How to Remove Water From Isopropyl Alcohol

[a_k]

Assuming 40% for the vodka the mixture would freeze at around -23°C/-10°F, for 70% ethanol it would be -48°C/-55°F.

I am curious too about the magic temperature where the water freezes out and the alcohol remains liquid.

[Martin]

See: http://www.ashland.com/pdfs/technical/A ... 0Point.pdf

I don't think you can separate alcohol (ethanol, methanol, isopropanol) from water through such methods.

[a_k]

Nor do I. Anyway, since it was claimed to be possible, I've prepared two solutions with 40% and 50% ethanol and cooled them to -23.2°C. Both solutions were clear.

The 40% solution developed to a slight slurry upon stirring but the crystals redissolved because the beaker was removed from the freezer and the temperature of the solution was slowly rising.

The second solution remained clear. The water in an ethanol/water solution does not freeze out.

[holorefugee]

Freezing in this scenario would begin at a temperature significantly below 0. The first material to freeze would not be water, but a dilute solution of alcohol in water. The liquid left behind would be richer in alcohol, and as a consequence, further freezing would take place at progressively lower temperatures, and the frozen material, while always poorer in alcohol than the (increasingly rich) liquid, would become progressively richer in alcohol. Further stages of removing frozen material and waiting for more freezing will come to naught once the liquid uniformly cools to the temperature of whatever is cooling it. If progressively colder temperatures are available, the frozen material will contain progressively larger concentrations of alcohol, and the fraction of the original alcohol removed with the solid material will increase. In practice, unless the removal of solid material carries away liquid, the degree of concentration will depend on the final temperature rather than on the number of cycles of removing solid material and chilling. Thermodynamics gives fair assurance, even without more information about alcohol and water than that they freely dissolve in each other, that even if temperatures somewhat below the freezing point of ethyl alcohol are achieved, there will still be alcohol and water mixed as a liquid, and at some still lower temperature, the remaining alcohol-and-water solution will freeze without an alcohol-poor solid being separable. In practice, while not able to produce an alcohol concentration comparable to distillation, this technique can achieve some concentration with far less effort than any practical distillation apparatus would require.

OK, while this works very well at 8% I think it becomes impractical to go from 90% to 99%. The loss of alcohol is too high. I will stick with heat distillation. In two passes I can go from 6% to 90% and a third pass will take me as high as the humidity will allow.

[holorefugee]

For IPA at -70F you would be at 90% and at -130F you would be at 100% and have no unfrozen alcohol. Since Dry ice is only -69.5 90% would be the easiest. I am glad I have a 1L still...

[dannybee]

Gypsum plaster, a building and modeling material, would likely dehydrate IPA. As building material it is easily available and very cheap.
Gypsum plaster is "Calcium Sulphate hemi-hydrate".
Pour some plaster powder in the IPA bottle, shake well and let them react. 290 g plaster will absorb 54g water:
2 (CaSO4·1/2H2O) + 3 H2O→ 2 (CaSO4·2H2O)

Also "Lime plaster" (calcium oxide) and "Cement plaster" are potential desiccants.

( http://en.wikipedia.org/wiki/Calcium_sulphate .....reacts slowly with water to return to the dihydrate state, a property exploited in some commercial desiccants.)
http://en.wikipedia.org/wiki/Calcium_oxide
http://en.wikipedia.org/wiki/Plaster

It is worth to give them a try.

I tried this it seems to work,but how do you test iso for purity %? i could salt it to see if thers any water left. it seem to have color the iso a bit. its better to seperate than salting out the iso

[Tony]

So I would guess you can measure the Specific Gravity to see if there is any water Danny. But I doubt it would read accurately if it were say 98% IPA. But what I would worry about as I think you mentioned is any residue. If you were to use a hot blow dryer to it would it make some kind of strange film?

[Joe Farina]

So I would guess you can measure the Specific Gravity to see if there is any water Danny. But I doubt it would read accurately if it were say 98% IPA.

Yes, that's another problem. I have a hydrometer, but it would be difficult to see the difference in the 98 to 100 range. When it's all said and done, maybe it's best to just bite the bullet and buy gallon containers of 100 IPA from the cheapest source.

[Tony]

Yes, that's another problem. I have a hydrometer, but it would be difficult to see the difference in the 98 to 100 range. When it's all said and done, maybe it's best to just bite the bullet and buy gallon containers of 100 IPA from the cheapest source.

I agree Joe, but there is a green part of me that would like to keep the stuff as long as possible. Also, I never know when exactly it getting "bad". When results start going south, there are so many possibilites. It would have been nice to be able to recharge the stuff before each session.

[Dinesh]

Yes, that's another problem. I have a hydrometer, but it would be difficult to see the difference in the 98 to 100 range

It may be possible. Get an object (preferably a large object) that partially floats. Graduate it with horizontal lines or dots on the side. Put it in various known concentrations of alcohol (there'll be error bars, but the error will be dependant on the same hydrometer, so they should be consistent). You should then be able to extrapolate to the 98% odd limit.
For what it's worth, the proportion, n, of something with a density of (sigma)that partially floats in a fluid of density (rho) is:
n = (sigma)/(rho)

So, the (small) change in flotation height,
dn = -(sigma)/((rho)^2)* d(rho.

So, the denser the original object, the greater is the change in flotation level.