Nitrosonium ion catalysis: aerobic, metal-free cross-dehydrogenative carbon–heteroatom bond formation †

Catalytic cross-dehydrogenative coupling of heteroarenes with thiophenols and phenothiazines has been developed under mild and environmentally benign reaction conditions. For the first time, NO x + was applied for catalytic C–S and C–N bond formation. A comprehensive scope for the C–H/S–H and C–H/N–H cross-dehydrogenative coupling was demonstrated with 4 60 examples. The sustainable cross-coupling conditions utilize ambient oxygen as the terminal oxidant, while water is the sole by-product. The formation of carbon–heteroatom bonds is fundamental for the synthesis of natural products, pharmaceuticals and materials science. 1 To overcome the requirement for pre-functionalized starting materials, cross-dehydrogenative coupling (CDC) has emerged as a highly efficient strategy. 2 Transition-metal-catalyzed C–S and C–N bond formation has been widely reported. 3 Cost, toxicity and oxygen sensitivity of catalysts limit the general applicability. 4 Consequently, metal-free synthesis has gained increasing interest. 5 Different metal-free approaches for the C–H/S–H CDC have been reported. 6 Additionally, the unique dehydrogenative amination with phenothiazines has received significant attention. 7 High temperatures, excess of oxidants and harmful solvents are common limitations.

The formation of carbon-heteroatom bonds is fundamental for the synthesis of natural products, pharmaceuticals and materials science. 1 To overcome the requirement for pre-functionalized starting materials, cross-dehydrogenative coupling (CDC) has emerged as a highly efficient strategy. 2 Transition-metal-catalyzed C-S and C-N bond formation has been widely reported. 3 Cost, toxicity and oxygen sensitivity of catalysts limit the general applicability. 4 Consequently, metal-free synthesis has gained increasing interest. 5 Different metal-free approaches for the C-H/S-H CDC have been reported. 6 Additionally, the unique dehydrogenative amination with phenothiazines has received significant attention. 7 High temperatures, excess of oxidants and harmful solvents are common limitations.
Nitronium and nitrosonium salts are inexpensive, stable and non-toxic single-electron oxidants. 8 Radner's group reported the synthesis of biaryls using NOBF 4 as catalyst (Fig. 1a). 9 Ambient oxygen was identified as the terminal oxidant and water as the by-product. 10 Later, Wang's group reported the catalytic intramolecular C-C bond formation (Fig. 1b). 11 Under acidic reaction conditions, NO + is generated in situ from NaNO 2 . The oxidative coupling of phenols is well studied. 12 Recently, our group reported the NO + catalyzed coupling for the construction of C-C bonds. 13 Despite the impact of NO + as catalyst for oxidative C-C bond formation, the application in carbon-heteroatom bond formation via C-H bond functionalization is unprecedented. Herein, we demonstrate the first NO x + catalyzed C-H/S-H and C-H/N-H CDC under mild and environmentally benign reaction conditions (Fig. 1d). Nitrosonium salts are capable to convert thiols to disulfides. 14 Oxidation of thiols proceeds via transient S-nitrosation and recombination of S-centred radicals. Due to the low bond dissociation energy (BDE) of phenols and thiophenols, the possibility for a radical-radical recombination reaction of phenoxy and sulfur radicals was hypothesized. 15 A multi parameter optimization for the cross-coupling of p-cresol (1a) and 4-chlorothiophenol (2a) was performed (Table S1, ESI †). To our delight, 3a was isolated in excellent yield, by using NO 2 BF 4 as the catalyst. Hexafluoroisopropanol (HFIP) was identified as the best solvent, due to its acidic character and the unique ability to stabilize radical intermediates. 16 Initially, the scope for the cross-coupling of phenols and thiophenols was studied (Scheme 1). The reaction was scaled to 1 mmol, which did not alter the outcome of the reaction. Functional groups at the para-position of phenols were well tolerated (3b-3d). 2,4-Substituted phenols yielded products 3e-3h in good to excellent yields, covering electron-rich and sterically demanding functional groups. Product 3g was isolated in quantitative yield and product 3l was synthesized with high paraselectivity. Electron-rich product 3j revealed selectivity for the meta-position of phenol. The same outcome was observed for product 3k by blocking the ortho-and para-positions. Dearomatization and subsequent 1,4-addition appeared to be an alternative pathway. 3l was isolated in moderate yield, using a 3,5-subsituted phenol. Polysubstituted phenols allowed the isolation of products 3m-o in 76-98% yields. Naphthol derivatives were compatible, affording products 3p, 3q. Next, substituted thiophenols were tested. Different functional groups on the para-position were well tolerated (3s-v). Alkyl or chloro substituents at the orthoand meta-position afforded products 3w-3y in good yields. Double thioarylation was not observed under the developed conditions. Alkyl and benzyl thiols did not yield the desired products either. To stress the utility of the obtained products, 3a was transformed into benzothiophene 3z by applying a dual C-H bond activation strategy. 17 Next, the thioarylation of indoles was studied (Scheme 2). Unprotected indoles gave better results than N-protected analogues. This result makes the reaction conditions more attractive for other applications. The cross-coupling of indole 4a and thiophenol 2a yielded 5a in 77% yield. Scaling the reaction to 1 mmol gave 5a unaffectedly. Functional groups with different electronic properties at the indole skeleton were well tolerated (5b-f). Further, thiophenols were decorated with functional groups at the para (5h-j), ortho (5k-l) and meta (5m) position. Product 5n was isolated in 51% yield bearing a napthyl moiety. Polysubstituted products 5o-5s were synthesized in good yields, covering combinations of electron-rich and electron-deficient functional groups. Next, substituted pyrroles were tested. Product 5t was isolated in 64% yield. 2-Substituted pyrrol yielded the double functionalized product 5v in good yield. To further stress the applicability, 5l was transformed into the indol-fused benzothiophenes 5v. 18 Next, the time course of the cross-coupling reactions was analysed by GC-MS-FID (Fig. 2). Interestingly, thiophenol 2a was fully converted to disulfide 6a, prior to the coupling step with phenol 1a (Fig. 2A). In contrast, indole 4a and thiophenol 2a underwent synchronous cross-coupling without initial Scheme 1 Scope with respect to phenols (1) and thiophenols (2). Reaction conditions: 1 (0.1 mmol, 1 equiv.), 2 (2 equiv.), HFIP (0.05 M), at room temperature under air atmosphere. Yields are given for isolated products after column chromatography. a Reaction carried out with 1 mmol of phenol 1a. b 1. 3a (0.3 mmol, 1 equiv.), MeI (1.5 equiv.), NaH (1.2 equiv.) in DMF (0.1 M) at room temperature; 2. (CF 3 CO 2 ) 2 Pd (20 mol%), AgOAc (5 equiv.), K 2 CO 3 (1.5 equiv.) in PivOH (0.3 M) at 130 1C for 2 d.
Scheme 2 Scope with respect to indoles (4) and thiophenols (2). Reaction conditions: 4 (0.1 mmol, 1 equiv.), 2 (2 equiv.), HFIP (0.05 M), at room temperature under air atmosphere. Yields are given for isolated products after column chromatography. a Reaction carried out with 1 mmol of indol 4a. b 5l (0.08 mmol, 1 equiv.), Pd(PPh 3 ) 2 Cl 2 (5 mol%), CsPiv (2 equiv.) in N,N-dimethylacetamide (0.1 M). formation of 6a (Fig. 2B). Conducting the coupling reaction with disulfide 6a confirmed the discrete recombination selectivities (Schemes S2 and S5, ESI †). Consequently, disulfide 6a is an intermediate for the coupling with phenol, but indoles undergo direct recombination with thiophenols. Further control experiments were conducted (ESI † for the details). No product was formed in the presence of radical trap butylated hydroxytoluene (BHT) and the product formation was inhibited under inert gas atmosphere (Schemes S2 and S4, ESI †). Ambient oxygen was crucial to maintain the catalytic activity. Initiation of the coupling reaction through oxidation of weak heteroatomhydrogen bonds was studied by methylating the starting materials. Drastically reduced conversion was observed, which excludes a direct single-electron-transfer (Schemes S2 and S4, ESI †). Thiophenol 2a was quantitatively oxidized to disulfide 6a in the presence of NOBF 4 (Scheme S3, ESI †). Conversion to disulfide 6a was tied to the presence of air and did not take place in the presence of BHT (Scheme S3, ESI †).
Based on the control experiments, a mechanism for the cross-coupling of phenol 1a and indole 4a with thiophenol 2a was proposed (Fig. 2C). Initially, intermediate A is formed by S-nitrosylation of 2a. Homolytic cleavage releases radical B, which forms disulfide 6a by recombination with itself. Phenol 1a is oxidized to form a phenoxy radical. Delocalization of the phenoxy radical (C) and attack at the disulfide bond leads to release of radical B. Rearomatization furnishes product 3a.
Oxidation of indole 4a gives intermediate D, which undergoes recombination with B. Oxidation of indole 4a and thiophenol 2a occurs synchronously, leading to the formation of 5a before disulfide 6a begins to form. Further, NO 2 BF 4 oxidizes the starting materials to form NO 2 . NO 2 dimerizes to form N 2 O 4 , which undergoes disproportionation to NONO 3 . 11a Water is released upon protonation by HBF 4 and nitrosonium and nitronium tetrafluoroborate are regenerated. NOBF 4 is capable of oxidizing the substrates in the same way as the nitronium salt. Oxidation of substrates results in the formation of nitrogen monoxide, which has to be oxidized by ambient oxygen, in order to maintain the catalytic cycle.
Based on the revealed radical mechanism, we hypothesized that phenothiazines represent suitable substrates for the radical recombination with phenols, due to the low N-H BDE and the unique ability to stabilize radical intermediates. 19 Multi parameter optimization was performed (Table S2, ESI †). Sustainable C-H bond amination of phenols with phenothiazines was achieved using NaNO 2 as catalyst, omitting halogenated reagents and solvents. The developed conditions overcome environmental issues of known methods. 7 Next, the scope of the catalytic C-H/N-H cross-dehydrogenative coupling was studied (Scheme 3). Initially, the cross-coupling of 4-methoxyphenol (7a) and different phenothiazines was performed. Products 9a-d were isolated in good yields. Different 4-substituted phenols yielded the desired cross-coupling products in moderate to excellent yields (9e-j). Cross-coupling with differently substituted phenothiazines was achieved for several phenols. Product 9k was isolated in good yield as single regioisomer. Polyfunctionalized products 9l-q were isolated in moderate to good yields. Phenoxazine proved to be compatible Scheme 3 Scope for the cross-dehydrogenative C-H bond amination. Reaction conditions: 8 (0.2 mmol, 1 equiv.), 9 (3 equiv.), 4 : 1 toluene/PivOH (0.05 M), 0 1C to rt under air atmosphere. Yields are given for isolated products after column chromatography. a NOBF 4 (0.02 mmol, 10 mol%) was used as catalyst.
for the cross-coupling reaction (9r). 2-Naphthol yielded the desired product 9s in 73% yield. Finally, the cross-coupling of 2-phenylindole and phenoxazine was successfully performed. However, synthesis of 9t worked superior if NOBF 4 was used as catalyst. The underlying mechanism for the aerobic C-H bond amination proceeds analogously as described before via direct radical-radical recombination under aerobic conditions (Scheme S8, ESI †).
In summary, we have reported the first application of NO x + as efficient and environmentally friendly catalyst for carbonheteroatom bond formation. The operationally simple and sustainable protocol enables the C-H/S-H and C-H/N-H CDC. Ambient oxygen serves as stoichiometric oxidant and water is generated as by-product. A broad scope was demonstrated in good yields and regioselectivities. The reported methodology offers mild reaction conditions and does not require an excess of reagents or any specialized equipment. A. P. A. acknowledges the support of the DFG (AN 1064/4-1) and the Boehringer Ingelheim Foundation (Plus 3). L. B. is supported by the Verband der Chemischen Industrie e.V. We gratefully acknowledge Ms R. Perinbarajah for assistance. Open Access funding provided by the Max Planck Society.

Conflicts of interest
There are no conflicts to declare.