Transition-metal and oxidant-free approach for the synthesis of diverse N-heterocycles by TMSCl activation of isocyanides.

A highly efficient TMSCl-mediated addition of N-nucleophiles to isocyanides has been achieved. This transition-metal and oxidant-free strategy has been applied to the construction of various N-heterocyles such as quinazolinone, benzimidazole and benzothiazole derivatives by the use of distinct amino-based binucleophiles. The notable feature of this protocol includes its mild reaction condition, broad functional group tolerance and excellent yield. This journal is © The Royal Society of Chemistry.

In the past decades, isocyanides have proved themselves to be irreplaceable structural scaffolds in organic synthesis.[1] The chemistry of isocyanides is characterized by the great diversity of transformations that includes multicomponent reactions (MCRs, such as Passerini and Ugi reaction),[2] transition metal-catalyzed insertions (also called imidoylative reaction),[3] as well as isocyanide-mediated radical cascade reactions.[4] Generally, the isocyanide group can act as a mild nucleophile by electrophilic activation in the presence of carbonyl, imine or transition-metal catalysts, which allow further transformations after the incorporation of isocyanide core into starting material (Scheme 1a). In contrast, the reactions of isocyanides with external nucleophiles are particularly challenging because of the poor electrophilicity of isocyanides, and most of these reactions require highly reactive organometallic nucleophiles (Scheme 1b).[5] Only a few reports achieved the direct additions of weak nucleophiles to isocyanides by Lewis acid complexation[6] or NHC catalyst (Scheme 1b).[7] Therefore, the development of new catalyst system for the activation of isocyanide as electrophilic reagent would be highly desirable.

Scheme 1

Strategies for isocyanide activation.

On the other hand, nitrogen-containing heterocycles are invaluable building blocks in organic chemistry and are considered to be “privileged” structure in medicinal chemistry.[8] In this context, the construction of N-heterocycles has been a major research topic in synthetic chemistry.[9] Among these reports, isocyanides have emerged as C1 synthons for the synthesis of various N-heterocycles via isocyanide insertion reactions[10] (similar to carbon monoxide[11]). For example, bisnucleophile agents A could be applied to the synthesis of N-heterocycles B through isocyanide insertion-cyclizations by the use of transition metals (such as Pd, Co, Ni, etc.)[12] or I2/TBHP catalytic system[13] (Scheme 1c). However, these reports suffer from the use of expensive transition metals or peroxide reagents. Meanwhile, in light of the success of Lewis acid promoted nucleophilic additions to isocyanides. We envisaged that the use of Lewis acid might catalyse the nucleophilic addition of A to isocynide,[14] and subsequent cyclization of the formamidine intermediate could deliver the corresponding N-heterocycles C (Scheme 1d). Thus, an unprecedented transition-metal and oxidant-free approach to access various N-heterocycles using isocynide as C1 source could be achieved.

Our study commenced with the reaction between 4-methylaniline (1a) and tert-butyl isocyanide (2a) in acetonitrile at 70 °C. A survey of reaction parameters was summarized in Table 1. First, no desired product was observed in the absence of Lewis acid catalyst (Table 1, entries 1). Then, 1.0 equivalent of CuCl was selected as the Lewis acid based on the literature report,[14] formamidine product 3a could be obtained in 50% yield after stirring for 24 h (entry 2). Then, a series of transition metal-based Lewis acids such as AgCl, FeCl3 and ZnCl2 were also evaluated in the same reaction condition, and the results were still unsatisfactory (entries 3–5). Next, we chose Brønsted acids[15] such as CF3COOH, and TfOH as the activation reagents for this reaction (entries 5–7). Only a trace mount of formamidine 3a was detected along with unreacted starting material. Fortunately, in the presence of BF3·Et2O, the reaction could afford the corresponding product 3a in 55% yield (entry 8). Surprisingly, further optimization of the reaction conditions revealed silicon-based Lewis acid TMSCl could catalyse the reaction with 85% yield (entry 9).[16] To the best of our knowledge, the nucleophilic activation of isocyanides using silicon-based Lewis acid has not yet been reported.[17] Meanwhile, catalyst loading had obvious effects on the reaction yields. A slightly increased yield was observed with 1.5 equiv. of TMSCl, while decreasing the amount of TMSCl to 0.5 equiv. resulted in a lower yield of 3a (entries 10 and 11). A survey of other reaction media revealed that the overall results could not be improved (entries 12–14). In addition, a lower yield was obtained when the reaction was performed at room temperature (entry 15). Finally, formamidine 3b could also be obtained in high yield using 4-chloroaniline 1b as nucleophile (entry 16).

Optimization of the reaction conditionsb

Entry	Catalyst (equiv.)	Temperature (°C)	Solvent	Product	Yieldb (%)
1	—	70	CH3CN	3a	0
2	CuCl (1.0)	70	CH3CN	3a	50
3	AgCl (1.0)	70	CH3CN	3a	Trace
4	FeCl3 (1.0)	70	CH3CN	3a	Trace
5	ZnCl2 (1.0)	70	CH3CN	3a	55
6	CF3COOH (1.0)	70	CH3CN	3a	0
7	TfOH (1.0)	70	CH3CN	3a	10
8	BF3·Et2O (1.0)	70	CH3CN	3a	55
9	TMSCl (1.0)	70	CH3CN	3a	85
10	TMSCl (1.5)	70	CH 3 CN	3a	90
11	TMSCl (0.5)	70	CH3CN	3a	71
12	TMSCl (1.5)	70	DCE	3a	79
13	TMSCl (1.5)	70	THF	3a	59
14	TMSCl (1.5)	70	Toluene	3a	80
15	TMSCl (1.5)	rt	CH3CN	3a	52
16	TMSCl (1.5)	70	CH3CN	3b	92

Reaction conditions: 1 (0.2 mmol), 2a (0.3 mmol), catalyst (0.5–1.5 equiv.), solvent (2 mL), 24 h.

Isolated yields.

With the optimal conditions in hand, we applied this strategy to the synthesis of various quinazolinones[18] by employing 2-aminobenzamides 4 as bisnucleophile agents (Table 2). In general, the reaction works well when R1 was an aromatic group. Substituents at para-positions bearing either electron-donating or electron-withdrawing groups can afford the desired products in good to excellent yields (5a–5f). The cyclization products with substituents at meta-positions were also obtained in good yields (5g, 5h), while lower yield was observed with substituent at ortho-position (5i). Then, substrates with aliphatic groups, such as methyl, n-propyl, benzyl, propargyl, etc., were also employed in this reaction to give the corresponding products in 84–90% yields (5j–5o). Next, 2-aminobenzamides with various R2 groups were evaluated in the standard condition, and functionalized quinazolinones were generated in 88–93% yields (5p–5u). It is worth noting that 2-aminobenzene sulfonamide could also be tolerated in this reaction, affording the cyclization products 5v in 87% yield.

Substrate scope for the synthesis of various quinazolinonesa

Reaction conditions: 4 (0.2 mmol), 2a (0.3 mmol), TMSCl (1.5 equiv.), CH3CN (2 mL), 70 °C, 24 h. Isolated yields.

The scope of this methodology has been also extended to the synthesis of other N-heterocycles by simply changing the amino-based binucleophiles (Table 3). First, diverse o-phenylene-diamines were subjected to the same reaction conditions. To our delight, the reaction proceed smoothly in all cases regardless of the electronic and steric properties of the substituents, giving the corresponding 1H-benzo[d]imidazole derivatives[19] in moderate to good yields (7a–7m). Furthermore, N-methyl and N-phenyl-o-phenylenediamine were also tolerated in this reaction, delivering 2-aminobenzimidazole 7n and 7o in 84% and 80% yields respectively. It is worth noting 2-amino-benzenethiol could undergo the same transformation to furnish benzo[d]thiazole product 7p in 92% yield. However, the reaction failed to generate benzo[d]oxazole 7q with o-aminophenol under identical condition. Finally, diversified facile synthesis of benzimidazo[1,2-c]quinazolines 7r and 7s could be achieved in reasonable yields.

Substrate scope for the synthesis of other N-heterocyclesa

Reaction conditions: 6 (0.2 mmol), 2a (0.3 mmol), TMSCl (1.5 equiv.), CH3CN (2 mL), 70 °C, 24 h. Isolated yield.

2.0 equiv. of TMSBr in 2 mL C2H5OH was used.

To gain an insight into the reaction mechanism, several control experiments were performed as presented in Scheme 2. Initial radical inhibition studies using TEMPO and BHT indicated that the reaction does not proceed through a radical pathway (Scheme 2a). The reaction of 2-aminobenzamides 4a with 2a by the standard condition under N2 provided 5a in 86% yield, revealing that oxygen is not participated in this reaction (Scheme 2b). In the meantime, the generation of BuNH2 as byproduct was confirmed by GC-MS.[20]

Scheme 2

Control experiments. (a) Radical inhibiton studies. (b) Standard conditon under N2 conditions.

The following reaction mechanism is proposed based on our experimental observations and previous literature reports.[20] First, nucleophilic addition of bisnucleophile agents A to tert-butyl isocyanide 2avia TMSCl activation could generate formamidine intermediate D. Then intramolecular nucleophilic addition of formamidine D could deliver the cylization intermediate E. Finally, β-elimination of intermediate E could afford the desired product C along with byproduct BuNH2 (Scheme 3).

Scheme 3

Plausible reaction mechanism.

The present activating strategy was also applied to the synthesis of a biologically active molecule Erlotinib (FDA-approved tyrosine kinase inhibitor).[21] The reaction of starting material 8 with isocyanide 2a was performed under the standard condition, affording the key intermediate 9 in 92% yield. Subsequent chlorination and amination reactions could afford Erlotinib in 74% yield over two steps (Scheme 4).

Scheme 4

Synthesis of biologically active compounds.

Conclusions

In conclusion, we have developed an efficient silicon-based Lewis acid system for the activation of isocyanides. Based on this strategy, a new robust transition-metal and oxidant free method for the construction of various N-heterocycles could be realized using isocyanide as methine source. Quinazolinone, benzoimidazole, and benzothiazole derivatives could be obtained in good to excellent yields under mild conditions. The present strategy opens a powerful pathway for the activation of isocyanides, and further studies on the application of this methodology are currently underway.

Conflicts of interest

There are no conflicts to declare.