Facile synthetic entry into rotationally restricted 9-arylacridines

Introduction

Biaryl, biheteroaryl and aryl-substituted heterocyclic compounds with the central sp²-sp² or sp²-sp³ bond can be severely restricted in rotational freedom when substituted in the vicinity of the central bond joining the two subsystems of the molecule [1]. There is no general synthetic approach to the synthesis of such compounds, as exemplified by some selected recent reports [2–9]. In this communication, we describe a new method to synthesize 9-arylacridines with a severe steric hindrance between the acridine system and the aryl substituent. This work is a continuation of the chemistry of the anionically activated trifluoromethyl group in 2-(trifluoromethyl)aniline (1, Scheme 1) that has previously resulted in the development of numerous new methodologies for the synthesis of many classes of heterocyclic compounds [10–13]. Briefly, the trifluoromethyl group is stable in the absence of an anionic center in the molecule but undergoes a facile elimination of fluoride when the anionic center is introduced at the γ position. Lithium reagents, for the most part, have been used previously to activate the trifluoromethyl group.

Scheme 1

Treatment of substrate 1 with aryllithiums, generated at -70°C by the reaction of 1-bromo-2,6-dimethylbenzene or 1-bromo-2-ethylbenzene with n-butyllithium, gave a complicated mixture of products, none of them major. The outcome was similar for reactions conducted in ether or tetrahydrofuran (THF), at low temperature or under reflux conditions. These results can be explained in terms of secondary lithiation reactions resulting in generation of benzyllithium reagents, in addition to the expected phenyllithium derivatives.

Different types of products, depending on temperature, were obtained by treatment of 1 with Grignard reagents generated from the two bromobenzenes mentioned above. At low temperature in the range of -70°C to 23°C, each reaction produced an acyclic trimeric product derived from three units of 1 and another acyclic trimeric product containing two fragments derived from 1 and a fragment of the aryllithium reagent (not shown). By contrast, conducting the reaction in THF under reflux conditions resulted in formation of a single major product in each case (Schemes 1 and 2). Products 8 and 10 were obtained in the respective yields of 35% and 66% following purification, and their structures are fully consistent with the analytical data including ¹H NMR, ¹³C NMR and high resolution mass spectra (HRMS). The oily compound 8 was additionally transformed into a solid hydrobromide salt that was characterized by elemental analysis. It is important to note that compounds 8 and 10 are devoid of an alkyl group of the aryl Grignard reagent. The mechanism, that takes this feature into account, is suggested in Scheme 1. In full agreement with the previously published reactivity of substrate 1 in the presence of lithium reagents, the first step involves ionization of 1 to generate a magnesium derivative 2. Elimination of MgBrF generates the aza-ortho-xylene 3 that is recognized as the major intermediate in the anionically activated chemistry of 1. A subsequent nucleophilic addition of the Grignard reagent with 3 is suggested to reinstate aromaticity by formation of 4. A similar sequence of reactions through the intermediary of 5 leads to the key diaryl-substituted non-aromatic intermediate 6 the electrocyclization of which or its magnesium derivative results in the formation of an aromatic ring. Compound 7 is the suggested direct precursor to the final product 8 by the formal elimination of methylmagnesium bromide. It is important to stress ‘the formal elimination’ because this transformation may involve a more complicated pattern that is consistent with the simple elimination of MeMgBr. In a similar way, the intermediate product 9 is the suggested precursor to product 10 (Scheme 2).

Scheme 2

Grignard reagents are mostly aggregated in solution, and the aggregation diminishes their reactivity [14]. An increased concentration of the more nucleophilic monomeric molecules can be achieved by increasing temperature. This feature is fully consistent with the suggested intermediary of 6 at an elevated temperature in comparison to the reaction conducted at low temperature. More specifically, the different outcomes of the reactions of 1 conducted at 23°C and under reflux conditions can be explained in terms of an increased nucleophilicity of the Grignard reagents at the elevated temperature. The preferred addition of 2 with aza-ortho-xylene 3 and analogs takes place at low temperature to produce the acyclic products mentioned above. This is due to high concentration of 2 and the presence of Grignard reagent in an aggregated form. By contrast, the more nucleophilic monomeric Grignard molecules successfully compete with 2 in the addition reaction with 3 at elevated temperature.

It should be noted that the two compounds described in this preliminary report are examples taken from a large library of compounds synthesized by using the general procedure. A full account on this chemistry will be published in due course.

Experimental