A new entry to o,o’ -dihalogenated deoxybenzoins by palladium-catalyzed α -arylation of 2'-chloroacetophenones

The palladium-catalyzed arylation reaction is a valid, advantageous tool for the construction of o,o' -dihalo-1,2-diarylethanones. After an exhaustive range of assays the arylation of 2'- chloroacetophenones is performed in good yields, thus overcoming the handicap imposed by the presence of the ortho -chloro substituent and controlling the regioselective monoarylation with dibromoarenes.


Introduction
The extensive use of 1,2-diarylethanones or deoxybenzoins over the last decades as versatile building blocks of a whole range of compounds 1 has promoted a search for efficient ways for their preparation.Led by the classical Friedel-Crafts arylation of arylacetic acid derivatives, several methodologies have been developed for this purpose, providing a good amount of alternatives available to the synthetic chemist.However, when the nature of the functional groups attached to the aromatic moieties of deoxybenzoins is labile or sensitive to the reaction conditions, or when specific functionalization of ortho and/or ortho' positions is required, most of the above procedures fail. 2 Our group has contributed to the development of new protocols for the generation of 1,2diarylethanones, especially in the field of palladium-catalyzed arylation of ketone enolates, providing simple, highly efficient approaches based on both homogeneous and heterogeneous catalysis that avoid the use of usually harmful transmetallating agents and allow in some cases catalyst recycling and reuse. 3Nevertheless, the latter protocols had never been tested in the synthesis of o,o'-dihalogenated deoxybenzoins, an appealing kind of difunctionalized diarylethanones as they are key starting materials to access complex heterocyclic systems, like the phenanthro [9,10]pyrimidines, isoxazoles and pyrazoles shown below. 4 This paper deals with our efforts towards a reliable, straightforward method to prepare such o,o-dihalo-1,2-diarylethanones employing a palladium-catalyzed arylation of acetophenones.

Results and Discussion
Before starting to evaluate different procedures to accomplish our aim, a reflexive thought about the most convenient substrates was required, as the use of polyhalogenated coupling partners could lead to a lack of regioselectivity. 5Following the initial work of Miura, 6 our experience in the synthesis of 2,3-diarylbenzofurans 7 had revealed that upon certain reaction conditions αarylation with 1,2-dibromoarenes can be stopped at the C-arylation stage to provide α-ortho- bromoaryl moieties (Scheme 1).With regard to the ortho-haloacetophenone coupling partner, the choice of the halide was mainly determined from the comparison of the dissociation energies C-I<C-Br<C-Cl, which are closely related to the oxidative addition, key step of the catalytic cycle proposed as the mechanism to explain arylation of ketone enolates.Thus, iodo derivatives were discarded due to their high reactivity which makes them less selective.In the same context, ortho-bromoacetophenones could also competitively act as arylating agents instead of as sources of enolate. 8ccordingly, 2'-chloroacetophenones 1 were chosen as suitable starting materials along with dibromoarenes 2, taking phenyl methyl ketone 1a and dibromobenzene 2a as model substrates in order to perform a range of assays resumed in Table 1.
Scheme 1 a The amounts of palladium catalyst and ligand are indicated in mol%.b GC-MS yields using tetrabutylammonium bromide as internal standard.c Both starting materials were recovered unreacted.d Complex mixture in which 3a was not detected e L= 2-(di-tertbutylphosphino)diphenyl.f Sealed tube g The dilution employed was twice the one of the previous assay (entry 32).
It was clear from the beginning that slight variations of temperature, relative amounts of palladium sources and ligands or even the metal counterion of the base provoke important variations in the reaction results.Another remarkable feature was that the conditions optimized for the isolation of 2-bromoaryl intermediates depicted in Scheme 1 (Table 1, entries 1-2) 6,7 provided complex mixtures in this case, and it was necessary to increase the temperature and the amount of base to detect the target ethanone 3a (Table 1, entries 3-4), though with relatively low yields (46%) and the occurrence of several side-products which made the purification task difficult.Then, a variety of procedures and variations based on previous work of the group of Hartwig, 9 Buchwald, 10 and Nolan 11 involving ligands or catalytic systems that had proven useful in other α-arylation reactions (P t Bu 3 , 2-(di-tert-butylphosphino)diphenyl, absence of ligands, Pd(PPh 3 ) 4 , BINAP, N-Heterocyclic Carbene [NHC] ligands like IPr and IMes) were investigated, affording in most cases negligible results.Only a combination of Pd 2 dba 3 and BINAP, NaO t Bu and THF (Table 1, entries 13-14) provided a yield similar to that of the initial assay shown in entry 3. It is particularly surprising to notice the lack of ketone 3a when the complex (IPr)Pd(allyl)Cl (entry 20) was employed as this system had been specifically designed for the αarylation of acetophenones.11a In most of the performed assays, the reaction mixture turned red after 0.5 h of reaction.The structures 6-8 depicted below were tentatively proposed as possible complexes that would explain such coloration 12 and also to some extent the hindrance to reach target ethanone 3. Several aspects can seriously prevent the formation of 3, such as a) the formation of palladacycle 6, formed after the expected oxidative addition can attack the enolate and coordinate to the o-chloride, b) the formation of complex 7, generated by an initial coordination to the bidentate enolate, or even c) the formation of organometallic compound 8, the product of an oxidative addition of C-Cl bond instead of C-Br, which could be stable enough to influence the reaction path.Accordingly, in order to avoid such deleterious side-processes, it was decided to increase the amount of ligand or to add a halide salt (NaI, LiCl, Bu 4 NI) in the hope that this would lead to an improved coordination to Pd.Both strategies failed (Table 1, entries 21-23).In view of the moderate results obtained when using Pd 2 dba 3 /BINAP (Table 1, entries 13-14), the effect of little variations in this system could not be discarded.Those variations performed in entries 27-33 featured an optimized procedure (Table 1, entry 30) where finally target ethanone 3a was obtained with good yield, in the presence of traces of monohalogenated deoxybenzoin 4, probably generated by a phenyl migration 13 or a dehalogenation process. 14Attemps to reduce the amount of the catalyst led to the enone 5, generated by an aldol condensation of reagent 1a.This side-reaction could not be avoided even when a higher dilution was used (Table 1, entries 32-33).
The optimized protocol for the preparation of 3a was applied accordingly to a range of commercially or readily available 2-chloroaryl methyl ketones 1 and dibromoarenes 2, as shown in Table 2.
With regard to the acetophenones, the best yields were afforded by 1a, probably due to the relative instability of the trifluoromethylated ketone 1b under these reaction conditions and due to a greater hindrance of the dimethylated ketone 1c to generate the corresponding enolate. 15oreover, and in accordance with previous reports on aryl monohalides and theoretical studies about oxidative addition, 3 electronically neutral and poor dibromoarenes 2a,d couple more efficiently than methylated or alkoxylated derivatives 2b-c.a GC-MS yields using tetrabutylammonium bromide as internal standard are featured in italics.Isolated yields are shown in parentheses.
Due to the presence of structurally similar by-products and starting acetophenones, a decrease to moderate to good yields obtained after purification, considerably lower than the GC-MS yields, has been noticed.However, the advance achieved in this context should be pointed out, particularly taking into account that the only existing general route to o,o'dihalodeoxybenzoins implies a long (three-, four-and even five-step) sequence with overall yields around 20-60% and with obvious disadvantages in terms of atom-economy and sustainability. 4n summary, a new, straightforward synthesis of structurally challenging ortho,ortho'dihalodeoxybenzoins has been carried out by means of a palladium-catalyzed arylation of 2chloroaryl methyl ketones with dibromoarenes.The promising potential of the presented protocol