Hydroxyalkylated heterocycles are one class of important heterocyclic compounds in agricultural and pharmaceutical field [1-4]. Over the years, many hydroxyalkylated heterocycles have been synthesized as antiviral [5], antagonist [6], antifungal [7], antimalarial [8] and so on (Fig. 1).

On the other hand, cross-dehydrogenative coupling (CDC) has become a captivating strategy for bond construction, due to the "excellent step and atom economy" properties [9-14]. As one type of CDC reactions, the direct hydroxylation of heterocycles with alcohols has also received great attention because of the "readily available", "inexpensive" and "biodegradable" advantages of alcohols [15]. In 2011, Wang et al. [16] published the C2-alkylation of azoles with alcohols and ethers using tert-butyl hydroperoxide (TBHP) as the oxidant at 120 ℃ (Scheme 1a). And in 2017, di-tert-butyl peroxide (DTBP) oxidized CDC reactions of thiophenes and pyridines with alcohols and cyclic ethers at 130 ℃ were developed by Kianmehr and co-workers [17] (Scheme 1b). Next year, Yu group [18] reported DTBP-mediated C2-hydroxyalkylation reactions of chromones with alcohols at 140 ℃ (Scheme 1c). These reactions well developed heteroarenes hydroxyalkylated methods, but they still exhibit a series of disadvantages, such as relatively high reaction temperature, inflammable and explosive properties of organic peroxides oxidants. In 2019, our group [19] reported a mild and convenient direct hydroxyalkylation reaction of benzothiazoles with a variety of alcohols in the presence of K₂S₂O₈ in H₂O at 65 ℃ (Scheme 1d). Despite the fact that this methodology is milder than the reactions mentioned above, it is still carried out under heating condition.

During the past decade, photoredox catalysis, by using visible light as a renewable energy source, has facilitated a number of organic transformations that involve a single-electron transfer (SET) pathway [20-27]. There is likewise a growing trend of introducing photocatalysis into the hydroxyalkylation reaction of heterocycles. In 2016, DiRocco group [28] developed an iridium complex catalyzed hydroxymethylation of heteroaromatic bases (Scheme 1e). This reaction was carried out at room temperature, but it suffered from certain limitations such as complex handling procedure, expensive transition-metal catalyst and inevitable metal residues. Inspiringly, Lei et al. [29] explored a mild and metal-free photocatalytic cross-coupling of (iso)quinolines with different alcohols by using selectfluor as oxidant (Scheme 1f). In order to extend the hydroxyalkylation of various heterocycles, combining our previous researches [30-36] on photochemical reactions and green organic synthesis, we herein report a mild and convenient, visible light-induced direct hydroxyalkylation of 2H-benzothiazoles with alcohols by using selectfluor as oxidant at room temperature.

Initially, we chose 2H-benzothiazole (1a) and ethanol (2b) as the model substrates to investigate the reaction conditions. The results were outlined in Table 1. Generally, the model reaction was carried out under a nitrogen atmosphere irradiated by 30 W blue LEDs at room temperature. Firstly, a variety of oxidants were investigated in conjunction with acetonitrile (MeCN) as the solvent. It was identified that the desired hydroxyalkylated product can be afforded in 34% yield using selectfluor (3 equiv.) as a visible light-activated oxidant, but the reaction could not be promoted by utilizing TBHP or K₂S₂O₈ (Table 1, entries 1-3). In order increase the yield of the reaction, we added trifluoroacetic acid (TFA) (1.5 equiv.) as the additive. Markedly, the yield increased to 61% (Table 1, entry 4). Subsequently, the acid was changed to p-toluenesulfonic acid monohydrate (TsOH·H₂O), acetic acid (AcOH) and hydrochloric acid (HCl), but the yields were inferior to that of TFA (Table 1, entries 5-7). We further tested the amount of selectfluor, but no improvement was observed with less or more than 3 equiv. of selectfluor (Table 1, entries 8 and 9). In addition, the product could be obtained in 71% yield as the amount of TFA was reduced to 1.0 equiv., but the yield decreased when the amount increased (Table 1, entries 10-12). Examination of a range of common solvents showed that the reaction could be carried out efficiently in acetonitrile while reaction without solvent or with solvents such as methylene chloride (CH₂Cl₂), N, N-dimethylformamide (DMF), dimethyl sulfoxide (DMSO), acetone all did not give any product (Table 1, entries 13-17). Further studies showed that reducing or increasing the amount of acetonitrile resulted in a diminished yield of 3ab (Table 1, entries 18 and 19). What is more, the results showed that the product yield decreased to 65% when the reaction time was shortened to 16 h, but no clear improvement in yield was observed when the reaction time increased to 24 h (Table 1, entries 20 and 21).

Subsequently, the gram-scale synthesis experiment was carried out. Notably, a comparable 67% yield was obtained under the optimized reaction conditions when the model reaction was performed in 10 mmol scale (Scheme 2). Moreover, the UV-vis spectra of selectfluor and the reaction compounds were measured, the results demonstrated that there is an overlap between the absorption spectrum of them with the emission spectrum of Blue LEDs with λ = 405 nm (see the Supporting information for further details). Then, the experiments with the light on and off were conducted to prove the importance of light exposure. The reaction was inhibited in the absence of light, but was activated after light exposure, thus further confirming the crucial role of visible light in the reaction (Fig. 2).

With the optimized reaction conditions in hand, the scope of the alcohols (2) in the direct hydroxyalkylation with 2H-benzothiazole (1a) was investigated. As shown in Scheme 3, a broad range of structurally diverse aliphatic alcohols including methanol, ethanol, propanol, 2-propanol, butanol, 2-methyl-1-propanol, 2-butanol, pentanol, 3-methyl-1-butanol and 2-pentanol underwent direct hydroxyalkylation reactions with 2H-benzothiazole to afford the corresponding cross-coupling products smoothly in moderate to good yields (3aa-3aj). The results showed that the yields of the corresponding products obtained from the aliphatic primary alcohols gradually decreased with the increase of the carbon chain (3ab, 3ac, 3ae and 3ah). Notably, the hydroxyalkylation of 2H-benzothiazoles with the secondary alcohols had higher yields than those of the primary alcohols (3ac, 3ae, and 3ah compared to 3ad, 3ag, and 3aj, respectively). It was delightful that ethylene glycol and 1, 3-propanediol were smoothly compatible with this system providing the desired hydroxyalkylation products in 56% and 59% yields, respectively (3ak and 3al). The direct alkylation of 2H-benzothiazole with cyclopentanol also provided the target product in 48% yield (3am). Regrettably, 2H-benzothiazoles with allyl alcohol or benzyl alcohol only gave trace amount of the hydroxyalkylation products (3an and 3ao).

Next, we turned our attention to explore the scope of substituted 2H-benzothiazoles. Various substituted 2H-benzothiazoles (1) and alcohols (2) were compatible with the catalytic system and the results are summarized in Scheme 4. In general, ethanol, propanol, 2-propanol reacted with substituted 2H-benzothiazoles containing an electron-withdrawing group (nitro, chloro, cyano, and acetyl) gave the desired products in 40%-66% yields (3ba, 3bb, 3be, 3bf, 3bg, 3bi and 3bj). Meanwhile, these alcohols reacted with those 2H-benzothiazoles containing an electron-donating group (6-methoxyl, 7-methoxyl) delivered the products in 35%-46% yields (3bc, 3bd, 3bh, 3bk, 3bl). The substrate scope was further extended to a series of alcohols with longer carbon chains including butanol, 2-methyl-1-propanol and 2- butanol, which were also well-tolerated (3bm-3br). However, 5-aminobenzothiazole and 6-aminobenzothiazole only gave trace amount of the target products (3bs and 3bt), which might be due to the oxidation side reactions of the amino group.

To probe the applicability of this catalytic system, we attempted to broaden the generality of this hydroxyalkylation reaction by employing ethers as the substrates. As can be seen from Scheme 5, the C(sp³)-H bond adjacent to the oxygen atom of ethers could also be activated under our protocol. Ethers (diethyl ether, isopropyl ether) and cyclic ethers (tetrahydrofuran, 1, 3-dioxolane, 1, 4- dioxane) were all compatible with the conditions providing the alkylation products in moderate to good yields (5a-5e). According to our previous work [19], the alkylation of 2H-benzothiazole and 1, 3-dioxolane would give two isomers. Notably, only the 2-position of 1, 3-dioxolane was alkylated in our protocol and single regioisomer (5d) was found. Nevertheless, methyl tert-butyl ether was unsuitable for this reaction (5f), which might be due to the unstability of the methyl tert-butyl ether free-radical intermediate and the steric hindrance of the bulky tert-butyl.

To explore the possible mechanism, a radical trapping experiment was performed by addition of a radical scavenger TEMPO (2, 2, 6, 6-tetramethyl-1-piperidinyloxy) to the reaction system (Scheme 6). The transformation was completely inhibited and TEMPO-trapped complex was observed by LC-MS, which suggested that the reaction might proceed via a radical pathway.

It has been reported that selectfluor is not only a most efficient electrophilic fluorinating reagent, but also a powerful oxidant as well as a convenient mediator of several "fluorine-free" functionalizations of organic compounds [37-39]. According to the radical trapping experiment, we proposed a plausible mechanism for the hydroxyalkylation reaction of 2H-benzothiazole with isopropanol (Scheme 7). Under visible light irradiation, selectfluor A was converted to an N radical cation B and an F radical [40, 41]. Afterward, the generated electrophilic N radical cation B would abstract a hydrogen atom from isopropanol 2c to provide an ammonium ion C and a hydroxyalkyl radical 2ca [29]. Subsequently, 2H-benzothiazole 1a was protonated by acid to form 1aa, the electron-deficient 2H-benzothiazole 1aa would capture the relatively nucleophilic hydroxyalkyl radical 2ca and deliver the corresponding radical cation adduct 1ab [42]. Then, the intermediate 1ab was deprotonated to a radical 1ac. A single-electron transfer (SET) was followed between radical 1ac and another selectfluor giving iminium ion 1ad and N radical cation B [43]. Finally, further deprotonation of 1ad would then afford the desired hydroxyalkyl coupling product 3ac.

In conclusion, we have developed a mild and convenient method for the C2-hydroxyalkylation of 2H-benzothiazoles with diverse alcohols, which was mediated by selectfluor under the blue LEDs irradiation at room temperature. This hydroxylkylation reaction tolerates a wide range of functional groups. Besides, ethers were also compatible in this reaction, leading to corresponding C2 ether-substituted 2H-benzothiazoles with high regioselectivity. As such, we expect that this new strategy can be applied to medicinal and agricultural chemicals studies in the future.

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

The authors are grateful for the financial support by the National Natural Science Foundation of China (No. 30900959) and the Natural Science Foundation of Zhejiang Province (No. LY17C140003).

Supplementarymaterial related to this article can befound, in the online version, at doi:https://doi.org/10.1016/j.cclet.2020.05.022.