Boron-mediated directed aromatic C–H hydroxylation

Transition metal-catalysed C–H hydroxylation is one of the most notable advances in synthetic chemistry during the past few decades and it has been widely employed in the preparation of alcohols and phenols. The site-selective hydroxylation of aromatic C–H bonds under mild conditions, especially in the context of substituted (hetero)arenes with diverse functional groups, remains a challenge. Here, we report a general and mild chelation-assisted C–H hydroxylation of (hetero)arenes mediated by boron species without the use of any transition metals. Diverse (hetero)arenes bearing amide directing groups can be utilized for ortho C–H hydroxylation under mild reaction conditions and with broad functional group compatibility. Additionally, this transition metal-free strategy can be extended to synthesize C7 and C4-hydroxylated indoles. By utilizing the present method, the formal synthesis of several phenol intermediates to bioactive molecules is demonstrated.

P henols are structural constituents of pharmaceuticals, agrochemicals, polymers, and naturally occurring compounds and serve as versatile synthetic intermediates [1][2][3][4] . Bioactive molecules of particular interest are (hetero)arenes such as amides, indolines, and indoles-containing hydroxyl groups (Fig. 1a) [5][6][7][8][9][10][11] . The site-selective introduction of a hydroxyl group to a (hetero)arene is an important task in both chemical industry and organic synthesis. Traditional methods used for phenol preparation include nucleophilic aromatic substitution of activated aryl halides 12 and Sandmeyer-type hydroxylation 13 , as well as the transition-metal-catalysed hydroxylation of (hetero)aryl halides with hydroxide salts (e.g., KOH and NaOH) [14][15][16][17][18] , all of which require the presence of a (pseudo)halide in the (hetero) arenes. During the past decade, C-H functionalization has gained considerable momentum, holding great promise for avoiding the preinstalled functional groups [19][20][21][22][23][24][25][26][27][28] . Among these C-H functionalization techniques, hydroxylation is one of the most important C-H functionalization reactions 29,30 . As early as 1990, Fujiwara et al. explored the hydroxylation of benzene using O 2 as the oxidant enabled by Pd catalysis 31 . However, this pioneering work had several limitations, such as a low efficiency, poor selectivity, and harsh reaction conditions. Substrates bearing a chelating functional group can coordinate with the metal catalyst and undergo further C-H functionalization [32][33][34] . In this context, several groups have explored transition-metal-catalysed directed aromatic C-H hydroxylation using organic oxidants, hydrogen peroxide or molecular oxygen (Fig. 1b) [35][36][37][38][39][40][41][42][43][44][45][46] . While synthetically very attractive, most of these protocols still suffer from the use of expensive noble metals, such as Pd, Rh, Ru, and Ir, as catalysts. This requirement may be a significant limitation, especially for applications needing large-scale synthesis methods and for the removal of toxic trace metals from pharmaceutical products. From a synthetic perspective, the ability to prepare synthetically relevant scaffolds via regio-controlled C-H hydroxylation under mild conditions by using cheap oxidants and avoiding the use of transition metals would be of great importance.
The transition-metal-catalysed C-H borylation reaction has emerged as an effective method for the construction of arylboronic acids and their derivatives [47][48][49] . Recently, our group 50 and the Ingleson group 51 reported a general strategy for the mild directed C-H borylation of (hetero)arenes using BBr 3 as both the reagent and catalyst under metal-free conditions [52][53][54][55] . BBr 3 is an attractive borylation agent because it is a commercially available in multigram to kilogram quantities and is cheaper than most common boron reagents. To further extend the utility of this strategy, here, we developed a one-pot directed C-H borylation/ oxidation protocol to access numerous structurally diverse phenols, whose regioselectivity is not easily accessed by traditional strategies (Fig. 1c). Replacing the transition-metal-catalysed C-H hydroxylation process by a boron-mediated strategy offers an

Results
Reaction design. We initiated our study by investigating the reaction of N-pivaloyl amide 1a with BBr 3 ( Table 1). As a result, we discovered that the use of 1.0 equivalent of 1a with 1.1 equivalents of BBr 3 in DCM at room temperature for 1 h led to the full conversion of the precursors and formation of boron complex 1b. Then, the treatment of 3.0 equivalents of NaBO 3 in a THF and H 2 O (1:1) co-solvent led to the in situ formation of the hydroxylated product, and phenol 1c was isolated with a 92% yield ( Table 1, entry 1). Other boron halides, such as BF 3 , were not efficient for this reaction (Table 1, entry 2), and BCl 3 only afforded a trace amount of the product (Table 1, entry 3). When the reaction was carried out using ClBcat or 9-BBN, we did not observe any C-H borylation or hydroxylation products (Table 1, entries 4-5). The substrate 1a′ bearing an N-Me group failed to achieve this transformation, confirming the importance of the Npivaloyl moiety for achieving both a high reactivity and selectivity (Table 1, entry 6). To our delight, other common oxidants, such as oxone and H 2 O 2 , were also effective for this hydroxylation process, generating the desired product 1c in slightly lower yields ( Table 1,

entries 7-8).
Scope of the methodology. We first examined the scope of the ortho-selective C-H hydroxylation of amides (Fig. 2). When the simple N-pivaloyl amide 2a was employed as a substrate, hydroxylation proceeded at the ortho C-H bond, affording 2c with a 85% yield. Amides bearing methyl (3-5a), t Bu (6a), phenyl (7a), and halogen-containing motifs (8-13a) at the ortho, meta, and para positions underwent facile hydroxylation and afforded the corresponding products 3-13c in good to excellent yields. The amides bearing electron-withdrawing groups such as CF 3 (14-15a), COOMe (16a), and CN (17a) are particular noteworthy; these substrates produced ortho-hydroxylated products 14-17c with 66-80% yields. Electron-donating groups such as OTBS (18c) and SMe (19c) at the para position of the amides are tolerated. Substrate 20a bearing a methoxy group can undergo ortho C-H hydroxylation and O-demethylation to generate the corresponding product 20c with a 81% yield. Notably, the phenyldiazenyl substituent in substrate 21a, which is also susceptible to C-H borylation, remained intact during the reaction. Other Npivaloyl amides, including N-methylaniline (22c), tetrahydroquinoline (23c), and indoline (24c), are also tolerated for C-H hydroxylation. This protocol is compatible with heterocyclic motifs such as thiophene 25c. Polyaromatic substrates 26-28c were also shown to be highly reactive. As a prominent structural motif, N-arylpyrrolidinones have been used in Ru(II)-catalysed C-H hydroxylation 56,57 . We found that the boron-mediated directed C-H hydroxylation of N-phenylpyrrolidinone (29a) in the presence of BBr 3 could provide the desired product 29c with a 79% yield. The system was compatible with the different paraand meta-substitution patterns in the phenyl ring of the Narylpyrrolidinone backbone (30-36c). In addition, this C-H hydroxylation method is not limited to N-arylpyrrolidinones. Lactams such as 37-38a, oxazolidin-2-one 39a, and thiophene 40a could also undergo C-H hydroxylation at the ortho position, affording good yields of products 37-40c. Subjecting N-pivaloyl amides 41-45a, which are substrates bearing two N-pivaloyl directing bonds, to our system resulted in the selective formation of the difunctionalization products 41-45c in 60-89% yields. These bisphenols could be utilized as precursors for construction of polymers 58 .  We next investigated the scope of the C7 selective C-H hydroxylation of indoles (Fig. 3a). We found that indole 46a could generate 7-hydroxyindole 46c with a 88% yield by a cascade C-H borylation/oxidation/DG removal protocol, in which the N-Piv group can be removed automatically during work-up with K 2 CO 3 . Indoles bearing methyl (47-49a) substituents at the 4-6 positions underwent facile hydroxylation and afforded the corresponding products in 74-85% yields. Again, halogencontaining motifs (F, Cl, and Br, 50-56a) work very well in the C7 selective borylation process. In addition, substrate 57a contains a phenyl substituent also delivering coupled product 57c with a 70% yield. We further examined the scope of using C3pivaloyl indoles as coupling partners with BBr 3 ; these compounds reacted with a high regioselectivity to produce C4-hydroxylated indoles (Fig. 3b) 59 . We first evaluated the influence of the N-H protection groups on the indoles. Notably, the free indole 58a could provide the desired product 58c with a 59% yield. The treatment of the indoles 59-60a bearing N-Me and N-Bn groups in the system provided a 71% and 54% isolated yields of the corresponding C4-hydroxylation products 59-60c. Indole 61a bearing an N-Ts protection group can promote the reactivity of this transformation, affording the product 61c with a 85% yield. Regarding the scope of the indole framework, diverse substituents, including methyl (62c), F (63a), Cl (64-65c), Br (66-67c), I (68c), and phenyl (69c) are tolerated.
Synthetic applications. To further demonstrate the potential synthetic applications of this method, we showed three examples to compare existing strategies with our developed C-H hydroxylation method. Previously, using N-acetylindoline 70 as a model substrate for the total synthesis of the potent caspase-8 inhibitor (+)-haplocidine and its N1-amide congener (+)-haplocine, the precursor acetoxy-indoline 71′ was generated with a 84% yield by the palladium-catalysed C7 hydroxylation of indoline 60 . Based on the boron-mediated strategy, we prepared product 71 from substrate 70 with a 71% yield, in which N-acetyl can be used as a directing group (Fig. 4a). Trauner and co-workers 61 reported the evolution of the total synthesis of exiguamines,  where nitrovinylindole 74 was a key intermediate. To simplify this synthesis process, we provided an alternative route to 74 using the developed C-H hydroxylation protocol. The indole substrate 72 was regio-selectively hydroxylated at the C7 position and further deprotected and then protected as a benzyl ether to yield 7-hydroxy-6-bromoindole derivative 73 with a 63% yield. Then, indole 73 was formylated and condensed with nitromethane to yield nitrovinylindole 74 with a 89% yield (Fig. 4b).
The Renata group 62 recently identify a concise synthetic route to access tambromycin. During the study, they were drawn to a thallium-mediated C-H hydroxylation of indoles at the C4 position, suffering from highly variable yields and a lack of scalability. Inspired by this result, we finally focused our attention on the boron-mediated strategy to synthesize indole 78. Using Nmethyl indole 75 as a substrate, C4-hydroxylation was identified as a viable approach to access the desired indole fragment 76 after etherification with MeI. To our delight, the removal of a pivaloyl group from 76 was readily accomplished by a reverse Friedel-Crafts reaction in the presence of TsOH and glycol, providing a good yield of 77. Further C3 formylation and oxidation could provide a good yield of the key building block 78, which was facile to convert to tambromycin (Fig. 4c).

Discussion
In summary, we have developed an efficient boron-mediated system that is capable of mimicking the chelation-assisted metallic system to achieve directed C-H hydroxylation. The use of this method for the preparation of substituted phenols and downstream-functionalized products showcases the strategic opportunity to use this strategy for the synthesis of biologically active compounds. The reaction provides a simple new bond disconnection protocol for constructing these motifs with different regioselectivities and broader functional group compatibilities than existing methods.

Methods
General procedure for the synthesis of phenol 1c. A flame-dried 25 mL Schlenk tube was flushed with argon, and N-pivaloyl amide 1a (0.2 mmol, 1.0 equiv) and dry DCM (0.5 mL, 0.4 M) were introduced. A solution of BBr 3 (1.0 M in DCM, 0.22 mL, 1.1 equiv) was added slowly under an argon atmosphere. The mixture was stirred at room temperature for 1 h. After stirring, the solvent was removed under vacuum directly. NaBO 3 ·4H 2 O (92.3 mg, 0.6 mmol, 3.0 equiv), 0.5 mL of THF, and 0.5 mL of H 2 O were sequentially added to the reaction mixture and stirred at room temperature for another 1 h (monitored by TLC). After that, the excess water was removed by filtration with MgSO 4 and then washed with EtOAc (10.0 mL × 3). The filtrate was collected, and the crude mixture was directly subjected to column chromatography on a silica gel, using petrol ether/EtOAc (10/1) as the eluent to give the desired product 1c as a white solid (41.5 mg, 92%).

Data availability
The authors declare that the data supporting the findings of this study are available within the article and Supplementary Information file or from the corresponding author upon reasonable request. The X-ray crystallographic coordinates for the structures reported in this study have been deposited at the Cambridge Crystallographic Data Centre (CCDC) under deposition numbers CCDC 1910134. These data can be obtained free of charge from The Cambridge Crystallographic Data Centre via www.ccdc.cam.ac. uk/data_request/cif.