Category Archives: Process research

Publication of the week, number 93, 4th September 2015

This weeks sees a new, chiral, route into a best selling drug developed by an Irish group from the Royal College of Surgeons in Ireland and Kelada Pharmachem. Ltd.

(S)-Pregabalin was (is) manufactured and sold by Pfizer and being now off patent, also by several generic manufacturers.

Screen Shot 2015-09-03 at 08.23.43

Various routes to it have been developed in particular a Rh(DUPHOS) enantioselective hydrogenation, which has the disadvantage of cost! Enzymatic methods are also used as are various other chiral routes. Even a racemic process has been published employing a resolution step loosing 50% of the material. Our Irish friends have developed a chiral PTC route involving an organocatalytic conjugate addition of nitromethane to a 1,4-acceptor-

Screen Shot 2015-09-03 at 08.44.33

The nitroisoxazole is a masked carboxylic acid which is liberated by base hydrolysis. This compound is made in 4 steps from acetylacetone and involves addition to hydroxylamine, a nitration and an aldol reaction with isovaleraldehyde followed by formation of the mesylate and elimination. The overall yield is 88%. There are a few interesting comments in the paper to this sequence: “The condensation of acetylacetone and hydroxylamine is a well-known reaction that has been described in many papers.” Looking at the reference they cite, it is only one to a paper in Chemical Communications. “This reaction takes place in water and involves dissolving an opportune hydroxylamine salt and a base to make an aqueous solution of hydroxylamine (pH ≈ 8) to which acetylacetone was slowly added under vigorous stirring. The addition of acetylacetone to a solution of hydroxylamine is exothermic, and on a 10 kg scale the temperature of the solution went from 18 °C to 35−40 °C (5 added in portions during a period of 1 h)”. So here a thermal challenge for anyone. I suppose the heat capacity of water keeps the exotherm under control. I wonder what happens when the stirrer breaks?

The nitration “This involved dissolving compound 6 in cold (0−5 °C) sulfuric acid then adding nitric and then excess of sulfuric acid keeping the temperature below 15−20 °C. The subsequent heating of the reaction mixture generated an exotherm which brought the temperature from 50 to 58 °C. This data is in line with observation recorded for other aromatic nitrations which is explained considering the generation of a +NO2 ion. Quenching of reaction using water ice bath gave almost quantitative yield of desired 7, which is highly water insoluble and precipitated immediately”. So another interesting exotherm here.

The aldol reaction is also not trivial as “After several attempts, a ternary mixture of water, tetrahydrofuran, and methanol was identified as the best media to ensure high yields of aldol compound“. The elimination of the mesylate is also exothermic.

Various quinidinium  salts were screened for the nitromethane addition, the best was

Screen Shot 2015-09-03 at 09.18.10

which gave the addition product

Screen Shot 2015-09-03 at 09.23.48

in 96% yield and 88% ee. The ee% of this compound is concentration dependant. Running the nitromethane addition higher than 0.08M causes a decrease in %ee. Crystallisation brought the ee up to >99%. Attachment of the catalyst to polystyrene allowed re-use (up to 10 times) without loss of enantioselectivity.

There are a few other tit-bits of information in the paper, the nitro acid (from hydrolysis of the nitroisoxazole) “is remarkably acid sensitive, and it is prone to decompose rapidly if heated or if brought to a pH < 3. On the contrary, sodium carboxylate salt of 2 is very stable and could be isolated and stored for long periods“. And “With the low grade Ni/Ra available to us and in order to operate at 1 atm, 2 equiv of Ni/Ra were necessary to obtain quantitative yields of compound 1. The high loading of Ni/Ra was required to avoid the formation of products of partial reduction. We have observed that the starter nitroacid 2 disappeared just after 2 h; however, the purity and quantity of the reduced compound, i.e., pregabalin, increased with elongation of reaction time. We have also verified that Ni/Ra could be reused for at least 8 cycles before it lost in part its activity and required regeneration. We have also demonstrated that Pd/C and H2 or ammonium formate are equally good reductants allowing to obtain desired (S)-1″. 

Now this paper appears in OPRD, a journal well respected for the quality of the process work presented there. In my opinion this paper, while giving us a procedure for the synthesis of this important compound lacks some vital information.  I count 8 steps, not 6 as claimed in the paper. Plus there are several thermally challenged reactions/reagents in this sequence; hydroxylamine, nitrations, nitromethane di-nitro compounds, mesyl chloride, distillations and so on. But there are no details in the experimental nor the supplementary material to assist us in a thermal evaluation of this process. This data should have been presented, especially in such a journal. To present such chemistry, done on a 10 kg scale in a 20 L RBF, is  somewhat irresponsible, given the author’s comments (above). Proper cooling, stirring and thermal control is vital for such chemistry and has not been adequately demonstrated in this paper. So some caution is required if you wish to repeat this on a similar scale.

As such I don’t really see this as being a worthy successor to current production processes. Noteworthy, however, is the polymer bound organo-catalyst and a method for it’s production.

851 total views, no views today