In process research, throughout the development, you can spend a lot of time chasing down and identifying side products both real and imaginary! Identifying and synthesising metabolites is also a vital part of the job.
The imaginary side products usually arose from the QA department. Now some of those working there have chemical knowledge and spend their days making process chemists lives hard. Not only in questioning every little detail about the specifications you have set for your compound but interrogating you on the possible side products that may just be somewhere in your API! For example, I was developing a compound containing a para disubstituted aromatic ring where one of the groups was a pendant carboxylic acid. The questions arose, “What happens if you have the meta and ortho substituted aromatic, does it do the same chemistry and have you detected anything that may be these final compounds?” Now my answer was; we have included those compounds in our specifications for the starting material and have never seen any of the other two compounds. So I had to make all regioisomers into the final API to prove that there were none there!
How many side products did I isolate from doing the following simple reaction?
There were four, three from non-selective reduction of the carbonyl groups (compounds A, B & C) and one (compound D) whose formation remains a mystery.
This accounts for the low yield of this reduction. Compound A was the significant side product and a chromatographic step was required to remove these compounds. This was especially important as those with free OH groups would also participate in the next step of the sequence. Now, these side products are not reported in the literature. The literature yield is >80% for this reduction, I never managed to reach that, even by varying reaction conditions the best I achieved was 65%. Coincidently to this post a publication has appeared dealing with the isolation, characterisation and chemical synthesis of side products found during the manufacture of dronedarone hydrochloride, an anti-arrhythmic agent acting on cardiac ion channels.
It would be very useful to the community if the publications reported the actual yields obtained instead of everything being >90%. Also, any side products observed should be adequately described.
Anyway I wish all the readers here a happy 2014, may all your compounds be crystalline and your side products characterised.
2,337 total views, 1 views today
