9,10-Dioxa-1,2-diaza-anthracene derivatives from tetrafluoropyridazine

Reaction of tetrafluoropyridazine with catechol gives a tricyclic 9,10-dioxa-1,2-diaza-anthracene system by a sequential nucleophilic aromatic substitution ring annelation process, further extending the use of perfluoroheteroaromatic derivatives for the synthesis of unusual polyfunctional heterocyclic architectures. The tricyclic scaffold reacts with amines and sodium ethoxide providing a short series of functional 9,10-dioxa-1,2-diaza-anthracene systems.


Introduction
Drug discovery programmes are continually searching for viable synthetic routes to highly novel classes of heterocyclic compounds with the aim of exploring chemical 'drug-like' space [1] and uncovering valuable biological activity for hit-tolead generation of new chemical entities by parallel synthesis techniques. The wide variety of relatively simple heterocyclic structural types that have not been synthesised [2], the rela-tively low level of structural diversity in all known organic structures [3] and, indeed, the perceived lack of structural diversity in pharmaceutical companies' compound collections have often been suggested to be among the bottlenecks in drug discovery programmes [4]. Methodology for the ready synthesis of new organic frameworks is still required and, in this context, heterocyclic scaffolds based on novel molecular archi-Scheme 1: Synthesis of novel tricyclic heterocycles from pentafluoropyridine.
tecture that bear multiple functionality and can be rapidly processed into many analogues by parallel synthesis are particularly valuable [5,6].
In a continuing research programme, we have demonstrated that perfluorinated heteroaromatic derivatives are very useful starting scaffolds for the synthesis of a variety of heteroaromatic [7], [5,6] and [6,6]-bicyclic [8][9][10][11], and polycyclic heterocyclic systems [12]. Perfluoroheteroaromatic derivatives are either commercially available or can be accessed by halogenexchange processes by reaction of the corresponding perchloroheteroaromatic system and potassium fluoride [13]. No special techniques for handling perfluoroheteroaromatic compounds are required, apart from the usual laboratory precautions, because these systems are generally volatile, colourless liquids. We established that highly novel tricyclic scaffolds, such as the relatively uncommon dipyrido[1,2-a:3′,4′-d]imidazole system 1, could be synthesised from pentafluoropyridine in a single step [12], exemplifying our general strategy for the synthesis of highly novel classes of polyfunctional heterocyclic compounds. Several dipyrido[1,2-a:3′,4′-d]imidazole analogues 2 were prepared by the displacement of the remaining ring fluorine atoms by nucleophilic aromatic substitution processes (Scheme 1).
We were interested in further expanding the use of highly fluorinated heterocycles for the preparation of novel heterocyclic structures and focussed upon the synthesis of ring fused systems that could be derived from the reaction of tetrafluoropyridazine (3) with catechol (4). In principle, two possible systems 5 and 6 may be formed depending upon the regioselectivity of the nucleophilic aromatic substitution processes (Scheme 2).
Both 5 and 6 have ring fluorine atoms present that may, in principle, be displaced by nucleophiles which could lead to the synthesis of many analogues of these systems. The dioxa-1,2-diazaanthracene (or 3,4-difluorobenzo [5,6] [1,4]dioxino [2,3c]pyridazine also referred to as benzodioxinopyridazine) systems are very rare heterocyclic structures and only a handful of analogues based upon this molecular skeleton have been synthesised, mainly by the reaction of chlorinated pyridazines with catechol [14][15][16].
In this paper, we describe the synthesis of dioxa-1,2-diazaanthracene derivatives by the sequential reaction of commercially available tetrafluoropyridazine with catechol, and a short series of nucleophiles.

Results and Discussion
Initially, we carried out reactions of tetrafluoropyridazine (3) with one and two equivalents of sodium phenoxide as a model substrate for catechol (Scheme 3).
Reaction of one equivalent of sodium phenoxide with (3) gave product 7 arising from substitution of fluorine located at the site para to activating ring nitrogen, consistent with earlier studies involving reactions between tetrafluoropyridazine and various nucleophiles [13]. Similarly, reaction of two equivalents of It seems reasonable to assume that initial substitution occurs at the 4-position of 3, analogous to the reaction between 3 and phenoxide, to give intermediate 5a. At this point, we would expect cyclisation to occur at position 5 to give product 6, again by analogy to the outcome of reaction between 3 and excess phenoxide. However, since nucleophilic aromatic substitution reactions are frequently reversible [13], conversion of 6 must occur via intermediate 5a and lead to the most thermodynamically stable product 5 (Scheme 5).
The utility of the dioxa-1,2-diaza-anthracene system 5 as a scaffold for array synthesis was assessed in representative reactions with a short series of nucleophiles (Scheme 6).
Nucleophilic substitution of fluorine at the 4-position occurs regiospecifically to afford products 9a-c according to 19 F NMR analysis of the corresponding reaction mixtures. The 19 F NMR resonances located at ca. −90 ppm are characteristic of fluorine atoms located at sites ortho to a ring nitrogen atom. X-ray crystallography of the allylamino derivative 9b (Figure 1), and a comparison of NMR spectral data, confirms the structures of these analogues.   Again, the regiospecificity of these reaction processes occurs because of the activating effect of ring nitrogen directly opposite the site of nucleophilic substitution.

Conclusions
A small range of dioxa-1,2-diaza-anthracene analogues 5 and 9 have been synthesised from tetrafluoropyridazine in two efficient steps, further expanding the application of highly fluorinated heterocycles for the synthesis of rare heterocyclic architectures.