The OSM Blog

The synthetic scheme currently used to produce series 4 amides is not a very satisfying one:

While the reactions are all workable, both sources of diversity (the "north-west" amide and the "north-east" aryl, here R1 and R2) are introduced before the low-yielding cyclisation step, and the particularly interesting amide is determined by the first reaction in the scheme. As a result, four distinct reactions are needed for every series 4 amide synthesised. Compared to the more efficient scheme used to access the ether series, this is a problem worth addressing.

Series 4 ether scheme. Diversity introduced in final step, from triazolopyrazine prepared in bulk.

A possible option might be protection of the 2-chloro-6-pyrazine carboxylic acid, a substitution/condensation/oxidative cyclisation scheme as above, and then deprotection followed by amide condensation. A more elegant and atom-efficient solution could involve direct carbonylation of an already-cyclised triazolopyrazine system, with addition of an amine occurring subsequently or perhaps in one pot. Also important is that this would give me some interesting chemistry to play with for my thesis.

Alternative routes to series 4 amides. Top: as currently used. Middle: Pd-catalysed carbonylation, either via a carboxylic acid or by direct amide formation. Bottom: oxidation of a benzylic carbon, possibly during oxidative cyclisation. Diversifying amide introduction steps shown in blue.

(Pd-catalysed) direct carbonylation

The more interesting of the two schemes proposed. As with Jo's scheme for the ether series, a chloro-substituted triazolopyrazine would be synthesised, followed by carbonylation with carbon monoxide and a palldium catalyst. I have found two examples of similar reactions in the patent literature (US8546394 claims a reaction but seems not to provide experimenal details, WO2011112766 (A2) describes carbonylating a bromo triazolopyrazine), but no journal articles. Both of these patent examples react the aryl halide with CO under pressure in an alcohol (MeOH or EtOH) to form the corresponding methyl or ethyl ester. A one-step route would instead use an inert solvent and excess R1-NH₂ amine to directly form the amide. Literature examples of this exist for simpler aromatic systems e.g. 10.1021/ol0498287 (1 mol % Pd(BINAP)Cl₂, 3.4 atm CO, 100^o) and 10.1002/anie.2007029 (less favourable, uses DMSO as a solvent). A decent review of Pd-catalysed carbonylation is 10.1021/om800549q.

I plan to try one of these reactions soon. Using the method from 10.1021/ol0498287:

1 eq Ar-Cl, 1.3 eq Et₃N, and 1 mol % Pd(BINAP)Cl₂ in MeOH. Heat at 100 ^oC under 50 psig of CO for ~4 h.

Chloropyrazine would appear to be a good model substrate, and if/when that works I'll try the reaction on chlorotriazolopyrazine.

Metal carbonyls as a CO source

More recent research efforts have replaced gaseous CO with hexacarbonyl transition metal complexes, such as W(CO)₆ and Mo(CO)₆. Other efforts have removed palladium from the reaction entirely, using the Mo or W as both carbonyl-carrier and catalyst (10.1021/jo1002592, 10.1021/jo101611g). One paper (10.1021/om051044p) runs a similar reaction in pure water under microwave superheating, and uses it as a step in the synthesis of a HIV-1 protease inhibitor (initially without palladium, but with the use of a palladacycle when reacting with aryl chlorides). Whether or not this is a synthetically useful reaction remains to be seen, but running a medicinally interesting reaction in neat water is interesting if nothing else.

Oxidation of benzylic carbon

Far less interesting chemically, but wuuld probably solve the problem at hand. Beginning with 2-chloro-5-methylpyrazine, oxidation would lead to the desired carboxylic acid. This might happen during the oxidative cyclisation step, or may require a subsequent reaction. Either way, I see this as more a fall-back option and would prefer to focus on direct carbonylation.