Ivaylo Slavcheva,
Jas. S. Wardb,
Kari Rissanenb,
Georgi M. Dobrikov*a and
Svilen Simeonov*ac
aInstitute of Organic Chemistry with Centre of Phytochemistry, Bulgarian Academy of Sciences, Acad. G. Bonchev Str., Bl. 9, Sofia 1113, Bulgaria. E-mail: svilen.simeonov@orgchm.bas.bg
bUniversity of Jyvaskyla, Department of Chemistry, Survontie 9 B, 40014 Jyväskylä, Finland
cResearch Institute for Medicines (iMed.ULisboa), Faculty of Pharmacy, Universidade de Lisboa, Av. Prof. Gama Pinto, 1649-003 Lisbon, Portugal
First published on 15th July 2022
The base-promoted direct amidation of unactivated esters is among the most useful reactions for amide bond formation in contemporary organic chemistry. The intensive research in this area has led to the development of a number of new methods to achive this transformation. However, to date, the existing literature is more methodological and in many instances lacks practical directions. Therefore, the full potential of this transformation is yet to be revealed by broadening the substrate scope. In a search for new practical applications of the amidation reaction, herein we present a comprehensive study of a number of base-promoted direct amidations that encompass a wide range of amines and esters. Furthermore, we applied our findings in the synthesis of phosphoramidates and several industrially relevant products.
Due to their availability, the direct amidation of esters5 is among the most attractive methods to achieve the synthesis of amides. Different methods have been developed (Scheme 1), utilizing different bases, solvents and reaction conditions (Scheme 1a–d),6–9 including settings for flow chemistry (Scheme 1e),10 and mixer mill (Scheme 1f).11
However, despite their straightforwardness at first glance, these transformations are often challenging. Among the already developed methods, some limitations can be pointed out. In many instances the existing literature suffers from non-systematic approach, fractionalized data, lack of diversity in the used substrates and a complete overview of the reaction's scope and limitations. Virtually there are no reports of unsuccessful experiments, which can create a misconception that the reaction is universal towards all esters and amines. To date, the practical application, i.e. the implementation of the new reaction setting for the synthesis of valuable synthetic products, has received less research focus, and thus needs additional diligence.
Phosphoramidates are another important class of compounds, which finds applications spanning agriculture,12 organic synthesis,13 pharmaceutical and medicinal chemistry.14,15 To date, the synthesis of phosphoramidates mainly rely on oxidative cross-coupling or reduction routes. The preparation of these compounds by redox neutral methods remains rather unexplored and suffers from various drawbacks, such as the use of toxic and sensitive chlorophosphonates,17 laborious inorganic salt elimination methods18–20 or the use of harsh conditions21 and transition metal catalysis22,23 (Scheme 1).
Driven by these considerations, herein we report a detailed study of direct amidation of unactivated esters with amines, promoted by strong bases. Our work represents a systematic study of over 75 reactions that encompass large substrate scope, ranging from anilines and heterocyclic amines to aliphatic amines and guanidines, and stretching to the synthesis of phosphoramidates. We performed a thorough examination of the scope and the limitations of the reaction settings and applied them in the synthesis of several important derivatives.
We started our investigation by exploring the reaction between methyl 3-methyl benzoate (1a) and 4-methyl aniline (2) (Scheme 2a). Under similar conditions to those, reported by Kim et al. we observed moderate yield of compound 3 (62%, Table 1, entry 1), which is in accordance with the literature data. The use of DMF as a solvent did not provide any significant advantage (Table 1, entry 2). In an attempt to solve this issue we decided to increase the basicity by using DMSO/t-BuOK system, which is reported to behave as a superbase.26 To the best of our knowledge DMSO/t-BuOK has never been reported as a reaction medium for the direct amidation of esters. A rational explanation of this observation is the possible competition between the amine nucleophile and the formed in situ sulphur ylide nucleophilic species27 (Scheme 2b). To our delight we observed significant improvement of the reaction yield up to 94% (Table 1, entry 3), without the formation of the competitive side product 4 (Scheme 2b) and only hydrolysis of 1a was observed in trace amount as a side reaction. The attempt to decrease the required equivalents of t-BuOK resulted in a significant decrease of the yield (Table 1, entries 4 and 5).
Scheme 2 (a) t-BuOK promoted direct amidation; (b) possible side reaction under DMSO/t-BuOK conditions. |
Solvent | Concentration (amine) | Eq. (ester) | Eq. (t-BuOK) | Yield | |
---|---|---|---|---|---|
a Reaction time 5 min. | |||||
1 | THF | 1 mmol ml−1 | 1.3 | 2.0 | 62% |
2 | DMF | 1 mmol ml−1 | 1.3 | 2.0 | 70% |
3 | DMSO | 1 mmol ml−1 | 1.3 | 2.0 | 94% |
4 | DMSO | 1 mmol ml−1 | 1.0 | 0.3 | Traces |
5 | DMSO | 1 mmol ml−1 | 1.0 | 1.5 | 84% |
Having these reaction conditions in hand, we explored the amine scope by using methyl 3-methylbenzoate (1a) as a model compound (Scheme 3). The reaction proceeded smoothly with various substituted anilines (2, 5–7). The presence of aliphatic, halogen or cyano-substituents at the aromatic ring was tolerated and the desired amides were isolated in moderate to excellent yields. The scope of the reaction was broadened by the acylation of two very weakly nucleophilic fluorinated anilines (8 and 9) in high yields. Notably, the challenging acylation of 2-aminopyrimidine (10) and the highly deactivated aminoanthraquinone (11) was achieved in 94% and 55% yield respectively. The heterocyclic amine 12 provided the desired product 31 in 30% yield. The reaction scope was extended with imidamide 13 and guanidine 14 and the heterocyclic aromatic amine 15 which were acylated in good yields. Nevertheless, the applicability of the method to aliphatic amines proved very limited. The only aliphatic amine that has been successfully acylated under these conditions was piperidine (16). Surprisingly, the use of tetramethylguanidine 22 was also unsuccessful.
We extended the scope of the method towards the acylation of methyl 2-methoxybenzoate (1b) with 2-aminopyrimidine (10) and N,N′-diphenylguanidine (23). The desired products 36 and 37 were formed in 63% and 85% yield respectively. CH-acidic esters methyl 2-phenylacetate (1c) and ethyl acetate (1d) were also successfully subjected to direct amidation with 2-aminopyrimidine (10) and 4-methylaniline (2) in moderate yields. However, the reaction of 2-aminopyrimidine with methyl 2-hydroxybenzoate (1e) and the attempt to amidate lactone 1f with aniline 2 were unsuccessful.
We methodically tested the classes of amine substrate (Scheme 4), which did not react or provided poor yields under the previous reaction conditions. In contrast with the DMSO/t-BuOK system benzyl amine (20), primary (18) and secondary (40) alkylamines, were successfully acylated, as well as alkyl-phenylamine (21), tetramethylguanidine (22), and N-monosubstituted piperazine (41). The products 31 and 35, which were previously obtained in poor yields, were isolated in significantly improved yields, 72% and 85% (Scheme 4) versus 30% and 71% (Scheme 3), respectively. The substrate scope was extended to the sterically hindered amines 42 and 43, morpholine 44, aminoalcohol 45, biguanide 46 and aniline 47. Acetamide 39 was obtained in a higher yield compared to the DMSO/t-BuOK system. The opening of lactone 1f and the amidation of 2-hydroxybenzoate 1e were also successfully providing the desired products 60 and 61 in high yields. The amidation of 3-nitro-2-chloro benzoate 1g with aniline 2 rendered product 62 in poor yield due to extensive formation of side products, showing that this reaction conditions do not tolerate nitro substituents. Esters containing carboxy (1h) and hydroxyl (1i) groups were also successfully amidated.
Despite the numerous methods to achieve the synthesis of phosphoramidates16 the direct reaction of amines with triethylphosphate (65), without any catalyst, to the best of our knowledge, has never been reported before. Driven by these considerations, we decided to go beyond the carboxylic esters and apply our findings in the amidation of readily available 65. To our delight, this approach was successful and we were able to synthetize phosphoramidates 70–74. Both aromatic and aliphatic amines were applicable as substrates providing the desired products in very high yields (Scheme 5). Noteworthy, the synthesis of similar phosphoramidates was achieved by Tf2O-promoted activating strategy. However, this approach provided only low to moderate yields.29
Despite its great potential this method suffers from several drawbacks. It is generally not applicable to important classes of substrates, such as aliphatic and benzylic amines. It's worth mentioning that in all the unsuccessful experiments, the competing reaction of the esters with the DMSO, derived sulfur ylide, leading to the formation of product 4 (Scheme 1). This observation hampers further optimization of the reaction conditions (time, temperatures, etc.), due to the fact that in the case of slowly reacting amines, all the ester is predominantly consumed in the faster competitive reaction.
In contrast, the amidation of esters with THF/n-BuLi is applicable to a broader substrate scope. Various aliphatic and aromatic amines can be easily acylated with high to excellent yields. The reaction itself is clean and fast, with a straightforward work-up. The method is further applicable towards the synthesis of phosphoramidates. Nevertheless, the work with aggressive base such as n-BuLi that requires inert atmosphere is a major drawback. In several instances, the low solubility of some substrates in THF limited the scope of the reaction.
Amine substrates that were not applicable under both reaction condition are summarized in Fig. 1. Both reaction settings failed in the amidation of 1a with nitroanilines (75a–b) resulting in a complex mixture of products. No conversion was observed during our attempts to acylate aminophenol (75c), diphenylamines (75k–l) and primary amide (75h). Several amine substrates, containing sulfonamido group (75d–g) as well as some highly sterically hindered amines (75i) and heterocycles (75j, 75m–o) were also not applicable.
Having in hands this large amount of experimental data, we focused our attention towards some practical applications. Driven by the interesting selectivity of the DMSO/t-BuOK system that was found to effectively promote direct acylation of anilines, while not effective for alkyl and benzyl amines we attempted to selectively react 4-amino benzylamine (76) with 3-methyl methylbenzoate (1a) at the aniline nitrogen. Nevertheless, the reaction failed to provide the desired product 77. Surprisingly, when the same reaction was carried out in the presence of n-BuLi we observed remarkable selectivity at the aniline position, leading to the formation of 77 as a sole product in 70% yield (Scheme 6a). The structure of 77 was unambiguously confirmed by X-ray crystallography as hydrochloric acid salt, 77·HCl (See section X-ray and ESI†). This protocol represents a new way for protective group free functionalisation of anilines in a presence of competitive amino groups, which up to now has been achieved by protection/deprotection sequence,30 or through much more laborious synthetic protocols.31–33
We further demonstrated the potential of our findings in a n-BuLi promoted synthesis of N,N-diethyl-meta-toluamide (DEET) (78) (Scheme 6b). DEET is among the most common active ingredients in insect repellents. To date, it has been mostly prepared via classical amidation of the corresponding acylchloride34 and amidation of 3-methyl benzoic acid in presence of coupling agents35 Other methods based on transition metal catalysis36–39 and Grignard promoted amidation of 3-methyl benzonitrile40 could be also found in the literature. Herein, we report its synthesis in nearly quantitative yield (94%) from the readily available methyl 3-methylbenzoate and diethylamine (17) at r.t. for 5 min.
The practical potential of the DMSO/t-BuOK system was demonstrated in the “auto-amidation” of anthranylic ester 1j (Scheme 6c). The resulting compound 79 is a precursor in the synthesis of various natural compounds such as the alkaloids sclerotigenin (80a), circumdatin F (80b), benzomalvin A (80c), asperlicin C (80d) and others (Scheme 6). To date 79 was synthetized using a laborious two-step procedure, which includes reflux in H2SO4/MeOH for 4–5 days.41,42 In contrast, our protocol provided 79 in 5 minutes with 80% yield.
Traditionally the direct acylation of guanidines has been a challenging synthetic task. Therefore, we explored the utility of the DMSO/t-BuOK system for direct acylation of N,N′-diphenylguanidine 23 (Scheme 7a). Although several methods to achieve the synthesis of acylated diphenylguanidines exist, they require laborious synthetic procedures and use of metal catalysts (Pd, Hg, etc.).43–46 None of them is direct and they mostly exploit substituted thioureas as starting materials. To the best of our knowledge, the direct acylation of diphenylguanidine is being reported here for the first time. Furthermore, the reaction itself is very straightforward, fast, catalyst-free and high yielding. The structure of 81 was confirmed by single crystal X-ray diffraction (See section X-ray and ESI†).
The antiplatelet thrombotic agent 83 (Scheme 7b) was originally synthetized by acylation of N,N-dimethyl biguanide with benzoic anhydride in acetone at r.t. for 5 hours, with 71% yield.47 Our conditions provided its direct synthesis in 76% yield from N,N-dimethyl biguanide hydrochloride (82) and methyl benzoate at r.t. for 5 min.
The structure of 81 was found to crystallise in the P21/c space group, again with only a single molecule present in the asymmetric unit cell. Interestingly, despite the potential for hydrogen bonding, only a single intramolecular interaction is present, with all intermolecular contacts consisting of weaker C–H⋯O (C⋯O = 3.314(2) Å) or C–H⋯N (C⋯N = 3.483(2) Å) interactions, or π–π interactions (centroid to centroid distances = 3.88 Å) between the aromatic rings (Fig. 2).
Footnote |
† Electronic supplementary information (ESI) available. CCDC 2127765 and 2127766. For ESI and crystallographic data in CIF or other electronic format see https://doi.org/10.1039/d2ra03524c |
This journal is © The Royal Society of Chemistry 2022 |