[ugachmaaz]

I've been trying to design a device to increase the percentage of ethanol in a water-alcohol mixture without using distillation.

The way I'm trying to do this is by using a dry ice bath (dry ice+propanol). I then insert a stainless steel vessel in the bath with the said water-alcohol mixture (between 30-60% ethanol to begin with). Going by the ethanol-water phase diagram and the temperature of the ice bath (-78C), I should, in theory, be able to concentrate the mixture up to ~80% ethanol.

Where this goes wrong is that instead of cleanly separating into water ice and concentrated water-ethanol mixture, what I get is a kind of ice-alcohol slushy.

Any suggestions on how to get a clean separation and get ice that I can scoop out?

Thanks!

[Arkcon]

What you're trying to do is called (somewhat incorrectly) freeze distillation, and has been well understood for centuries.  https://en.wikipedia.org/wiki/Fractional_freezing#Freeze_distillation

The Wikipedia page says if you get the ethanol-water mixture too cold, too quickly, you can freeze the alcohol along with the water.

If you can filter the ice out quickly, and begin to enrich the reminder for ethanol.

[Enthalpy]

With 40-70% water, it's no surprise that the mix gets thick at -78°C. Several cooling steps look better, where ice is removed each time.

A nice setup would use the obtained (concentrated) water and alcohol to provide cold to the successive first separation containers, and dry ice only at the coldest container. Not only would this create the intermediate temperatures, it would also save dry ice.

That is, the mix would flow in one direction among the containers, the concentrates in the other direction, and dry ice would be consumed only at the coldest one. As the melting point depends on the alcohol-to-water proportion, melting of a concentrate can provide cold to freeze a lesser concentrate. One should just check the amounts of heat, and whether some complement is needed.