Follow along: IG @birectifier
A few months ago I used a Jamaican technique trusted to me to establish infected dunder and observed its activity for four weeks. This dunder took on substantial titrateable acidity and the pH moved around quite a bit, both up and finally down. The aroma evolved from cheetos to bubble gum and back to cheetos. Eventually I wanted to determine what aroma was where so I used a birectifier technique where a sample with no alcohol is subsidized with 100 ml of absolute alcohol and fractioned. This gives the birectifier the ability to meaningfully evaluate much more than just final distillates.
To prepare my sample I added 115 ml of 87% light ethanol and then 185 ml infected dunder to create a total of 300 ml (33% ethanol charge). This is slightly larger than the typical 250 ml charge, but it worked out wonderfully. I added a single drop anti foam and 2 pumice stones to prevent bumping.
Learning from the muck experiment, I created a fraction 4b because I suspected there would be residual ethanol in the dunder. I ended up collecting about 7 ml in this fraction and made the transition to fraction five when the temperature on the dephlegmator thermometer reached 82-83°C. Decision to switch was also coupled with visual observation of water vapor moving up the column.
Something to note is that there was no emulsion in fraction 5 which may imply there was very little esters despite a significant time under heat and high acidity as well as high starting ABV. Acid catalyzed esterification in the still may not be as significant as bio-catalyzed esterification in the fermet. This dunder should also have had residual acetic acid, but I never detected any. Possibly it was elongated or just obscured in the sensory matrix?
Evaluating the fractions illustrated two incredible things that could not be observed just by smelling the ferment. The first significant observation is that this bacteria culture was capable of chain elongation meaning it converted acetic and lactic acids into much higher value longer chain acids. Secondly, and quite significant is that the bacteria was able to produce rum oil.
Bacteria producing rum oil is never acknowledged anywhere in the literature and no scientists to my knowledge have studied infected dunder in isolation besides data on pH, TA, and density. What this means is that there are now two demonstrated methods of producing rum oil. The first would be the high pH method attributed to both Rafael Arroyo and Batavia Arak where rum oil is produced by fission yeast and now we see that rum oil can also be produced by extremely active bacteria under high acid conditions. Fission yeasts may still participate as the symbiotic partner in high acid ferments, but may not actually be responsible for the rum oil. This begs a lot of extra study.
The lack of esterification also needs better study.
Fraction 1: Extremely light, but there almost seems to be an extra note possibly of acetaldehyde. Nothing offensive or overbearing.
Fraction 2: Diminutive version of fraction 1. There is a faint extra note I would simply call dirty.
Fraction 3: Very neutral despite an emerging dirty acrid note.
Fraction 4: Definite notes of fusel oil, but probably no more than the light ethanol I inputted.
Fraction 4b: This was roughly 7 ml and implies how much alcohol in the dunder pushed the analyte past 100 ml of absolute alcohol. It has the same character as fraction 4 (Keep in mind, all the previous samples may have had a lower than usual ABV because the analyte size also increased).
Fraction 5: Rum oil character? This was not obvious when the fraction was coming across the birectifier unlike other examples of rum oil created using fission yeast. It either has a different rum oil character than I’ve seen expressed by fission yeast or there is another aroma layered over top of it. Acidic and there is actually a mouth feel to it, almost a thickness. [6.6080 mg/mL 16.506 ml titrant / 15 ml sample] Forgot to smell neutralized fraction.
Fraction 6: Very acidic on the palate. There is almost a dirtiness, but I’m not too sure if its just the character of acids beyond a certain threshold. As it lingers on my palate, a cheetos character comes into focus. [8.0094 mg/mL (g/L) 20.056 ml titrant / 15 ml sample] Neutralized fraction had no overwhelming aroma, but a subtle cheetos character.
Fraction 7: It feels like a duplicate of fraction 6. At this point I’m curious about neutralyzing the acids and seeing if any rum oil character persists underneath. [7.0589 mg/mL 17.632 ml titrant / 15 ml sample] Neutralized fraction had no overwhelming aroma, but a subtle cheetos character.
Fraction 8: I’m either anosmic, tasting this last fraction, or this one has less aroma. There is still very definite gustatory acidic on the palate. [5.6184 mg/mL 14.034 ml titrant / 15 ml sample] Neutralized fraction had no overwhelming aroma, but a subtle cheetos character.
Stillage: 5.0510 g/L / 8.411 ml titrant. Neutral stillage retained a fishy aroma.
The success of this investigation is making me think I can best use the birectifier to evaluate active ferments in the same way but cutting the added ethanol to 85 or 90 ml and possibly increasing the overall size of the charge to 325 or 350 ml. This may be a lot more useful than simply trying to use enough ferment to contain 100 ml.
I have previous experimented with evaluating ferments by switching to a larger boiling flask, but this dramatically cuts down on fractioning power and I contemplated building a taller 8 bulb birectifier (which historically existed). This all may not be necessary which would save a lot of equipment expense as well as having to develop a second heating routine.
To get these same insights from a ferment directly or a product like muck or dunder, a distillery would require liquid chromatography in addition to their normal gas chromatography so the birectifier method gives incredible versatility with no extra expense.
7/5/23. I used the same process on some high dunder I was investigating. The only interesting thing to note is that even at a high pH of 5.8, there was a lot of volatile acidity that made fraction 6 tart like lemonade. There was no disagreeable sweaty character. The fifth fraction had nothing obviously high value, but seemed to have a yeast body ester character.