Elrik - 15-10-2018 at 13:50
I got an old and partially used drum of denatured ethanol and I want to separate some into its component parts.
It consists of 80% ethanol, 10% isopropanol, 9% methanol, 1% MIBK
The manufacturer does not specify if the ethanol is 95% or dry, and the drum is old enough to buy its own alcohol, so I'm assuming the isopropanol has
had ample opportunity to form peroxides despite not contacting sunlight. Therefore distilling to high concentration could be dangerous, but I want to
reclaim 10 ml of that MIBK for a future experiment.
None of those components are reported as forming a formal azeotrope with each other.
My idea is this: Add some 10% ferrous sulfate (aq), set up my still with a 20 cm vigreux fractionating column and, with a good reflux ratio, take my
time distilling off the methanol and 95% ethanol to get good separation and then by the time the wet isopropanol began distilling over any peroxide
would have had ample time to react with the ferrous sulfate and when the wet MIBK came over I would not be distilling to dryness.
The MIBK could then be dried with salt, further dried with CaCl2 (anh.) and redistilled, this time to dryness.
Does that seem a safe way to process this? I'd prefer to avoid going through the bisulfite complex.
If anyone is suspicious of noobs, no I dont intend to drink the ethanol fraction. If I liked drinking alcohol I'd just brew it myself 
Deathunter88 - 16-10-2018 at 01:29
Seems like a good plan, you might have some trouble getting a clean separation on that MIBK though due to it's low concentration.
wg48 - 16-10-2018 at 09:25
As a general rule they pick adulterants that are hard to separate out of the denatured alcohol. Your denatured alcohol is no exception with one or two
of the denaturants within 2C of the alcohol. So your probably going to need a very long vigreux column or better a long packed column to separate
all the components to reasonable purity.
From memory to separate a two components with a 2C boiling point difference requires 20+ equivalent plates while a typical vigreux column 250mm long
is equivalent to only about 5 plates assuming strong reflux.
In short your going to need a much better column than a typical vigreux
Elrik - 16-10-2018 at 10:54
Yeah, I was expecting difficulty in getting full separation of the ethanol and isopropanol since, as you said, with water they boil 2° apart and dry
its still only 3.7°.
Half will most likely be used to make ethyl acetate and most of that will be used as solvent, but when I do need pure ethyl acetate those two are much
easier to fractionate. Some ethanol will be used to make other esters and ethyl iodide, possibly even ethyl ether if I need it and I certainly wont
want to make any isopropyl ether! At the worst, when I need much more pure ethanol I could reform it from refined ethyl acetate for small amounts or
ferment sugar beets for a litre.
Given the very low MIBK yield I might intentionally allow some isopropanol to slip into the MIBK for convenience because after drying they could be
easily fractionated.
My main concern is just not detonating a RBF at the level of my abdomen
macckone - 17-10-2018 at 07:10
If you are concerned about peroxides, add some sodium metabisulfite.
That is the standard method with isopropyl ether.
Isopropyl alcohol itself doesn't form a peroxide but some amount of isopropyl ether can form in anhydrous conditions. That will form peroxides.
Isopropyl alcohol will also oxidize to acetone which also forms peroxides.
There are a wide number of reducing compounds that can destroy peroxides, use them.
Don't distill to dryness regardless of rather you expect peroxides as a flask can rapidly overheat and shatter.
For the record isopropyl ether is one of the worst peroxide formers.
Wet isopropyl alcohol should not form the ether but why take chances?
Elrik - 17-10-2018 at 13:39
Exactly, I dont want to take chances. Thats why in my proposed method I included ferrous sulfate, as I know its used to clean up diethyl ether. I also
know its usually used in combination with strong acid, which I wouldnt want to add to slowly distilling alcohol and ketone. From the proposed
mechanism [1] it looks like the acid is just to keep ferric hydroxide from precipitating which shouldnt be a problem with low peroxide content and
distillation. I'm short on bisulfite but not ferrous sulfate, and my hobby budget is always limited so basically I'm hoping acid free ferrous sulfate
can take the place of bisulfite. I also have a fair amount of ascorbic acid.
Oh, and dont worry, I dont distill anything to dryness directly on a heating mantle. Thats what oil baths are for, I like my glassware
macckone - 18-10-2018 at 21:16
With the potential for peroxide formation the distilling to dryness risk is explosion, not cracking your glassware. I recommended bisulfite because
it doesn't require acid. your vitamin c should also work but it can be degraded by heat.
Tsjerk - 19-10-2018 at 00:24
I would add some bisulfite, and don't distill to dryness. Just discard the first 15 percent and last 15 percent and you will have some reasonable pure
ethanol. I don't see how the methanol or isopropanol would interfere as solvent (ethanol solvent or as methyl/ethyl/isopropyl acetate). Isopropanol
ether is no fun, but maybe this ethanol source is not the best for making ether.
MIKB could be separated from the last 15 percent fraction after distilling off the isopropanol, as the 75% azeotrope with water. It boils rather low,
and separates after distillation (sounds like steam distillation to me? source; click)
Elrik - 19-10-2018 at 11:24
To be clear, I never intended to distill the alcohol to dryness. Thats why I planned to add aqueous ferrous sulfate, so there would be water left at
the end. Its only the distilled and dried MIBK fraction that I would distill to dryness [in an oil bath].
Methanol and isopropanol wouldnt normally be a problem when its used as a solvent. The problems would arise when I try to make esters, ether, or ethyl
iodide. I'm just trying for an easier option than making and distilling sugar beet wine for reagent ethanol, and hoping to get 10 ml of MIBK as a
bonus
I'll look again for metabisulfite tree stump remover so I can avoid shipping fees.
happyfooddance - 19-10-2018 at 20:58
If you are worried about peroxides, I think the fears may be overblown, but in any case distilling a small amount first (50ml maybe) would probably be
a good idea.
Probably more work than you are looking for, but you could isolate any/all of those components via esterification/de-esterification of the whole lot.
If you are creative you might think of a way to provide yourself with several reagents useful to you with very few steps.