A week and a half ago, a severe thunderstorm knocked out power for a couple of hours. Power outages and surges are kind of a pain: they mess up the fermenters, damage electrical equipment, and generally cause disruptions that last for two or three days before everything comes back online. As a result, there was a two day window where there were no samples to run and not much of anything to do at the lab. So I killed the time by doing a few of my own experiments.
The most interesting one was where I added about 0.2 mL each of methanol and isobutanol to 75 mL of high-purity 96% ethanol, and then diluted this up to 500 mL with water. Then I distilled the mixture in a lab retort flask with a Graham condenser, taking off several fractions through the distillation. The idea was to make a simple simulation of a pot still run with a 14.5% ABV charge and study how methanol and isobutanol (the second-most common fusel alcohol after isoamyl) concentrate in each fraction.
I took off a “foreshot” consisting of the first drips, about 0.75 mL, just enough to run on the GC. Then I took four fractions of 35-42.5 mL each, followed by a larger tail fraction of 110 mL. The ABV of the main four fractions fell from 68.18% for the first to 14.37% for the fourth, while the fifth was just 1.28%. There was far too little in the foreshot to run on the density meter; I guesstimated its ABV as 72%.
Methanol boils at 65 C, ethanol boils at 78 C, and isobutanol at 108 C. Many people think that what should happen is that the methanol will concentrate in the foreshots, the hearts are mostly just ethanol, and the tails will contain the isobutanol along with mostly water.
As you can see, this is very much the opposite of what actually happened:
MeOH-iBuOH Pot Still Experiment.xlsx- (44.26 KiB) Downloaded 189 times
Isobutanol’s concentration peaked in the foreshot and dropped off rapidly as the run continued. Meanwhile, methanol declined very slowly, and toward the end of the run it dropped off much less slowly than did ethanol. The ratio of methanol to ethanol did drop very slightly from the foreshot to the second fraction, but then it increased over the rest of the run - its concentration grew compared to that of ethanol beginning in the middle of the hearts and climbing through the tails.
The takeaway is that methanol comes out throughout a pot still run. It comes out disproportionately in the tails compared to ethanol and certainly compared to the fusels, but the majority of it comes out in the hearts right along with the ethanol. Discarding the foreshot doesn’t remove much more of it than would dumping a little portion of the middle of the hearts. Luckily, there isn’t enough methanol to worry about in really anything except brandy, and even there it still isn’t going to be more concentrated relative to ethanol than it is in the original wine.
The reason to dump the foreshot has nothing to do with methanol. A lot of other nasty stuff concentrates there instead, including acetaldehyde, acetal, ethyl acetate, fusels, and some other hydrophobic impurities. When distilling a fairly dilute ethanol solution, the water has a tendency to push out hydrophobic compounds, making them more volatile than you’d expect from their boiling points. Meanwhile, anything that is more hydrophilic than ethanol will be pulled in and will become relatively less volatile. Methanol is in this category, so it shows up more in the tails than ethanol itself does. Because the hydrophobicity of the longer-chain alcohols goes up with the number of carbon atoms, the relative volatility of the alcohols in a wash are in the reverse order to their boiling points, and methanol itself stays less volatile than ethanol up to about 40% ABV. This thread has more information.
Very interesting.