Sciencemadness Discussion Board

Separation of isomers by selective seeding and crystallisation?

sludger - 17-8-2019 at 02:24

This is a suggestion for separating isomers that avoids the expense of a column.

My idea is that a solution that is supersaturated in two or more isomers will spontaneously deposit crystals of different isomers if left long enough. So I would imagine it would selectively deposit one isomer before the other if provided with a suitable seed of that isomer. The solution could then be transferred to a second crystallisation vessel to deposit another isomer. (It might not even need a seed as the degree of supersaturation will now be different for the different isomers.) The process can then be repeated.

To separate two isomers you would therefore just need a small seed of one of them. Even if you haven't got a sample, you could get some by spontaneous crystallisation on the first run. Select a likely-looking crystal and amplify your product.

A cursory search found me a lot about selective crystallisation but I didn't see anything about using supersaturation and a seed to create the selectivity. But it may already be a standard method... or maybe there's a reason why it cannot work. Obviously it will depend on getting conditions right and things not going wrong as well as the substance itself. I'm just asking whether it's worth doing experiments or whether it's dead in the water from the start. Sorry if I'm reinventing the wheel.





[Edited on 17-8-2019 by sludger]

[Edited on 17-8-2019 by sludger]

[Edited on 17-8-2019 by sludger]

j_sum1 - 17-8-2019 at 02:49

I think the idea is feasible. I suspect the practice may be a bit fickle: especially if the crystals are small.

Edit
I assumed you were asking about optical isomers. On re-read, you did not speify that.

[Edited on 17-8-2019 by j_sum1]

Ubya - 17-8-2019 at 04:21

what kind of isomers are you referring to? many isomers have different chemical and physical properties, so they can be separated by crystallization, distillation, etc.
if by isomer you mean enantiomer, crystallization can work
Quote:

5-10% of all racemates are known to crystallize as mixtures of enantiopure crystals, so-called conglomerates.[2] Louis Pasteur was the first to conduct chiral resolution when he discovered the concept of optical activity by the manual separation of left-handed and right-handed tartaric acid crystals in 1849. In 1882 he went on to demonstrate that by seeding a supersaturated solution of sodium ammonium tartrate with a d-crystal on one side of the reactor and a l-crystal on the opposite side, crystals of opposite handedness will form on the opposite sides of the reactor.

if not, you need to turn your enantiomer into a diasteroisomer (different physical properties), for example a carboxylic acid with a chiral amine.