isop (Stranger)
06-13-02 00:20
No 320673
      Partitioning of substances between solvents  Bookmark   

If solubility of substance X is known in solvent A and in solvent B, and A and B are not miscible, can one simply calculate somehow how much will go into either when one has a container with a certain percentage of A and B? Does this go in a linear manner or are partitioning coefficients always determined empirically?

Couldn't find any info on this...

I'm wondering because of this example: GBL is known to be miscible with water and most organic solvents (e.g. DCM). Yet, I have seen more than one paper where GBL is extracted from water solution with a very small amount of DCM. So what makes GBL prefer DCM instead of H2O? Can such behaviour be predicted or even calculated in any way?

Thanks a lot in advance! crazy
06-13-02 00:42
No 320685
      Short answer  Bookmark   

No, partitioning coefficients can't be predicted in a reliable way.

Long answer, depends on the substance and solvents.
Things that hydrogen bond (alcohols and amines) can have really complicated partitioning effects.  Generally speaking most of the substance winds up in the solvent it is most soluble in.

In your example GBL is not very polar (someone correct me if it is highly polar) so it will dissolve better in a non-polar solvent.  Using triple separation even a 60:40 separation (assuming constant seperation) results in less than 7% remaining in the original solution.