[20] used indium(III) and gold(I) complexes, respectively, for the activation of NIS. However, the Lewis acid activation of NCS and chlorination of electron-deficient arenes is seldom reported to our knowledge. Inspired by this strategy, we set out to search for an inexpensive Lewis acid promoter to activate NCS or NBS for the halogenation of the less electron-rich coumarins.[21] Herein, we report a safe, convenient, and regioselective 3-halogenation protocol of coumarins promoted by CuX₂/ZnX₂ with the corresponding NXS as halide source.

In the beginning, control experiment for the chlorination reaction was conducted by heating coumarin (1a) with 5 equiv NCS in acetonitrile, but no reaction took place even after prolonged refluxing,[22] which indicated the NCS needed to be activated. Several Lewis acid promoters were evaluated for the chlorination of 1a using the standard protocol (Table S1 in the Supporting Information). To our delight, the chlorination of 1a could be successfully promoted by the Lewis acid catalysts investigated. Among them, the ­CuCl₂·2H₂O is the most efficient and affords the 3-chloro coumarin (2a) with 81% yield (entry 6, Table S1) based on consumed 1a with no other regioisomers or multichlorinated side products found. Other Lewis acid catalysts such as ZnCl₂, CdCl₂, or AlCl₃, etc. exhibited relatively lower catalytic activities for this chlorination. FeCl₃ showed a poor regio­selectivity for this reaction, which promotes the chlorination of 1a at both 3- and 4- positions with a ratio of 70:30. The reaction solvent also influenced the reaction: The yield of the desired product decreased when the chlorination was carried out in acetic acid, 1,4-dioxane, or DMF. We further optimized the dosage of NCS and CuCl₂·2H₂O promoter (­Table S2) to guarantee a complete conversion of 1a. The conversion increased when more NCS was used, and the ­CuCl₂·2H₂O amount should be maintained at 1.6–2.4 times (molar ratio) that of NCS for better yields (entries 1–5, Table S2). Finally, when a molar ratio of 1:7:14 (coumarin/NCS/CuCl₂) was used, 2a was obtained in 62% yield with full conversion of 1a (entry 7, Table S2). The CuCl₂·2H₂O can be half replaced with ZnCl₂ to complete the chlorination with similar yield of the 2a (entry 9, Table S2). Anhydrous CuCl₂ can promote chlorination with the same yield as the CuCl₂·2H₂O, but more than 99 hours reaction time is needed. The addition of extra water can accelerate the halogenation, however, the yield decreased to 46% (entries 12–14, Table S2).

With the optimized chlorination conditions in hand, we extended this reaction to a series of coumarins substituted by methyl, methoxy, hydroxyl, chlorine, bromine, acetoxy, and dimethylamino groups. This method can also be applied to the bromination of these substrates using NBS, which is promoted by CuBr₂. The results of these halogenation reactions are summarized in Table [1]. For detailed reaction conditions including the ratios of reactants, reaction temperature and time, see Tables S3 and S4 in the Supporting Information.

Table 1 The 3-Halogenation of Substituted Coumarins^a

Entry	Substrate	Product	Yield (%)
1  2			2a, X = Cl, 62b 3a, X = Br, 61c
3  4			2b, X = Cl, 45b 3b, X = Br, 49c
5  6			2c, X = Cl, 84b 3c, X = Br, NRc
7  8			2d, X = Cl, 71b 3d, X = Br, 77c
9 10			2e, X = Cl, NRb 3e, X = Br, NRc
11 12			2f, X = Cl, 74b 3f, X = Br, 77c
13 14			2g, X = Cl, 84b 3g, X = Br, 66c
15 16			2h, X = Cl, 95b 3h, X = Br, 99c
17 18			2i, X = Cl, 68b 3i, X = Br, 73c
19 20			2j, X = Cl, 76b 3j, X = Br, 85c
21 22			2k, X = Cl, 58b 3k, X = Br, 61c

^a Isolated yield. NR = no reaction.

^b Chlorination conditions: The materials were refluxing in 20 mL MeCN (ca. 82 ℃) until the coumarin is consumed. For 2a–e: A mixture of coumarin (1.0 mmol), NCS (7.0 mmol), CuCl₂·2H₂O (7.0 mmol), and ZnCl₂(7.0 mmol) is used; for 2f–j: coumarin (1.0 mmol), NCS (1.0 mmol), CuCl₂·2H₂O (2.0 mmol), and ZnCl₂(2.0 mmol) is used; for 2k: coumarin (1.0 mmol), NCS (1.0 mmol), and CuCl₂·2H₂O (1.6 mmol) are used and reacts at RT (room temperature).

^c Bromination conditions: The starting materials were refluxed in 20 mL MeCN (ca. 82 ℃) until the coumarin is consumed. For 3a–c: A mixture of coumarin (1.0 mmol), NBS (5.0 mmol), and CuBr₂ (10 mmol) is used; for 3d,e,h–j: Coumarin (1.0 mmol), NBS (2.0 mmol), and CuBr₂ (3.2 mmol) is used; for 3f,g: coumarin (1.0 mmol), NBS (1.5 mmol), and CuBr₂ (2.4 mmol) are used and reacts at RT, for 3k: coumarin (1.0 mmol), NBS (1.0 mmol), and CuBr₂ (1.6 mmol) are used and reacts at RT.

As illustrated in Table [1], the amount of NCS is varied from 1 to 7 equiv, depending on the electronic properties of the substituents: it costs only one or two equivalents of NCS to chlorinate the activated coumarins (2f–k), but 7 equiv of NCS are needed to complete the chlorination of the substrates with electron-withdrawing or weakly electron-donating groups (2a–d). Less amount of NBS (1–5 equiv) was needed to complete the bromination reaction. Some halogenation reactions can be achieved at room temperature (2k, 3f,g,k). Notably, the complete bromination of 1k to afford 3k can be achieved in less than one minute at room temperature, which otherwise would take 72 hours to complete when NH₄OAc was used as the activation reagent.[15] In regard to the halogenation yield, the halogenation of the active substrates gives good to excellent yields (2f–k, 3f–k), and even the less active coumarins gives the 3-halo coumarins (2a–d; 3a–d) in moderate to good yields (49–84%). In the case of 4-hydroxyl coumarin, no anticipated 3-halo products (2e, 3e) were found, possibly owing to the lower reactivity of the C=C double bond, caused by chelating of copper with 4-hydroxyl group, which decreases the electron density. The bromination of 4-chloro coumarin (1c) and the halogenation of 4-bromo coumarin gives no anticipated products, obviously due to the steric hindrance of the 4-halogen atom.

According to previous reports,[23] Lewis acid catalyzed halogenation reactions might proceed through either copper(III) mechanism or electrophilic aromatic substitution (EAS) mechanism. As described above, the reaction conditions and yields are greatly affected by the substituents: electron-donating substituents favor the halogenation, while electron-withdrawing substituents impede the halogenation, which is in agreement with an EAS mechanism. A control experiment shows the chlorination reaction under O₂ atmosphere or argon protection gave similar results, this further supports the EAS mechanism.[23c] A plausible mechanism is illustrated in Scheme [2] with the chlorination of coumarin (1a) by NCS/CuCl₂ as an example.

Scheme 2 The proposed halogenation mechanism

The chlorination proceeds via the highly reactive chlorine ‘cation’ (Cl⁺) generated in situ. Firstly, copper chlorides coordinate with the carbonyls of NCS (A). This complexation in A substantially increased the electron-withdrawing power of the carbonyls, leading to the heterolytic cleavage of the N–Cl bond to expel a chlorine cation (Cl⁺) as shown in B and C. The highly electrophilic Cl⁺ regioselectively added to the 3-position of the coumarin ring to give a more stable benzyl carbocation D. Subsequent proton transfer from D to the N atom in C yields 3-chloro coumarin 2a as the chlorinated product and succinimide, which is still coordinating with copper (NHS-Cu). This mechanism necessitates the use of two equivalents of copper chloride relative to NCS. Electron-donating substituents increase the electron density of the coumarin to facilitate the electrophilic attack of the positive halogen and render a higher stability of the carbocation intermediate D, thus leading to higher halogenation efficiency. As a result, the halogenation of 7-methoxy coumarin gave the 3-halocoumarins (Table [1], 2h, 3h) quantitively, while the coumarins with chlorine groups (2b,c, 3c) gave relative lower yields.

Table 2 The Halogenation of Heteroarenes^a

Entry	Substrate	Product	Yield (%)
1 2			5a X = Cl, 99b 6a X = Br, 98c
3 4			5b X = Cl, 69b 6b X = Br, 96c
5 6			5c X = Cl, 72b 6c X = Br, 86c
7 8			5d X = Cl, 98b 6d X = Br, 98c

^a Isolated yield.

^b Chlorination conditions: The materials were refluxing in 20 mL MeCN (ca. 82 ℃) until the heteroarene was consumed. For 5a,c: Heteroarene (1.0 mmol), NCS (7.0 mmol), and CuCl₂·2H₂O (14.0 mmol) are used. For 5b: Naphthoquinone (1.0 mmol), NCS (10.0 mmol), and ­CuCl₂·2H₂O (20.0 mmol) are used. For 5d: Methoxysalen (1.0 mmol), NCS (2.0 mmol), and CuCl₂·2H2O (4.0 mmol) are used and reacts at RT.

^cBromination conditions: For 6a,b: Naphthoquinone (1.0 mmol), NBS (5.0 mmol), and CuBr₂ (10 mmol) are used. For 6c,d: Heteroarene (1.0 mmol), NBS (1 mmol) and CuBr₂ (1.6 mmol) are used, and 6d is prepared at RT.

We also tested the halogenation of other heteroarenes using this copper activation strategy (Table [2]) to access their valuable halo derivatives. For example, the selective halogenation of the naphthoquinones would provide an important platform for further derivatization of this important quinone.[24] The halogenation of naphthoquinones (4a,b) can regioselectively afford the monohalogenated products (5a, 6a,b) in quantitative yields, except the chlorination of 2-methyl-1,4-naphthoquinone (5b, 69%). There was an early report of the bromination of 2-methyl flavone (4c) using NBS alone gave the methyl brominated product.[25] Interestingly, upon using copper activation, a selective 3-bromination takes place on the heteroaromatic ring (5c, 6c), which also supports the EAS mechanism. The halogenation of 8-methoxypsoralen (4d) also affords quantitative yields of monohalogenated products (5d, 6d).

In conclusion, we developed a CuX₂/ZnX₂ promoted regioselective halogenation of coumarins by NCS or NBS with moderate to excellent yields.[26] The NXS dosage depends on the electronic properties of the substituent groups and varies from 1–7 equiv of the substrates, and the metal halide needs to be used in an amount 1.6–2 times that of NXS. Due to the usage of inexpensive reagents, the reaction can be scaled up although excess NXS, and metal halides are often needed to complete the conversion for some relatively electron-deficient coumarins without electron-donating substituents or even with electron-withdrawing substituents. This methodology can be applied to other less electron-rich or rather electron-deficient heterocyclic compounds such as 1,4-naphthoquinones, flavone, and methoxypsoralen to give monohalogenated derivatives. Therefore, this protocol provides an alternative approach to the halogenation of heteroaromatic compounds using copper activation of NXS with safe workup. We believe this concise procedure may also be applied to the halogenation of other aromatic compounds.

• References and Notes

• 1a Vardhan Reddy KH, Brion J.-D, Messaoudi S, Alami M. J. Org. Chem. 2016; 81: 424
• 1b Luong TT. H, Brachet E, Brion J.-D, Messaoudi S, Alami M. Eur. J. Org. Chem. 2015; 2015: 1771
• 1c Bruneau A, Brion J.-D, Messaoudi S, Alami M. Org. Biomol. Chem. 2014; 12: 8533
• 1d Min M, Kim B, Hong S. Org. Biomol. Chem. 2012; 10: 2692
• 1e Nicolaou KC, Bulger PG, Sarlah D. Angew. Chem. Int. Ed. 2005; 44: 4442
• 2a Ruiz-Castillo P, Buchwald SL. Chem. Rev. 2016; 116: 12564
• 2b Sun C.-L, Shi Z.-J. Chem. Rev. 2014; 114: 9219
• 2c Weix D. J. Acc. Chem. Res. 2015; 48: 1767
• 2d Johansson Seechurn CC. C, Kitching MO, Colacot TJ, Snieckus V. Angew. Chem. Int. Ed. 2012; 51: 5062
• 3a Huang B, Zhao Y, Yang C, Gao Y, Xia W. Org. Lett. 2017; 19: 3799
• 3b Voskressensky LG, Golantsov NE, Maharramov AM. Synthesis 2016; 48: 615
• 3c Song S, Sun X, Li X, Yuan Y, Jiao N. Org. Lett. 2015; 17: 2886
• 3d Schmidt R, Stolle A, Ondruschka B. Green Chem. 2012; 14: 1673
• 4a Tang R.-J, Milcent T, Crousse B. J. Org. Chem. 2018; 83: 930
• 4b Zysman-Colman E, Arias K, Siegel JS. Can. J. Chem. 2009; 87: 440
• 5 Medina FG, Marrero JG, Macias-Alonso M, Gonzalez MC, Cordova-Guerrero I, Teissier Garcia AG, Osegueda-Robles S. Nat. Prod. Rep. 2015; 32: 1472
• 6 Harada K, Kubo H, Tomigahara Y, Nishioka K, Takahashi J, Momose M, Inoue S, Kojima A. Bioorg. Med. Chem. Lett. 2010; 20: 272
• 7a Xi GL, Liu ZQ. J. Agric. Food Chem. 2015; 63: 3516
• 7b Kostova I, Bhatia S, Grigorov P, Balkansky S, Parmar VS, Prasad AK, Saso L. Curr. Med. Chem. 2011; 18: 3929
• 8 Stefani HA, Gueogjan K, Manarin F, Farsky SH. P, Zukerman-Schpector J, Caracelli I, Pizano Rodrigues SR, Muscara MN, Teixeira SA, Santin JR, Machado ID, Bolonheis SM, Curi R, Vinolo MA. Eur. J. Med. Chem. 2012; 58: 117
• 9a Hu Q, Tan Y, Liu M, Yu J, Cui Y, Yang Y. Dyes Pigm. 2014; 107: 45
• 9b Guo H.-M, Tanaka F. J. Org. Chem. 2009; 74: 2417
• 9c Zak J, Ron D, Riva E, Harding HP, Cross BC. S, Baxendale IR. Chem. Eur. J. 2012; 18: 9901
• 10a Xiao D, Martini LA, Snoeberger RC, Crabtree RH, Batista VS. J. Am. Chem. Soc. 2011; 133: 9014
• 10b Brown AM, Antila LJ, Mirmohades M, Pullen S, Ott S, Hammarström L. J. Am. Chem. Soc. 2016; 138: 8060
• 11a Azizian J, Mohammadi AA, Bidar I, Mirzaei P. Monatsh. Chem. 2008; 139: 805
• 11b Sharghi H, Jokar M. Heterocycles 2007; 71: 2721
• 11c Rodríguez-Domínguez JC, Kirsch G. Synthesis 2006; 1895
• 11d Rodriguez-Dominguez JC, Kirsch G. Tetrahedron Lett. 2006; 47: 3279
• 11e Polito L, Cravini M, Poletti L, Lay L. Synth. Commun. 2006; 36: 2203
• 12 Kim KM, Chung KH, Kim JN, Ryu EK. Synthesis 1993; 283
• 13 Kim K.-M, Park I.-H. Synthesis 2004; 2641
• 14 Das B, Venkateswarlu K, Krishnaiah M, Holla H. Tetrahedron Lett. 2006; 47: 8693
• 15 Das B, Venkateswarlu K, Majhi A, Siddaiah V, Reddy KR. J. Mol. Catal. A: Chem. 2007; 267: 30
• 16 Ganguly NC, De P, Dutta S. Synthesis 2005; 1103
• 17 Venkateswarlu K, Suneel K, Das B, Reddy KN, Reddy TS. Synth. Commun. 2009; 39: 215
• 18 Racys DT, Sharif SA. I, Pimlott SL, Sutherland A. J. Org. Chem. 2016; 81: 772
• 19 Zhou C.-Y, Li J, Peddibhotla S, Romo D. Org. Lett. 2010; 12: 2104
• 20 Leboeuf D, Ciesielski J, Frontier AJ. Synlett 2014; 25: 399
• 21 We have applied for a Chinese patent CN 105541772 to protect the rough findings of this idea before the optimization. In the patent, the activation of NCS using CuCl₂·2H₂O, AlCl₃, ZnCl₂ or FeCl₃ (1–5 equiv) and the chlorination reactions to afford 2a, 2c, and 2d are described, however, optimized reaction conditions for full conversion of the starting materials are not specified.
• 22 In the control experiment, no reaction happened by mixing coumarin with either NCS or CuCl₂ alone in refluxing acetonitrile even after 36 h.
• 23a Hao W, Liu Y. Beilstein J. Org. Chem. 2015; 11: 2132
• 23b Du Z.-J, Gao L.-X, Lin Y.-J, Han F.-S. ChemCatChem 2014; 6: 123
• 23c Prakash GK. S, Mathew T, Hoole D, Esteves PM, Wang Q, Rasul G, Olah GA. J. Am. Chem. Soc. 2004; 126: 15770
• 23d Galabov B, Nalbantova D, Schleyer PV. R, Schaefer HF. Acc. Chem. Res. 2016; 49: 1191
• 24 Jardim GA. M, da Silva Junior EN, Bower JF. Chem. Sci. 2016; 7: 3780
• 25 Huang W, Ding Y, Miao Y, Liu M.-Z, Li Y, Yang G.-F. Eur. J. Med. Chem. 2009; 44: 3687
• 26 3,7-Dichloro-2H-chromen-2-one (2b) – Typical Procedure To a 50 mL round-bottom flask, 7-chloro coumarin (180.6 mg, 1 mmol), NCS (935.2 mg, 7 mmol), CuCl₂·2H₂O (2387 mg, 14 mmol), and 20 mL anhydrous MeCN were added and refluxed until the reaction completed. The cooled mixture was concentrated in vacuum, dispensed in 25 mL 5% NaHSO₃ aqueous solution and extracted with 25 mL EtOAc for three times. The organic layer was combined, washed with 10 mL water and dried over anhydrous Na₂SO₄. After the solvent was removed, the crude product was purified by silica gel column chromatography to afford 96.2 mg 2b (44.9%). 3,7-Dichloro-2H-chromen-2-one (2b) White solid, mp 123–124 ℃ (acetone/PE, 2:1, V/V). ¹H NMR (600 MHz, CDCl₃): δ = 7.84 (s, 1 H), 7.40 (d, J = 8.3 Hz, 1 H), 7.38 (d, J = 1.9 Hz, 1 H), 7.30 (dd, J = 8.3, 1.9 Hz, 1 H). ¹³C NMR (151 MHz, CDCl₃): δ = 156.6, 152.9, 139.3, 137.9, 128.0, 125.7, 122.5, 117.4, 117.2, 100.0. IR (KBr): ν_max = 3103, 3061, 3040, 1736, 1720, 1702, 1604, 1484, 1403, 1348, 1318, 1245, 1206, 1144, 1123, 1077, 998, 969, 939, 916, 865, 825, 752, 681, 617, 595, 462 cm^–1. HRMS (ESI): m/z calcd for C₉H₅Cl₂O₂ [M + H] (³⁵Cl): 214.9661; found: 214.9661.