Xu Meng,
Yanmin Wang,
Chaoying Yu and
Peiqing Zhao*
State Key Laboratory for Oxo Synthesis and Selective Oxidation, Lanzhou Institute of Chemical Physics, Chinese Academy of Sciences, Lanzhou 730000, P. R China. E-mail: zhaopq@licp.cas.cn; Fax: +86 931 8277008; Tel: +86 931 4968688
First published on 29th May 2014
An efficient and mild heterogeneously CuCl2/nano-TiO2-catalyzed aerobic synthesis of imidazo[1,2-a]pyridines from 2-aminopyridines and ketones has been developed using air as the oxidant in the absence of any ligands and additives. This strategy was compatible with a large range of substrates, including unactivated aryl ketones and unsaturated ketones and went through the C–H bond functionalization mechanism instead of I−-assisted Ortoleva-King reaction to provide the corresponding imidazo[1,2-a]pyridines in good yields with low catalyst loading (0.8 mol%). Moreover, the heterogeneous catalyst can be successfully employed in gram-scale synthesis and reused many times without the significant loss of catalytic activity.
Imidazo[1,2-a]pyridines are valuable motifs and exhibit many important biological activities, like antibacterial,7 antiviral,8 antitumor9 and anti-inflammatory properties.10 Therefore, imidazo[1,2-a]pyridines are widely present in a number of best-selling pharmaceuticals,11 such as zolimidine,12 zolpidem,12 alpidem,13 olprinone,14 necopidem,15 saripidem.15 Traditionally, imidazo[1,2-a]pyridines were obtained by condensations between 2-aminopyridines and per-functionalized carbonyl compounds under various conditions (Scheme 1, eqn (1)).16 Recently, Lei's group discovered a silver-mediated oxidative coupling/cyclization using 2-aminopyridines and terminal alkynes to provide imidazo[1,2-a]pyridines (Scheme 1, eqn (2)).17
In 2013, Hajra's group reported a Fe-catalyzed method using 2-aminopyridines and nitroolefins (Scheme 1, eqn (3)).18 Subsequently, several homogeneous CuI or Cu/ZnI-catalyzed oxidative methods using 2-aminopyridines and unactivated methyl ketones in the presence of ligands or additives were reported, respectively, by Hajra, Adimurthy, Ji and Kumar's group at almost the same time (Scheme 1, eqn (4)).19 Most recently, Su and co-workers discovered a homogeneous CuI/O2 system for the synthesis of imidazo[1,2-a]pyridines by a reaction between 2-aminopyridines and unactivated methyl ketones and disclosed an iodine-promoted Ortoleva-King reaction rather than the previously reported C–H functionalization, which was involved in this transformation most probably (Scheme 1, eqn (4)).20 In the course of investigating applications of heterogeneous catalysts in our research group,4a,e,21 we proposed a simple heterogeneous system for the synthesis of imidazo[1,2-a]pyridines from unactivated ketones and 2-aminopyridines via metal-catalyzed aerobic C–H functionalization under ligand-, additive- and iodine-free conditions, which would be a useful complement to the previous methodologies.22
Here, we wish to describe a heterogeneous CuCl2/nano-TiO2-catalyzed aerobic synthesis of imidazo[1,2-a]pyridines from 2-aminopyridines and readily available ketones using air as the oxidant under iodine-, ligand- and additive-free conditions (Scheme 2). The catalytic system tolerates a variety of substrates and could be employed in the synthesis of zolimidine on a gram-scale. More importantly, the heterogeneous catalyst can recycle three times without the loss of activity.
| Entry | Catalyst | Solvent | Temp. (°C) | Yieldb (%) |
|---|---|---|---|---|
| a Reaction conditions: 2-aminopyridine (0.6 mmol), acetophenone (0.5 mmol), heterogeneous catalyst (12 mg, 0.8 mol%), solvent (0.6 mL), air, 24 h.b Isolated yields.c 10 mol% of CuCl2-2H2O was used.d 12 mg of nano-TiO2 was used.e 10 mol% of CuCl2-2H2O and 12 mg of nano-TiO2 were used. | ||||
| 1 | Cu–Zn/Al–Ti | DCB | 100 | 0 |
| 2 | Cu–Zn/Al–Ti | DCB | 130 | <5 |
| 3 | Cu–Mn/Al–Ti | DCB | 120 | 0 |
| 4 | Fe–Zn/Al–Ti | DCB | 120 | 0 |
| 5 | Cu–Zn/nano-TiO2 | DCB | 130 | 8 |
| 6 | Cu–Ni/nano-TiO2 | DCB | 130 | <5 |
| 7 | Ni–Zn/nano-TiO2 | DCB | 130 | 0 |
| 8 | Cu–Fe/OMS-2 | DCB | 130 | <5 |
| 9 | Cu(OH)x/OMS-2 | DCB | 130 | <5 |
| 10 | CuCl2/nano-TiO2 | DCB | 130 | 88 |
| 11c | CuCl2 | DCB | 130 | 12 |
| 12d | Nano-TiO2 | DCB | 130 | 0 |
| 13e | CuCl2 + nano-TiO2 | DCB | 130 | 15 |
| 14 | Cu(OH)x/nano-TiO2 | DCB | 130 | 23 |
| 15 | CuOx/nano-TiO2 | DCB | 130 | 0 |
| 16 | Ru/nano-TiO2 | DCB | 130 | 0 |
| 17 | Pt/nano-TiO2 | DCB | 130 | 0 |
| 18 | CuCl2/nano-TiO2 | DMSO | 130 | 16 |
| 19 | CuCl2/nano-TiO2 | DMF | 130 | <5 |
| 20 | CuCl2/nano-TiO2 | o-Xylene | 130 | 12 |
| 21 | CuCl2/nano-TiO2 | Dioxane | 100 | 65 |
| 22 | CuCl2/nano-TiO2 | MeNO2 | 100 | <5 |
| 23 | CuCl2/nano-TiO2 | THF | 60 | <5 |
| 24 | CuCl2/nano-TiO2 | EtOH | 70 | 87 |
The supported Cu(OH)x on manganese oxide-based octahedral molecular sieve (Cu(OH)x/OMS-2), which is an excellent heterogeneous catalyst for biomimetic homo-coupling of terminal alkynes,23a and basic Cu(OH)x/nano-TiO2, which can successfully catalyze Huisgen [3 + 2] cycloaddition,23b are not efficient for this cyclization, although the latter gave 23% yield (Table 1, entries 9 and 14). On the other hand, the copper oxide does not have catalytic activity at all in the experiment (Table 1, entry 15). More experiments showed that no reaction proceeded or low yield was obtained in the presence of nano-TiO2 powder alone or the catalyst precursor of CuCl2-2H2O (Table 1, entries 11 and 12). Moreover, the catalytic activity of a mixture of CuCl2-2H2O and nano-TiO2 was much lower than that of the CuCl2/nano-TiO2 (Table 1, entry 13), which means that the highly dispersed copper species are effective for this cyclization. Specifically, CuCl2 can catalyze the reaction while nano-TiO2 does not have any catalytic activity; therefore, nano-TiO2 probably plays a role of an appropriate support for the heterogeneous catalyst because of its large surface area, small particle size and the interaction between Cu species and TiO2.24 To improve the yield of reaction, other metal oxides were used, such as Ru and Pt, although the results were negative (Table 1, entries 16 and 17). Next, we found that the solvent could affect the reaction significantly (Table 1, entries 18–24). Specifically, polar organic solvents, like DMF and DMSO, hardly assisted with the reaction while only dioxane brought about moderate yield of the desired product among other examined solvents (Table 1, entries 18–23). Surprisingly, the best result was obtained by using protic EtOH as a solvent under mild conditions (Table 1, entry 24).
After confirming the catalytic transition metal, a lot of supports were examined for optimization as well (Table 2). In Table 2, we used the impregnation method to load CuCl2 on various solid supports and ran the cyclization under the established reaction conditions: EtOH as a solvent at 70 °C under air for 24 h. As a consequence, nano-size Al2O3, ZrO2 and CeO2 were not efficient supports, although they catalyzed the reaction slightly (Table 2, entries 2–4). SiO2, OMS-2 (manganese oxide-based octahedral molecular sieve), activated carbon and ATP (attapulgite) gave traces to approximately 10% yield (Table 2, entries 5–7 and 9) while CuCl2/Al–Ti gave moderate yield of the desired product (Table 2, entry 8). Finally, it was proven that nano-TiO2 is the best support for the heterogeneous catalyst under mild ligand- and additive-free conditions (Table 2, entry 1).
| Entry | Catalyst (0.8 mol%) | Isolated yield (%) |
|---|---|---|
| a Reaction conditions: 2-aminopyridine (0.6 mmol), acetophenone (0.5 mmol), heterogeneous catalyst (12 mg, 0.8 mol%), EtOH (0.6 mL), 70 °C air, 24 h. | ||
| 1 | CuCl2/nano-TiO2 | 87 |
| 2 | CuCl2/nano-Al2O3 | 15 |
| 3 | CuCl2/nano-ZrO2 | 26 |
| 4 | CuCl2/nano-CeO2 | 10 |
| 5 | CuCl2/SiO2 | 8 |
| 6 | CuCl2/OMS-2 | 11 |
| 7 | CuCl2/C | <5 |
| 8 | CuCl2/Al–Ti | 34 |
| 9 | CuCl2/ATP | <5 |
To verify whether the observed catalysis is derived from the solid catalyst CuCl2/nano-TiO2 or the leached copper species, the cyclization of 1a and 2a was carried out, and then the catalyst was removed after filtering a totally converted reaction mixture (acetophenone as a substrate). Next, 1.0 equiv. of a different ketone (2-acetylpyridine) was added as an additional substrate to the filtrate, and then the filtrate was treated with the remaining amount of 2-aminopyridine (>1.2 equiv.). Consequently, only trace amounts of another product 3u was discovered, while 66% yield of 3u was obtained, if the fresh CuCl2/nano-TiO2 was added to the filtrate (Scheme 3). ICP-AES was used to analyze the filtrate that was removed by CuCl2/nano-TiO2, which confirmed that no copper species was detected in the filtrate (Cu: below 0.001%). These results indicated that the catalysis is from the solid catalyst rather than the leached metal species, and the catalyst is unambiguously heterogeneous in nature.
After the cyclization was completed, CuCl2/nano-TiO2 was easily retrieved from the reaction mixture by centrifugation and filtration. The retrieved catalyst could be reused at least three times with a slight decrease in catalytic ability (Table 3).
| Cycle | 1 | 2 | 3 | 4 |
|---|---|---|---|---|
| a Reaction conditions: 2-aminopyridine (0.6 mmol), acetophenone (0.5 mmol), CuCl2/nano-TiO2 (12 mg, 0.8 mol%), ethanol (0.6 mL), air, 70 °C, 24 h. | ||||
| Isolated yield of 3b (%) | 87 | 79 | 72 | 53 |
With the optimized conditions in hand, we sought to define the scope of the aryl ketones using 2-aminopyridine 1a as the substrate. As shown in Table 4, electron-rich and -poor aryl ketones could undergo cyclizations with 2-aminopyridine successfully, including sensitive NO2-substituted acetophenone. Specifically, electron-poor ketones could give the corresponding products in excellent yields, while electron-poor ketones offered relatively low yields of imidazo[1,2-a]pyridines. Moreover, steric hindrance could influence the reaction, which made ortho-substituted aryl ketones give moderate yields of the desired products (Table 4, 3d, 3j and 3p). Interestingly, the ester group and N,N-dimethyl-substituted ketones also performed smoothly (Table 4, 3m and 3q). Remarkably, cyano was tolerated in the cyclization chemoselectively and offered excellent yield of imidazo[1,2-a]pyridine rather than 1,2,4-triazole under Cu-catalyzed aerobic oxidative conditions (Table 4, 3r).25 Finally, heteroaromatic ketones, such as furan, thiophene, pyridine and thiazole, were effective substrates and provided good yields respectively (Table 4, 3s–v).
| a Reaction conditions: 1a (0.6 mmol), 2 (0.5 mmol), CuCl2/nano-TiO2 (0.8 mol%, 12 mg), EtOH (0.6 mL), air, 70 °C, 24 h, isolated yields. | ||
|---|---|---|
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
Next, more efforts were made to further explore the scope of the heterogeneous cyclization. In Table 5, alkenyl-substituted imidazo[1,2-a]pyridine, which was obtained with difficulty by previous methods,16m,n was synthesized by this simple heterogeneous method using α,β-unsaturated ketone (Table 5, 4a). Furthermore, propiophenone was also a successful substrate for the reaction (Table 5, 4b). Subsequently, various substituted 2-aminopyridines were investigated and provided moderate to good yields (Table 5, 4c–h), including CF3- and multi-substituted 2-aminopyridines. Unfortunately, nitro substituted 2-aminopyridine could not participate in the cyclization (Table 5).
| a Reaction conditions: 1 (0.6 mmol), 2 (0.5 mmol), CuCl2/nano-TiO2 (12 mg, 0.8 mol%), EtOH (0.6 mL), 70 °C, 24 h, air, isolated yields. | ||
|---|---|---|
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
To demonstrate the operational simplicity and broad generality of this heterogeneous protocol, a gram-scale reaction was carried out using 2-aminopyridine and 4-methylsufonyl acetophenone under slightly modified reaction conditions (Scheme 4). The synthesis of the marketed drug zolimidine 5 was achieved on a gram-scale with 56% isolated yield under heterogeneous CuCl2/nano-TiO2-catalyzed ligand- and additive-free conditions.
Finally, some control experiments were conducted to further understand the mechanism of this heterogeneous copper-catalyzed oxidative cyclization (Scheme 5). The cyclizations were carried out under different atmospheres and results showed that the products were obtained under air or O2, while it failed to yield the desired product efficiently under N2, which indicated that the oxygen plays a role of an oxidant in the synthesis of imidazo[1,2-a]pyridines (Scheme 5, eqn (1)). According to the previous methods of homogeneous Cu-catalyzed synthesis of imidazo[1,2-a]pyridines from 2-aminopyridines and ketones,19 especially Su's latest work on CuI-catalysis,20 it is reasonable that the reaction mechanism of the previous copper-catalysis involved a catalytic Ortoleva-King reaction16g more probably rather than the copper-catalyzed C–H functionalization because of the use of iodine in the previous methods.22g,26 To prove this proposal, the cyclizations were conducted directly using I2 without any copper source (Scheme 5, eqn (2)). Interestingly, the cyclizations preformed smoothly no matter how much I2 was used in the reaction, which means that a-iodoacetophenone rather than imine is formed firstly for further Ortoleva-King reaction and copper does not play a crucial role when the cyclization is catalyzed by CuI or Cu/ZnI systems. More importantly, when the imine formed from 2-aminopyridine and acetophenone was employed as a starting substrate, the desired product was obtained in 78% yield under a heterogeneous CuCl2/nano-TiO2 system. On the other hand, the reaction failed when a catalytic amount of I2 was used as a catalyst (Scheme 5, eqn (3)). Under our optimized conditions, the absence of iodine anion rules out the possibility of the Ortoleva-King reaction in terms of mechanism.
Based on the observations and summary from control experiments and the works of Su and Han's group,20,27 a plausible mechanism involving oxidative C–H functionalization is postulated in Scheme 6.19 Firstly, imine A was generated via the condensation of 2-aminopyridine and acetophenone. Then, enamine B was formed by tautomerization of imine A.28 Subsequently, intermediate C was formed via oxidative addition between Cu(II) species and enamine B under the present conditions.29 After reductive elimination, the desired product imidazo[1,2-a]pyridine was given along with the reduced Cu(I) species, which was oxidized by oxygen to finish the catalytic cycle.
:
ethyl acetate = 3
:
1 as an eluent) to yield the corresponding product. The identity and purity of the products was confirmed by 1H and 13C NMR spectroscopic analysis.
Footnote |
| † Electronic supplementary information (ESI) available. See DOI: 10.1039/c4ra03299c |
| This journal is © The Royal Society of Chemistry 2014 |