Primary aromatic and heteroaromatic amides constitute an important class of carboxylic acid derivatives. Owing to their facile further functionalization, for example, their reduction into primary amines, dehydration into nitriles, the formation of heterocycles and amino acid derivatives, as well as because of their inherent bio-activity, the selective synthesis of benzoic acid amides continues to attract significant attention in synthetic organic chemistry. (Scheme 1 and 2).1–4
A number of useful methodologies have been developed for the selective formation of primary aromatic amides (Scheme 3).5–8 For example, hydration of their corresponding benzonitriles,5 or the straightforward conversion of benzoic acids or acid chlorides with ammonia.6 Other, less widely used synthetic strategies involved the rearrangement of benzaldoximes,7 or oxidations of benzyl amines, benzyl alcohols, or benzaldehydes.8 In addition, organometallic carbonylation reactions have also been successfully used in the synthesis of amides.9 In this respect, the palladium-catalyzed Heck carbonylation of aryl halides has become a powerful tool for the synthesis of all kinds of substituted aromatic and vinylic carboxylic acid derivatives.10 However, there are only a few known methods for the synthesis of synthetically useful primary amides. In general, in these reactions ammonia synthons are employed as a nucleophilic partner for the carbonylation. Some such synthons include hexamethyldisilazane (HMDS), formamide, N-tert-butylamides, hydroxylamine, and even a titanium–nitrogen complex has been described.11 Surprisingly, until very recently, the straightforward use of ammonia in Heck carbonylations was not reported because of the low nucleophilicity of ammonia and strong coordination with palladium(II).12, 13
As part of our continuing interest in palladium-catalyzed carbonylation reactions,14 we very recently reported the first general synthesis of primary amides starting from aryl bromides or chlorides.15 Key to this success was the use of a palladium/nBuP(1-adamantyl)2 (cataCXiumA) catalyst system.16 In continuance of this work, herein we report an improved catalyst system based on the commercially available Pd(OAc)2 and dppf.
During investigations on the influence of different bidentate ligands on the aminocarbonylation of bromobenzene, which we examined as a benchmark reaction, we discovered that 1,1′-bis(diphenylphosphino)ferrocene (dppf) gave improved results in this transformation.17 Typically, catalytic experiments were performed at a low pressure of CO (2 bar) and NH3 (2 bar) at 100 °C in the presence of 2 mol % Pd(OAc)2 and 3 mol % of the respective ligand (Table 1). Whilst 1,2-bis(diphenylphosphino)ethane (dppe), 1,2-bis(diphenylphosphino)benzene, 1,5-bis(diphenylphosphino)pentane (dpppe), and 1,2-bis(diphenylphosphino)metane (dppm) gave discouragingly low yields, the addition of 1,2-bis(diphenyl-phosphinomethyl)benzene afforded moderate conversion and yield (Table 1, entries 1–5). To our delight, dppf afforded full conversion and 93 % yield of benzamide (Table 1, entry 6), which was slightly better than our previous system (86 % yield; Table 1, entry 12). Furthermore, in the presence of 1,3-bis(diphenylphosphino)propane (dppp), bis-[2-(diphenylphosphino)-phenyl]ether, xantphos, (+)-2,3-O-isopropylidene-2,3-dihydroxy-1,4-bis(diphenylphosphino)butane (diop), and 1,4-bis(diphenylphosphino)butane (dppb) good results were obtained (Table 1, entries 7–11). Other commercially available monodentate ligands gave only low conversions and yields (1–47 %; Table 1, entries 13–16). Notably, with this novel system also at significantly lower catalyst loadings an excellent product yield was maintained. Hence, in the presence of 0.2 mol % of Pd(OAc)2, benzamide was obtained in 96 % yield (Table 1, entry 17). Even without further optimization, we obtained 55 % of the amide with only 0.05 mol % of Pd(OAc)2 (Table 1, entry 18).
In Scheme 4, a proposed mechanism for the aminocarbonylation reaction is shown, based on mechanistic studies of similar palladium-catalyzed carbonylations.18 In agreement with previous studies, we suppose that the oxidative addition of the active palladium(0) species is rate-determining, because aryl chlorides react more sluggishly than aryl bromides (see Table 2). Clearly, ammonia plays an important role in this reaction both as reagent and base to regenerate the active catalyst.
With suitable reaction conditions in hand (Table 1, entry 6), more than twenty different aryl and heteroaryl halides were aminocarbonylated in the presence of dppf. In general, the corresponding primary amides were obtained in moderate to excellent yields (Table 2). There was no major difference in reactivity or selectivity with respect to the substitution pattern of the halide. Thus, ortho-, meta-, and para-alkyl-substituted benzenes were successfully transformed into their corresponding primary amides in high yields (80–98 %; Table 2, entries 2–5). Similarly, 1- and 2-bromonaphthalene both showed excellent selectivity (Table 2, entries 6 and 7). Notably, 2-bromonaphthalene was not efficiently transformed into the primary amide under our previous conditions. Both electron-rich and electron-deficient aryl bromides gave good yields (83–97 %) of the primary amide products (Table 2, entries 8–10). More difficult substrates such as cyano-, nitro-, and acetyl-functionalized aryl bromides furnished the desired amides in 37–54 % yield (Table 2, entries 11, 12, and 13). On the other hand, heteroaryl substrates were efficiently transformed and the reaction showed good functional group tolerance (Table 2, entries 14–15). For example, nicotinamide was prepared from 3-bromopyridine in 94 % yield (Table 2, entry 14).
Finally, our catalyst system was successfully applied in the aminocarbonylation of aryl chlorides (Table 2, entries 16–21). Activated, inactivated, and heterocyclic chlorides were converted into their corresponding primary amides in moderate to good yields (30–78 %). In general, the yields were better than in the presence of the palladium/nBuP(1-adamantyl)2 (cataCXiumA) system. In addition to aryl halides, phenyl triflate was also transformed into benzamide in high yield (80 %; Table 2, entry 22).
In summary, an improved and general palladium-catalyzed aminocarbonylation of aryl and heteroaryl halides with CO and ammonia has been established. High catalyst turnover numbers of up to 1100 have been obtained. Primary amides are accessible under comparably mild conditions in good yield. The applicability and functional-group tolerance of the presented system is shown in the aminocarbonylation of more than 22 different aryl and heteroaryl halides.