Saturday Night Synthesis: Chlordiazepoxide
Good evening and welcome to the Christmas's special of S.N.S. Now, the main feature of this special is that there is nothing which may even remotely remind you that Christmas is coming.
What we have, instead, is the nice synthesis of chlordiazepoxide. This is obviously a benzodiazepine, so, it has sedative properties. Because benzodiazepines will be featured in the Psycho Week, I'm not to describe their pharmacology.
For what concerns this particular drug, interestingly, it was the first benzodiazepine to be launched on the market and the biotransformations it undergoes, once absorbed, lead to remarkably active metabolites.
Now, let's take a look at this accidentally discovered synthesis.
We begin with a Friedel-Craft acylation, which requires zinc chloride, instead of traditional aluminum trichloride. However, because of the nature of the two molecules, the very first reaction taking place is between the amino group and the acyl chloride, yielding an amide. Then, the acylation occurs, but, because the solution will contains unreacted amines, these can attack the carbonyl.
Actually, this amino group will therefore yield an imine: the pH of the solution has been properly adjusted, I guess.
No matter, we end up with a very unstable adduct: that Schiff base easily reacts with water and this is very nice, since we finally have what we were looking for.
We add hydroxylamine and, then, that particular acyl chloride. And here comes the tricky question: which group will react with the carbonyl?
Now, this whole first part of the synthesis, so far, was designed so that the hydroxyl would attack the carbonyl, but, surprisingly, it doesn't, whereas that nitrogen succeeds.
We've got now another adduct which is likely to lead to a much thermodynamically stabler molecule, the final product, luckily.
Methylamine sort of ignites this 'rearrangement' or intramolecular series of reactions, if you prefer.
Please forgive me if the arrows don't perfectly point towards the atom but they seem to indicate bonds: I'm still learning how to use that programme.
What I hope you to check is if I've correctly drawn what reacts with what.
To sum up, by the way, a pretty successful mistake, eh?











