Palladium-catalyzed oxidative cross-coupling for the synthesis of α-amino ketones.

A novel oxidative cross-coupling reaction for the synthesis of α-aryl α-amino ketones in the presence of palladium catalysts using T+BF4 - as an oxidant has been developed. This transformation was achieved by direct C-H oxidation of α-aminocarbonyl compounds and arylation. The mild reaction has a broad reaction scope and gives desired α-aryl α-amino ketones in moderate to excellent yields. This journal is © The Royal Society of Chemistry.

Transition metal-catalyzed oxidative coupling reactions involving the formation of C–C bonds from C–H bonds have attracted considerable attention, indicating excellent atom economy and environmental friendliness.[1] α-Amino carbonyl compounds have important roles in natural products and are the key structural units of natural products.[2] These compounds have also been used in organic chemistry to synthesize biological activites, therapeutic agents, quinazolines, imidazoles, pyrazines, indoles, and pyrroles.[3] Therefore, the direct oxidative C–H functionalization has gained significant attention for the synthesis of a series of α-amino carbonyl compounds.[2] For example, Li's group employed an oxidative coupling reaction to synthesize α-amino carbonyl compounds from N-glycine derivatives by direct C–C bond formation under the catalysis of copper salts.[5] Subsequently, stoichiometric amounts of chemical oxidants, such as DTBP, DDQ, TBHP, and 2,2,6,6-tetramethylpiperidine-1-oxoammonium tetra-fluoroborate (T+BF4−), have been applied to these reactions.[4] In 2013, Yang's group described a novel protocol for a copper-catalyzed oxidative phosphonation reaction by using α-aminocarbonyls and diphenylphosphine ((1), Scheme 1).[7] Huang's group disclosed a general and efficient method for C–N oxidative cross-coupling through direct Csp–H bond functionalization of α-aminocarbonyl compounds with amines under the catalysis of copper salts ((2), Scheme 1).[6] In 2015, Yang's group developed a highly efficient route to synthetize chiral arylglycine derivatives via a palladium-catalyzed enantioselective direct C–H oxidation arylation reaction ((3), Scheme 1).[4] Furthermore, transition metal-catalyzed direct C–H functionalization by an oxidative cross-coupling reaction has been reported in the past few years.[8] Although significant advances have been made along these lines, the development of efficient synthetic methodologies for the synthesis of α-aminocarbonyl compounds via palladium-catalyzed oxidative cross-coupling still remains a challenging topic. Based on these considerable progresses, in this paper, we describe a highly efficient C–H oxidative cross-coupling reaction strategy for the synthesis of α-amino ketone compounds by palladium-catalyzed direct C–H oxidation and arylation reactions ((4), Scheme 1).

Scheme 1

Transition metal-catalyzed reaction for the synthesis of α-aminocarbonyl compounds.

In an initial study, we chose 2-((4-chlorophenyl)amino)-1-phenylethanone 1a and para-methylphenyl boric acid as the model substrate to evaluate different oxidants in the presence of 10 mol% Pd(OAc)2 with 2,2-bipyridine as a ligand in TFE at 60 °C (Table 1, entries 1–8). To our delight, the desired product 2a was obtained in 14% yield by using 2,2,6,6-tetramethylpiperidine-1-oxoammonium tetra-fluoroborate (T+BF4−)[4] as an oxidant (Table 1, entry 8). Based on these results, various ligands were used to carry out the reaction in the presence of 10 mol% Pd(OAc)2. As expected, the best result of 29% yield was obtained by employing L3 as a ligand (Table 1, entries 9–15). Then, different solvents were screened; using CH3OH as the solvent with the set reaction conditions gave comparable results (entry 25), but others gave lower yields (Table 1, entries 16–25). When the temperature was increased to 80 °C, the yield of 2a reached 71% (Table 1, entries 26 and 27). To our delight, the reaction could occur in the presence of 10 mol% of catalysts such as Pd(NO3)2, Pd(TFA)2, PdCl2, Pd(PPh3)2Cl2, Pd(PPh3)4, Pd(CH3CN)2Cl2, and Pd(acac)2, while the reactivity of Pd(PCy3)2Cl2 was better than others, affording the desired product 2a in 86% yield (Table 1, entries 28–37). Furthermore, control experiments showed that no or trace amounts of the desired product was obtained in the absence of Pd(PCy3)2Cl2 or T+BF4− (Table 1, entries 38 and 39).

Optimization of the reaction conditionsa,b

Entry	Catalyst	Ligand	Oxidant	Solvent	Yieldb
1	Pd(OAc)2	bpy	BQ	TFE	10%
2	Pd(OAc)2	bpy	Ag2CO3	TFE	13%
3	Pd(OAc)2	bpy	K2S2O8	TFE	Trace
4	Pd(OAc)2	bpy	Air	TFE	8%
5	Pd(OAc)2	bpy	PhI(OAc)2	TFE	Trace
6	Pd(OAc)2	bpy	Ph3CBF4	TFE	NR
7	Pd(OAc)2	bpy	C7H7BF4	TFE	Trace
8	Pd(OAc)2	bpy	T+BF4−	TFE	14%
9	Pd(OAc)2	L1	T+BF4−	TFE	24%
10	Pd(OAc)2	L2	T+BF4−	TFE	13%
11	Pd(OAc)2	L3	T+BF4−	TFE	29%
12	Pd(OAc)2	L4	T+BF4−	TFE	15%
13	Pd(OAc)2	L5	T+BF4−	TFE	20%
14	Pd(OAc)2	L6	T+BF4−	TFE	12%
15	Pd(OAc)2	PPh3	T+BF4−	TFE	16%
16	Pd(OAc)2	L3	T+BF4−	DCE	21%
17	Pd(OAc)2	L3	T+BF4−	THF	31%
18	Pd(OAc)2	L3	T+BF4−	TOL	Trace
19	Pd(OAc)2	L3	T+BF4−	CH3NO2	16
20	Pd(OAc)2	L3	T+BF4−	DCM	18%
21	Pd(OAc)2	L3	T+BF4−	C2H5OH	37%
22	Pd(OAc)2	L3	T+BF4−	DME	40%
23	Pd(OAc)2	L3	T+BF4−	DMF	8%
24	Pd(OAc)2	L3	T+BF4−	Dioxane	32%
25	Pd(OAc)2	L3	T+BF4−	CH3OH	47%
26c	Pd(OAc)2	L3	T+BF4−	CH3OH	23%
27d	Pd(OAc)2	L3	T+BF4−	CH3OH	71%
28d	Pd(NO3)2	L3	T+BF4−	CH3OH	56%
29d	Pd(TFA)2	L3	T+BF4−	CH3OH	43%
30d	PdCl2	L3	T+BF4−	CH3OH	67%
31d	Pd(PPh3)2Cl2		T+BF4−	CH3OH	60%
32d	Pd(PPh3)4		T+BF4−	CH3OH	38%
33d	Pd(CH3CN)2Cl2		T+BF4−	CH3OH	74%
34d	Pd(acac)2		T+BF4−	CH3OH	58%
35d	Pd(PhCN)2Cl2		T+BF4−	CH3OH	63%
36d	Pd(cod)Cl2		T+BF4−	CH3OH	66%
37 d	Pd(PCy 3 ) 2 Cl 2		T + BF 4 −	CH 3 OH	86%
38d	Pd(PCy3)2Cl2			CH3OH	Trace
39d			T+BF4−	CH3OH	No

Reaction conditions: 1a (0.1 mmol), para-methyphenyl boric acid (1.2 equiv.), catalyst (10 mol%), ligand (10 mol%) and oxidant (1.2 equiv.) was stirred in solvent (1 mL) at 60 °C under Ar for 20 h.

Yield of the isolated product.

100 °C.

80 °C.

With the optimal reaction conditions in hand (Table 1, entry 37), we explored the C–H oxidative cross-coupling reaction of 2-((4-chlorophenyl)amino)-1-phenylethanone 1a with arylboric acids, as shown in Table 2. We first surveyed different substituents of arylboric acids with electron-donating groups, such as methyl, ethyl, isopropyl and methoxy, and found that they gave the desired product in 80–86% yields (Table 2, entries 2a–2d). Meanwhile, the steric effect was examined using the meta- and ortho-methyl phenylboric acids under identical conditions (Table 2, entries 2e and 2f). However, the steric effect in this transformation was very significant; only trace amounts of the product was obtained when ortho-methyl phenylboric acids were introduced for the optimization of reaction conditions (Table 2, entry 2f). When arylboric acids with different electron-donating or electron-withdrawing groups afforded the desired products in excellent to moderate yields (Table 2, entries 2g–2m). Moreover, in order to further expand the substrate scope, we selected potassium phenyltrifluoroborate as the arylated reagent under the optimized reaction conditions; the corresponding α-alkylation product 2i was obtained in 45% yield (Table 2, entry 2i).

Reaction conditions screeninga,b

Reaction conditions: 1a (0.1 mmol), para-methyphenyl boric acid (1.2 equiv.), Pd(PCy3)2Cl2 (10 mol%), and 2,2,6,6-tetramethylpiperidine-1-oxoammonium tetra-fluoroborate (T+BF4−) (1.2 equiv.) was stirred in CH3OH (1 mL) at 80 °C under Ar for 20 h.

Yield of the isolated product.

Potassium phenyltrifluoroborate as arylated reagents.

Furthermore, the naphthalen-1-ylboronic acid and benzo[1,3]dioxol-5-ylboronic acid could also afford α-aminocarbonyl compounds 2n and 2o in 73–77% yields (Table 2, entries 2n and 2o). Of particular note is the heterocyclic boronic acid, which was also compatible for the reaction (Table 2, entries 2p and 2q). Moreover, the introduction of various electron-withdrawing or electron-donating substituents on the aniline moeity gave the corresponding α-aminocarbonyl compounds in 30–88% yields (Table 2, entries 2r–2x); the electronic effect and the steric effect in this transformation was very notable (Table 2, entries 2t–2v). Next, different substituent groups of α-carbonyl compounds bearing different functional groups were additionally examined and the corresponding products were generated in moderate yields (Table 2, entries 2y and 2z).

To investigate the mechanism of this transformation, experiments were carried out. The desired product was obtained in the range of 86% to 65% and 86% to 45% yield when 2.0 equivalents of radical-trapping reagents 2,2,6,6-tetramethyl-1-piperidinyloxy (TEMPO) and 2,6-di-tert-butyl-4-methylphenol (BHT) were used, respectively, under standardized reaction conditions (Scheme 2). To our delight, the key α-imino intermediate A was detected by GC-MS (see ESI†). Based on the observed experimental results and pioneering reports,[4] we have described a plausible mechanistic pathway in Scheme 3. Initially, the arylpalladium intermediate B was produced via a transmetallation reaction of Pd(PCy3)2Cl2 with aryl boric acid, which attacks the α-imino intermediate A obtained by the in situ oxidation of 1a by T+BF4− to form the complex C. Then, an aryl group was added to the imine to generate intermediate D. Finally, the product 2a was obtained upon dissociation in the presence of H+. At the same time, the palladium catalyst was regenerated and synchronized into the next catalytic cycle (Scheme 3).

Scheme 2

Radical-trapping experiment.

Scheme 3

Proposed mechanism.

In summary, we have achieved a novel pattern for the synthesis of α-aryl α-amino ketone compounds via Pd(ii)-catalyzed oxidative coupling of α-aminocarbonyl compounds with arylboric acids. This reaction occurs via direct C–H oxidation and arylation reactions. The coupling of α-aminocarbonyl compounds gave functionalized α-aryl α-amino ketone compounds in moderate to excellent yields.

Conflicts of interest

There are no conflicts to declare.