Reports: ND154452-ND1: Metal-Free Functionalization of Aromatic Rings - Remarkably Mild Carbon--Heteroatom and Carbon-Carbon Bond Formations, and Dehydrogenative Cross-Coupling Reactions

William Chain, PhD, University of Delaware

We are exploring transformations that leverage the nitrogen functionality of aniline to enable selective reactivity on the aromatic ring.  Transient oxidation of the amino group affords us the ability to introduce carbon–heteroatom and carbon–carbon bonds under exceptionally mild reaction conditions.  The oxidation event exploits the typical liability experienced with anilines and enables controllable, metal-free, and environmentally friendly reactions.  Subsequent O-acylation temporarily transforms the aniline ring into an electron deficient species allowing selective and controlled introduction of new carbon–heteroatom and carbon–carbon bonds on the aromatic ring, while retaining the classical regiochemistry associated with electrophilic aromatic substitution.

The original goals of this proposal were to harness the elevated reactivity of N,N-dialkylaniline-N-oxides to: (1) complete the development of new carbon–heteroatom bond formations on aromatic rings, (2) to develop direct carbon–carbon bond formations on aromatic rings, (3) to develop metal-free dehydrogenative cross-coupling reactions of N,N-dialkylanilines, and (4) to develop the direct trifluoromethylation of N,N-dialkylanilines.

In the first year of support, we have made considerable progress in the manipulation of the aniline-N-oxide platform.  We recently described success in goals 1 and 2 with protocols for the C-alkylation of N,N-dimethylaniline-N-oxides, as well for the synthesis of amino alcohols, amino triflates and tosylates (Figure 1).

Figure 1.  Recently described bond formations on the aniline N-oxide platform.

In each protocol, an acylation event at oxygen that gives a p–system is believed to facilitate a [3,3]-sigmatropic rearrangement to excise the weak nitrogen-oxygen bond and form the new carbon–heteroatom or carbon–carbon bond of interest.  Preliminary computational studies support this conclusion.    

We demonstrated the direct formation of carbon–carbon bonds utilizing ethyl malonyl chloride as an acylating agent that could present a carbon–carbon p–system for the subsequent [3,3]-sigmatropic rearrangement.  In principle, any acylating agent with an electron-withdrawing substituent that might favor or facilitate formation of the enol tautomer of the acylated N-oxide should undergo the carbon–carbon bond forming event.  A single literature reference reported ketene as a viable participant in reactions with aniline-N-oxides, albeit with several side products attributed to multiple mechanistic pathways that might excise the weak nitrogen-oxygen bond.  We revisited diketene as a carbon source as well as cyanoacetyl chloride, both of which should undergo reactions analogous to ethyl malonyl chloride.  We described successful carbon–carbon bond formations with both of these substrates, and did not encounter any of the difficulties reported in prior studies with ketene.  Following acylation and decarboxylation, we isolated the corresponding methyl ketones and benzyl nitriles cleanly (Figure 2).

Figure 2.  New carbon–carbon bond formations utilizing aniline-N-oxides.

We proposed the formation of new carbon-heteroatom bonds again utilizing appropriate p–systems in rearrangement reactions.  We experienced successful carbon–nitrogen bond formation utilizing phenylisocyanate as a nitrogen source.  Following rearrangement, a decarboxylation event analogous to the reactions described above yields the secondary amine directly (Figure 3).  Further studies utilizing nitriles and diimides as nitrogen atom sources to give primary and secondary amines are ongoing.  

Figure 3.  Carbon–nitrogen bond formations utilizing aniline-N-oxides.

One major liability that we have encountered at the aniline-N-oxide oxidation level is the simple Polonovski-type elimination reaction that may occur prior to group transfer, which typically produces a small amount of dealkylated aniline products.  However, during our attempts to generate polyfunctionalized anilines by successive oxidation cycles, we found that substrates that contain a nucleophilic function ortho to the aniline can undergo a Mannich-type reaction in which the previously unproductive iminium ion is captured to give substituted indolines products.  We took advantage of this observation and developed a protocol for the convenient synthesis of a variety of indolines with opportunities for further elaboration.  The reaction appears to be general and tolerant of both electron-withdrawing and electron-donating groups on the aromatic ring (Figure 4).  We found both the benzyl esters and nitriles to be competent participants in the Polonovski-Mannich cyclization reaction.

Figure 4.  Indoline formation by a Polonovski-Mannich cyclization.

We have also made progress with goal 3, the development of metal-free dehydrogenative cross-coupling reactions with aniline-N-oxides.  We are studying this goal by two strategies – the [3,3]-sigmatropic rearrangement to relay appropriate coupling partners from nitrogen to carbon, and the introduction of exogenous carbon nucleophiles to expel a nitrogen-bound leaving group.  We have experienced success by the in situ formation and capture of benzyne, followed by a group transfer reaction (Figure 5).  Extension of this method to more substituted coupling partners is ongoing, as well as the introduction of exogenous carbon nucleophiles.

Figure 5.  Biaryl formation by the coupling of aniline-N-oxide and benzyne.