Chemical and Pharmaceutical Bulletin
Online ISSN : 1347-5223
Print ISSN : 0009-2363
ISSN-L : 0009-2363
Communication to the Editor
Site-Selective α-Alkylation of 1,3-Butanediol Using a Thiophosphoric Acid Hydrogen Atom Transfer Catalyst
Hiroyasu NakaoHarunobu Mitsunuma Motomu Kanai
Author information
JOURNAL FREE ACCESS FULL-TEXT HTML
Supplementary material

2022 Volume 70 Issue 8 Pages 540-543

Details
Abstract

Herein, we developed secondary-alcohol-selective C–H alkylation of 1,3-butane diol by combining an acridinium photoredox catalyst and a thiophosphoric acid hydrogen atom transfer (HAT) catalyst. The use of non-coordinating solvent such as dichloromethane (DCM) improved secondary α-alkoxy C–H selectivity by lowering bond dissociation energy (BDE) through intramolecular hydrogen bonding.

Introduction

Catalytic C–H activation has been attracting attention as an ideal organic synthesis method.14) Hydrogen-atom transfer (HAT) catalysis in particular is a powerful strategy for the direct functionalization of inert C–H bonds.5,6) Recently, the combination of HAT catalysts and photoredox catalysts has enabled the functionalization of a variety of C–H bonds under mild conditions such as room temperature and visible light.7,8) The development of HAT catalysts that can achieve regioselective C–H activation promises even greater utility.911) Especially for substrates containing multiple hydroxy groups, regioselective C–H activation at specific α-hydroxy positions would lead to efficient synthesis of sugar and polyol derivatives.1221) However, there are only a few examples of regioselective carbon–carbon bond formation of polyols via C–H activation. In 2017, Minnaard and colleagues achieved 3-position-selective C–H alkylation of unprotected sugars using the combination of a quinuclidine catalyst, an iridium photoredox catalyst, and a hydrogen bonding catalyst13) (Chart 1a). Taylor and colleagues reported regioselective C–H alkylation of unprotected sugars using bond-weakening catalysts such as diphenyl boronic acid and diphenyl tin chloride14,17) (Chart 1a). Although C–H functionalization was achieved with high selectivity in these reports, substrates have been limited to cyclic alcohols with fixed conformations, and only one example has been reported for a linear substrate.14) Therefore, we focused on the linear 1,3-diol moiety, a common structure in biologically active molecules, to perform regioselective C–H functionalization reactions. In this work, we achieved secondary-alcohol-selective C–H functionalization of linear 1,3-diol by using a thiophosphoric acid HAT catalyst2228) (Chart 1b). This is the first example of regioselective α-hydroxy C–H functionalization of linear 1,3-diols via HAT process.29)

Chart 1. Regioselective C–H Bond Functionalization of Polyols and Diols via HAT Catalysis

Results and Discussion

We began optimization of the reaction conditions using 1,3-butane diol (1a) and benzalmalononitrile (2a) as model substrates (Table 1). Benzalmalononitrile was selected because it gave the best yield in the initial screening. First, we investigated the combination of HAT catalyst and photoredox catalyst. Contrary to expectations, the HAT catalysts used for the alcohol α-functionalization did not afford any desired product (entries 1–6).3035) Tetrabutylammonium decatungstate (TBADT) provide the product with moderate yield (41%, entry 7).36) On the other hand, we found that the combination of thiophosphoric acid catalyst and acridinium photoredox catalyst, originally reported by our group, gave the product in high yield (88%, entry 8).2227) Since the reaction regioselectivity was moderate (1.6/1 r.r.), we further investigated the conditions. Among the conditions screened, it was found that the choice of solvent affected the regioselectivity. Notably, when using dichloromethane (DCM) as the solvent, a secondary α-alkoxy C–H selective product 3a was obtained in 49% (4.4/1 r.r., entry 14).

Table 1. Screening of Reaction Conditions

Under the optimized conditions, we checked the scope of benzalmalononitrile derivatives (Chart 2). Substrates bearing an electron-withdrawing group (ester: 2b), an electron-donating group (methoxy: 2c) and bromide (2d) afforded the corresponding products, respectively, in moderate yield with relatively high regioselectivity.

Chart 2. Regioselective C–H Bond Functionalization of 1,3-Butane Diol via HAT Catalysis

In order to investigate the possibility of a decrease in selectivity due to intermolecular HAT between the substrates 1a in acetonitrile solvent, we introduced deuterated substrates 1b and 1c to observe H–D shuffling (Chart 3). However, no decrease in the deuteration rate was observed for either substrate. These results suggest that the intermolecular HAT is slow. Therefore, the regioselectivity in the HAT step determines the product ratio (3/4).

Chart 3. Deuterium Labeling Experiments

Next, we checked whether the rate-limiting step changes with solvent. Kinetic experiments were performed in acetonitrile (MeCN) and DCM solvents (Chart 4). Kinetic isotope effects were observed (MeCN: KIE = 2.8, DCM: KIE = 2.8), indicating that the HAT step is the rate-limiting step in both solvents.

Chart 4. Kinetic Studies

The proposed reaction mechanism is shown in Chart 5. A sulfur-centered radical (RS·) 5 is generated through single-electron oxidation of thiophosphoric acid (Epa = 1.18 V vs. SCE)22) by a photo-excited acridinium catalyst (*Mes-Acr+: E1/2(*Mes-Acr+/Mes-Acr·) = 2.06 V vs. SCE) followed by deprotonation.37) The C–H bond abstraction step by this radical species 5 is the rate-limiting step, as shown in Chart 4. Intramolecular hydrogen bonding of 1a is affected by the polarity of the solvent38); in non-coordinating solvents such as DCM, 1a is known to make strong intramolecularly hydrogen bonding.39) The OH group of the secondary alcohol is particularly prone to hydrogen bonding.38) This is expected to increase the electron density of the secondary alcohol and lower the bond dissociation energy (BDE) of the C–H bond at the α-position.40) Therefore, the C–H bond with smaller bond dissociation energy (i.e., secondary α-alkoxy C–H) is selectively cleaved in DCM. Radical 6 adds to benzalmalononitrile 2a to afford 7. Finally, single-electron reduction of 7 by the reduced form of the photoredox catalyst (Mes-Acr·: E1/2(Mes-Acr·/Mes-Acr+)=−0.57 V vs. SCE) and the subsequent protonation afford product 8 and the oxidized form of the photoredox catalyst (Mes-Acr+), thus closing the catalytic cycle.37)

Chart 5. Regioselective C–H Bond Functionalization of Polyols via HAT Catalysis

Conclusion

We achieved regioselective C–H alkylation of 1,3-butane diol by using a thiophosphoric acid HAT catalyst. The selection of solvent was important for making intramolecular hydrogen bonding in 1,3-butane diol, resulting in a decrease in the BDE of the C–H bond at the α-position of the secondary alcohol. Regioselective C–H alkylation at the internal position was achieved when DCM, having a non-coordinating nature, was chosen as the solvent. Based on these preliminary results, experiments expanding the generality of substrates and applying the method to more complex substrates are currently ongoing.

Experimental

General Method

1H-NMR spectra were recorded on JEOL ECX500 (500 MHz for 1H-NMR), and JEOL ECS400 (400 MHz for 1H-NMR) spectrometer. For 1H-NMR, chemical shifts are reported in ppm with the solvent resonance as the internal standard (1H: CDCl3, δ 7.26; (CD3)2CO, δ 2.05). Data are reported as follows: s = singlet, br = broad, d = doublet, t = triplet, q = quartet, m = multiplet; coupling constants in Hz; integration. All deuterated solvents were purchased from Kanto chemical or Sigma-Aldrich (St. Louis, MO, U.S.A.). Column chromatographies were performed with silica gel Merck 60 (230–400 mesh ASTM).

Substrates for Table 1

Reagents were purchased from Aldrich, Tokyo Chemical Industry Co., Ltd. (TCI) (Tokyo, Japan), Kanto Chemical Co., Inc. (Tokyo, Japan), and Wako Pure Chemical Corporation (Osaka, Japan) and were used as received.

Synthesis

1,3-Butandiol (18.0 mg, 0.20 mmol), benzalmalononitlile (15.4 mg, 0.10 mmol), MesAcrClO4 (2.0 mg, 0.050 mmol), and thiophosphoric acid (TPA) (3.6 mg, 0.01 mmol) were dissolved in super dehydrated dichloromethane (1.0 mL) in a screw-capped vial under argon atmosphere. The vial was subjected to blue LED irradiation for 16 h under temperature control (approx. 27–29 °C). Then, the reaction mixture was evaporated and 1,1,2,2-tetrachloroehane was added as an internal standard. Purification of the crude mixture by column chromatography on silica gel afforded cyclized products. If products did not cyclize completely, an excess amount of Et3N was added and products were analyzed after evaporation. Regioisomers 3 and 4 were inseparable.

2-Amino-5-(2-hydroxyethyl)-5-methyl-4-phenyl-4,5-dihydrofuran-3-carbonitrile (3a)

A colorless oil; 3a/4a = 4.4/1 r.r. (8.3 mg) (inseparable stereoisomer, mixture of diastereomers (a + b), dr = 1.1 : 1 (a : b)); 1H-NMR ((CD3)2CO): δ = 7.36–7.17 (5H, m, a + b), 6.37 (2H, br s, a + b), 4.23 (s, 1H, b), 4.01 (s, 1H, a), 4.00–3.46 (2H, m, a + b), 2.17–1.75 (1H, m, a + b), 1.54 (3H, s, a), 1.47–1.07 (1H, m, a + b), 0.85 (3H, s, b); m/z Calcd for C14H16N2O2Na [M + Na]+ 267.1104. Found 267.1108.

Methyl 4-(5-amino-4-cyano-2-(2-hydroxyethyl)-2-methyl-2,3-dihydrofuran-3-yl)benzoate (3b)

A colorless oil; yield 18%, 3b/4b = 4.8/1 r.r. (5.4 mg) (inseparable stereoisomer, mixture of diastereomers (a + b), dr = 1.9 : 1 (a : b)); 1H-NMR ((CD3)2CO): δ = 8.01–7.98 (2H, m, a + b), 7.41–7.37 (2H, m, a + b), 6.45 (2H, br s, a + b), 4.36 (1H, s, b), 4.10 (1H, s, a) 3.90–3.33 (5H, m, a + b), 2.21–1.83 (1H, m, a + b), 1.57 (3H, s, a), 1.45–1.10 (1H, m, a + b), 0.87 (3H, s, b); m/z Calcd for C16H18N2O4Na [M + Na]+ 325.1159. Found 325.1157.

2-Amino-5-(2-hydroxyethyl)-4-(4-methoxyphenyl)-5-methyl-4,5-dihydrofuran-3-carbonitrile (3c)

A colorless oil; yield 38%, 3c/4c = 4.2/1 r.r. (10.4 mg) (inseparable stereoisomer, mixture of diastereomers (a + b), dr = 1.1 : 1 (a : b)); 1H-NMR ((CD3)2CO): δ = 7.27–7.16 (2H, m, a + b), 7.00–6.95 (2H, m, a + b), 6.30 (2H, br s, a + b), 4.48 (1H, s, b), 4.40 (1H, s, a) 3.90–3.27 (5H, m, a + b), 2.15–1.95 (1H, m, a + b), 1.55 (3H, s, a), 1.37–1.28 (1H, m, a + b), 0.85 (3H, s, b); m/z Calcd for C15H18N2O3Na [M + Na]+ 297.1210. Found 297.1210.

2-Amino-4-(4-bromophenyl)-5-(2-hydroxyethyl)-5-methyl-4,5-dihydrofuran-3-carbonitrile (3d)

A yellow oil; yield 50%, 3d/4d = 4.2/1 r.r. (16.1 mg) (inseparable stereoisomer, mixture of diastereomers (a + b), dr = 1.3 : 1 (a : b)); 1H-NMR ((CD3)2CO): δ = 7.55–7.52 (2H, m, a + b), 7.22–7.19 (2H, m, a + b), 6.43 (2H, br s, a + b), 4.26 (1H, s, b), 4.01 (1H, s, a), 3.90–3.51 (2H, m, a + b), 4.22 (1H, m), 2.10–2.00 (1H, m, a + b), 1.54 (3H, s, a), 1.47–1.11 (1H, m, a + b), 0.87 (3H, s, b); m/z Calcd for C14H15BrN2O2Na [M + Na]+ 345.0209. Found 345.0205.

Acknowledgments

This work was supported in part by JSPS KAKENHI Grant Numbers JP17H06442 (M.K.) (Hybrid Catalysis), JP20H05843 (Dynamic Exciton) and JP21K15220 (H.M.).

Conflict of Interest

The authors declare no conflict of interest.

Supplementary Materials

This article contains supplementary materials.

References
 
© 2022 The Pharmaceutical Society of Japan
feedback
Top