Recovering High Boiling Dipolar Solvents Miscible with Water for Recycling or Disposal At-Scale

Kilomentor has written blog articles covering the methodology in US 5628906 called Extraction Process that is particularly useful for working up reaction mixtures that contain suspended solids. These teachings are also applicable to simple, potentially lower cost recovery of high boiling organic solvents that are contaminated with large proportions of water.

Aqueous Drown Out Work Ups

Very often the simplest isolation of product from reactions in such solvents is a drown-out done by pumping the organic reaction solution into large volume of ice-cold water or water mixed with ice. This action where it works well precipitates the product which can be recovered from the mostly aqueous slurry. This has been done with dimethylsulfoxide, dimethylformamide, dimethylacetamide, N-methylpyrrolidone, sulfolane, and triglyme to name a few. 

After filtering off the desired precipitated solid operators are left with a homogeneous liquid comprising perhaps 80% water and 20% higher boiling organic solvent. This cannot legally be discharged into a sanitary sewer and would be expensive to send for destruction. Suppose instead we recharge this liquid (let us as an example say it is 80% dimethylformamide and 20% water) into the cold reactor from which the slurry just came. Suppose we then add what is called in US 5628906 a ‘secondary solvent’ ( let us, for example, use 1-butanol) in whatever, preferably minimal, volume that will cause the reactor contents to split into two phases; one predominantly water and the other, a mixture of the problematic high boiling water-miscible solvent (such as dimethylformamide from TABLE I in the patent ) and say 1-butanol (from TABLE II in the patent). Now we can cut the predominantly water phase and send it alone for waste. The mixture of 1-butanol and DMF, the primary and secondary solvents of the invention, can now be distilled to essentially recover both the 1-butanol as distillate bp. 117.7 C and DMF as pot residue bp. 152-154 C. The 1-butanol can be used in subsequent batches (it does not need to pass specifications since it does not encounter product). The DMF would need to be further purified to meet specifications if it is to be reused.

Note that in this example I chose 1-butanol to be the 'secondary solvent' to mix with my DMF because I expect there would be good hydrogen bond association between the alcohol hydroxyl and the amide while the two have a large difference in boiling points for the subsequent separation. If I were trying to separate a mixture of water and DMSO, I would choose toluene as my secondary solvent because I know there is good association between sulfolane which also has the sulfone functionality and all aromatics.