Open Access Article
Jiang
Nan
*a,
Xin
Ren
a,
Qiang
Yan
a,
Shilei
Liu
a,
Jing
Wang
a,
Yangmin
Ma
a and
Michal
Szostak
*b
aThe Youth Innovation Team of Shaanxi Universities, Shaanxi Key Laboratory of Chemical Additives for Industry, College of Chemistry and Chemical Engineering, Shaanxi University of Science and Technology, Xi'an 710021, China. E-mail: nanjiang@sust.edu.cn
bDepartment of Chemistry, Rutgers University, 73 Warren Street, Newark, New Jersey 07102, USA
First published on 27th February 2023
Direct functionalization of the C(O)–N amide bond is one of the most high-profile research directions in the last few decades; however oxidative couplings involving amide bonds and functionalization of thioamide C(S)–N analogues remain an unsolved challenge. Herein, a novel hypervalent iodine-induced twofold oxidative coupling of amines with amides and thioamides has been established. The protocol accomplishes divergent C(O)–N and C(S)–N disconnection by the previously unknown Ar–O and Ar–S oxidative coupling and highly chemoselectively assembles the versatile yet synthetically challenging oxazoles and thiazoles. Employing amides instead of thioamides affords an alternative bond cleavage pattern, which is a result of the higher
conjugation in thioamides. Mechanistic investigations indicate ureas and thioureas generated in the first oxidation as pivotal intermediates to realize the oxidative coupling. These findings open up new avenues for exploring oxidative amide and thioamide bond chemistry in various synthetic contexts.
conjugation and high resonance stabilization (Scheme 1a, left).1a–h Notwithstanding that the amide bond is well-recognized as the most stable carboxylic acid derivative, the highly selective C(O)–N activation of amides that enables us to sequentially perform chemical operations has attracted great attention in organic synthesis over the past few decades (Scheme 1a, right).2a–g This is mostly due to the ubiquitous presence of amide bonds in biochemistry, pharmaceuticals, and diverse bulk and fine chemicals as well as their versatility in synthetic settings.3a–f Indeed, amide bond activation has been experiencing major progress by means of carefully controlled steric repulsion, conformational restriction and electronic effects to successfully activate the N–C(O) bond by ground-state-destabilization of amidic resonance (Scheme 1b, left).4,5a–d,6,7 This mode of activation broadly relies upon disrupting
conjugation to increase negative charge at the nitrogen atom, while this mode should be contrasted with the tremendously important field of electrophilic activation of amides.4 The current well-developed reaction types are predominantly restricted to sterically twisted amides8 or electronically activated amides9 featuring electron-withdrawing groups (EWGs) that facilitate acyl addition or oxidative addition of the C(O)–N bond to a low-valent metal. Comparatively, formamides are much more broadly available than such activated amides; however, there are few examples of the site-specific C(O)–N activation of non-activated formamides (Scheme 1b, right).10
Formamides characterized by planarity and featuring electron-donating groups (EDGs) are not amenable to the ground-state-destabilization pathway. The present amide bond activation has been most successful using metal-catalyzed coupling and only implemented to generate architecturally linear molecules by single bond formation thus far.11
Thioamides, the closest O to S isosteres of amides in the strictest sense, have served as highly valuable motifs to accomplish synthetic processes.12 With respect to the structural and electronic characteristics, thioamides feature a shorter C
S bond than the analogous C
O bond in amides. The weaker electronegativity of sulfur and the higher
conjugation provide the higher contribution of the polar resonance form. This results in a stronger C(S)–N bond in thioamides, which together with potential side-reactions, such as desulfurization, electrophilic addition, and cyclization, poses a major challenge to chemoselectively cleave the C(S)–N bond (Scheme 1c).13 Indeed, studies on the activation of twisted thioamide bonds are exceedingly scarce, and chemoselective C(S)–N bond activation of non-twisted and non-electronically activated thioamides has not been documented to date.
In continuation of our long-standing interest to discover new modes of activation of amide bonds,7 herein, we disclose our findings on novel hypervalent iodine-induced twofold oxidative coupling of amines with amides and thioamides (Scheme 1d and e). Of note, the protocol accomplishes divergent C(O)–N and C(S)–N disconnection by the previously unknown Ar–O and Ar–S oxidative coupling. The method represents the first example of combining the inert C(O)–N/C(S)–N coupling with C–H functionalization in a single operation, providing a powerful platform to assemble the versatile yet synthetically challenging oxazoles and thiazoles that are common in pharmaceuticals and agrochemicals14a,b but currently lack a practical synthetic route.15 Employing amides instead of thioamides affords an alternative bond cleavage pattern, which is a result of the higher
conjugation in thioamides, furnishing C2–NR2-substituted benzoxazoles (Scheme 1f).16 This oxidative pathway outperforms the current dominant tactic of decorating the benzoxazole scaffold (Scheme 1g).17 Overall, these findings open up new avenues for exploring oxidative amide and thioamide bond chemistry in various synthetic contexts.
resonance.
| Entry | Variations from the standard conditions | Yield of 3ab (%) |
|---|---|---|
| a Performed on a 0.20 mmol scale. b Isolated yields. PIDA = phenyliodine(III) diacetate, PIFA = phenyliodine(III) bis(trifluoroacetate), and PhIO = iodosylbenzene. | ||
| 1 | None | 70 |
| 2 | Without IBX | 0 |
| 3 | PIDA instead of IBX | 23 |
| 4 | PIFA instead of IBX | 12 |
| 5 | PhIO instead of IBX | 10 |
| 6 | At 110 °C | 34 |
| 7 | O2 atmosphere | 40 |
| 8 | Reducing loading of IBX to 1.0 equiv. | 52 |
| 9 | Reducing loading of 2 to 2.0 equiv. | 45 |
| 10 | In DCE (0.25 M) | 17 |
| 11 | In 1,4-dioxane (0.25 M) | nd |
| 12 | In DMSO (0.25 M) | 26 |
Having identified optimal reaction parameters, we next sought to evaluate the substrate scope of this novel transformation by coupling various naphthylamines with N,N-dimethylthioformamide 2 (Table 2). As depicted in Table 2, a range of 2-naphthylamines reacted smoothly with full regioselectivity at the α-site of amine substrates, delivering thiazoles 3a–s in good yields. The electronically distinct substituents at 6- and 7-positions, such as methyl (3b), methoxy (3c and 3h), methoxycarbonyl (3d), cyclopropyl (3e), bromo (3f and 3g), furyl (3i), thienyl (3j), thianthrenyl (3k), and a series of diverse-substituted arene system (3l–3p) moieties, all proved to be competent substrates. The structure of 3b was unambiguously assigned by X-ray single crystallography. Notably, the high-value but usually challenging in oxidative coupling alkenyl (3q) group was retained intact under these conditions, thus providing a versatile handle for downstream modification. Likewise, π-electron-rich phenanthryl- and anthryl-derived thiazoles (3r–3s) were equally compatible and prepared in 70% and 78% yields, respectively. Furthermore, the use of 1-naphthylamine to engage in this protocol also furnished the target molecule 3aa in 56% yield. The naphthyl ring could be diversely decorated with cyclopropyl (3ab) or aromatics substituted with alkyl (3ad), sulfuryl (3ae), or nitro (3af) groups. It should be noted that 4-bromo-1-naphthylamine was not tolerated, but delivered the debrominated product 3aa. However, the readily oxidized dihydro-acenaphthylene unit was well-compatible and generated product 3ah in 56% yield. At the present stage, the reaction is not compatible with simple anilines (3ai) probably because of the more stable aromaticity of than that of naphthylamines.
We next employed N,N-dimethylformamide (DMF) as the structurally closest analog of N,N-dimethylthioformamide to participate in this novel methodology (Table 2). To our surprise, reacting 2-naphthylamine 1a with DMF resulted in the formation of NMe2-substituted naphthyl-fused oxazole 4a, wherein the alternative bond cleavage pattern occurred chemoselectively. Considering the unique selectivity with DMF to accomplish the synthesis of 2-amino-substituted benzoxazoles by oxidative coupling, we optimized the reaction conditions (details in the ESI†). By slightly tuning the conditions by replacing IBX with PIDA, the target 4a was isolated in 81% yield, and its structure was unambiguously confirmed by X-ray crystallography.
With the optimized conditions in hand, the scope of this benzoxazole formation was investigated (Table 2). As shown, various naphthylamine counterparts were amenable to this divergent reaction. Specifically, a collection of functionalized naphthylamines, encompassing methyl, methoxy, ester, cyclopropyl, and bromo groups, were subjected to this oxidative process to afford the expected products 4b–4h in 59–81% yields. Importantly, in this variant by-products resulting from C–N dissociation were not observed, indicating the key difference between
conjugation resonance under oxidative conditions (vide infra). Strikingly, the sensitive vinyl fragment (4i–4j) was well-compatible. An array of electronically distinct aryl systems (4k–4p) and even heterocycles (4q–4r) were also feasible while the simple amine behaved unsatisfactorily (4s). In addition, the substrate scope of amides was subsequently assessed. The results showed that linear (4a′–4d′), sterically bulky (4e′), and cyclic (4f′–4h′) tertiary amides all underwent smooth cyclization. Notably, this reaction could also be extended to secondary amides, thus resulting in the corresponding NH-secondary amine-tethered heterocycles (4i′–4k′) with high value in organic synthesis.
Next, we evaluated further functionalization of several pharmaceutical synthons by these two protocols (Scheme 2). Thus, DL-menthol and (−)-borneol derivatives could be easily converted into the corresponding thiazoles 5a and 5b in high yields. Likewise, the industrially important diacetone-β-fructose (DAF) and (−)-borneol synthons also performed well in this transformation to synthesize value-added oxazoles 5c and 5d in 77% and 75% yields.
To showcase the versatility and synthetical value of this new methodology, additional experiments were next conducted. First, scaling-up these two reactions to 5.0 mmol proved to be readily feasible, and both thiazole 3a and oxazole 4a were obtained in high yields, attesting to the scalability of the protocols (Scheme 3a). Second, shortening the reaction time was feasible to deliver oxazole 4a in 53% yield after one minute reaction time, which demonstrated exceedingly fast oxidative coupling (Scheme 3b). Next, although the direct synthesis of C2-NR2 substituted thiazoles was not feasible by this method, the product could be readily obtained through functionalizing the C–H bond of 3a with DMF as the aminating reagent (Scheme 3c).19a Furthermore, using copper catalysis,19b C–H arylation of 3a with iodobenzene was feasible to generate C2-arylated 7 in 64% yield (Scheme 3d). Likewise, despite the presence of several coordinating heteroatoms, Suzuki–Miyaura cross-coupling of 4f proceeded smoothly to deliver 8 in 89% yield (Scheme 3e). Next, the secondary naphthylamine 1′ could be cyclized to generate oxazole product 4a in 53% yield via C–N bond cleavage, ascribed to the in situ formed free naphthylamine (Scheme 3f). Finally, when formamide was used, the oxazole 9 was not formed, but the reaction produced 78% yield of formylation product 10 (Scheme 3g). This outcome proved the stability of 10 in the presence of hypervalent iodine and provided insights into the reaction mechanism.
To shed light on this highly chemoselective reaction pathway, several sets of control experiments as well as studies with key intermediates were performed (Scheme 4). As depicted in Scheme 4a and b, C2-free oxazole 11 failed to deliver product 4a, and conversely, C2–NMe2 thiazole 6 did not undergo the C–N bond cleavage to afford product 12. These two results clearly illustrate the divergence of oxidative mechanisms to build oxazole and thiazole rings. To further probe the key reaction intermediates, several possible derivatives were prepared and subjected to the reaction conditions. As such, amidine 13 with high-activity20 did not afford the cyclization product, while urea 12 was transformed into the target molecule 4a with a comparatively high efficiency cf. the reaction between 1a and DMF (Scheme 4c). On the other hand, employing amine-derived thioformamide 14 resulted in full recovery, which ruled out the possible intermediacy of thioformamide. However, the use of thiourea 15 resulted in the formation of the target product 3a with efficiency comparable to the reaction of 1a and N,N-dimethylthioamide (Scheme 4d). Finally, a deuterium labelling study with d7-N,N-dimethylthioamide showed in 22% D incorporation in the C2 position (Scheme 4e).
Overall, these mechanistic studies indicate that urea and thiourea species act as pivotal intermediates to enable the oxidative assembly of oxazoles and thiazoles. The key mechanistic difference results from the divergent
conjugation, where the stronger conjugation of the thioamide bond induces N–C cleavage under oxidative conditions. A tentative mechanism is proposed in Scheme 5. The reaction commences with electrophilic activation of the C
S and C
O bonds by the hypervalent iodine reagent,21 which is followed by the nucleophilic addition of amine. Subsequently, the first oxidation takes place to give thiourea 15 or urea 12. At this point, with respect to the generation of thiazoles, the electron-rich intermediate 15 reacts with the hypervalent iodine(VI) to induce a hydride shift to afford A, which undergoes the second oxidative coupling by C–H/C–S cyclization. For the generation of oxazoles, the product is obtained by electrophilic activation, the second oxidative coupling by C–H/C–O cyclization and isomerization in the presence of hypervalent iodine(III). In the case of thioamides, hypervalent iodine promotes amine dehydrogenation to form iminium.22a The key difference in reactivity between amides and thioamides stems from higher reactivity of thioamides to electrophiles22b–d and high affinity of sulfur towards hypervalent iodine reagents23a with a weaker thiocarbonyl bond that is more susceptible to reduction.23b
Footnote |
| † Electronic supplementary information (ESI) available. CCDC 2223988 and 2224291. For ESI and crystallographic data in CIF or other electronic format see DOI: https://doi.org/10.1039/d3sc00301a |
| This journal is © The Royal Society of Chemistry 2023 |