Guest Editor: S. Matsunaga
Beilstein J. Org. Chem. 2018, 14, 1546–1553. doi:10.3762/bjoc.14.131
Received 14 May 2018, Accepted 15 Jun 2018, Published 25 Jun 2018
Versatile C–H amidations of synthetically useful ferrocenes were accomplished by weakly-coordinating thiocarbonyl-assisted cobalt catalysis. Thus, carboxylates enabled ferrocene C–H nitrogenations with dioxazolones, featuring ample substrate scope and robust functional group tolerance. Mechanistic studies provided strong support for a facile organometallic C–H activation manifold.
Keywords: amidation; C–H activation; cobalt; ferrocene; mechanochemistry
C–H activation has surfaced as a transformative tool in molecular sciences [1-9]. While major advances have been accomplished with precious 4d transition metals, recent focus has shifted towards more sustainable base metals [10-17], with considerable progress by earth-abundant cobalt catalysts [18-22]. In this context, well-defined cyclopentadienyl-derived cobalt(III) complexes have proven instrumental for enabling a wealth of C–H transformations [23-41], prominently featuring transformative C–H nitrogenations [42,43] in an atom- and step-economical fashion [44-59]. Within our program on cobalt-catalyzed C–H activation [60-68], we have now devised C–H nitrogenations assisted by weakly-coordinating  thiocarbonyls [70,71], allowing the direct C–H activation on substituted ferrocenes [72-93] – key structural motifs of powerful transition metal catalyst ligands and organocatalysts (Figure 1) [94-97]. During the preparation of this article, the use of strongly-coordinating, difficult to remove directing groups has been reported [70,71]. In sharp contrast, notable features of our approach include (i) cobalt-catalyzed C–H amidations of thiocarbonylferrocenes by weak coordination, (ii) thermal and mechanochemical [98-100] cobalt-catalyzed ferrocene C–H nitrogenations, (iii) versatile access to synthetically useful aminoketones, and (iv) key mechanistic insights on facile C–H cobaltation.
We initiated our studies by probing various reaction conditions for the envisioned C–H amidation of ferrocene 1a (Table 1). Among a variety of ligands, N-heterocyclic carbenes and phosphines provided unsatisfactory results (Table 1, entries 1–3), while the product 3aa was formed when using amino acid derivatives, albeit as of yet in a racemic fashion (Table 1, entries 4–7). Yet, optimal catalytic performance was realized with 1-AdCO2H (Table 1, entries 8 and 9) [101-104], particularly when using DCE as the solvent (Table 1, entries 9–12). A control experiment verified the essential nature of the cobalt catalyst (Table 1, entry 13). In contrast to the thiocarbonyl-assisted C–H amidation, the corresponding ketone failed thus far to deliver the desired product, under otherwise identical reaction conditions.
Table 1: Thiocarbonyl-assisted C−H nitrogenation of ferrocene 1a.a
aReaction conditions: 1a (0.13 mmol), 2a (0.15 mmol), ligand (30 mol %), [Co] (5.0 mol %), solvent (1.0 mL). bReaction performed in the absence of [Cp*Co(CH3CN)3][SbF6]2. Yields of isolated product.
With the optimized reaction conditions in hand, we explored the robustness of the cobalt-catalyzed ferrocene C–H amidation with a variety of 1,4,2-dioxazol-5-ones 2 (Scheme 1). Hence, the chemoselectivity of the cobalt catalyst was reflected by fully tolerating sensitive electrophilic functional groups, including amido, chloro, bromo and nitro substituents in the para-, meta- and even the more congested ortho-position.
The versatile cobalt-catalyzed C–H amidation was not limited to mono-substituted ferrocenes 1 (Scheme 2). Indeed, the arylated ferrocenes 1b–d were identified as viable substrates likewise.
Moreover, differently substituted thiocarbonyls 1 were found to be amenable within the cobalt-catalyzed C–H amidation manifold by weak-coordination (Scheme 3).
Given the versatility of the cobalt-catalyzed C–H nitrogenation, we became intrigued to delineating its mode of action. To this end, C–H amidations in the presence of isotopically labelled co-solvents led to a significant H/D scrambling in proximity to the thiocarbonyl group. These findings are indicative of a reversible, thus facile organometallic C–H cobaltation regime (Scheme 4).
Next, intermolecular competition experiments revealed that electron-rich arylated thiocarbonylferrocene 1 reacted preferentially, which can be rationalized with a base-assisted internal electrophilic substitution (BIES) [24,105] C–H cobaltation mechanism. In addition, the electron-rich amidating reagent 2c was found to be inherently more reactive (Scheme 5).
As to further late-stage manipulation of the thus-obtained products, the amidated thiocarbonylferrocene 3aa could be easily transformed into the corresponding synthetically useful aminoketone 4aa (Scheme 6), illustrating the unique synthetic utility of our strategy.
Mechanochemical molecular synthesis has attracted recent renewed attention as an attractive alternative for facilitating sustainable organic syntheses . Thus, we were delighted to observe that the mechanochemical C–H nitrogenations proved likewise viable by thiocarbonyl assistance in an effective manner (Scheme 7).
In conclusion, we have reported on the unprecedented cobalt-catalyzed C–H nitrogenation of ferrocenes by weakly-coordinating thiocarbonyls. The carboxylate-assisted cobalt catalysis was characterized by high functional group tolerance and ample substrate scope. Mechanistic studies provided evidence for a facile C–H activation. The C–H amidation was achieved in a thermal fashion as well as by means of mechanochemistry, providing access to synthetically meaningful aminoketones.
|Supporting Information File 1: Experimental procedures, characterization data, and NMR spectra for new compounds.|
|Format: PDF||Size: 2.8 MB||Download|