The first few spirits I chose to fractionate with the birectifier [$1800USD] were all junk rums donated to the project. The initial challenge was mastering the timing of running the still to keep the output at 25 ml / 15 minutes even as the alcohol level changed. I solved this a little quicker than anticipated by putting a digital kill-a-wat monitor on my Variac to help construct a chart that told me the volts (energy input) I needed for every stage. After a few runs I had enough bearing to hit the targets well enough. I was ready to look under the hood at a real rum.
The rum I chose to begin with was Pampero Aniversario which was definitely a first love so many years ago alongside the old fully palm leaf wrapped Ron Zacapas. I remember when they were each only $20! I really think they were different then and probably represented a glut of heavenly rum stocks. Today, they probably are a victim of their own success and the old reserves have been run down. No one to blame but us drinkers!
I’ll tell you though, I’m no birectifier expert (yet), but under the hood, with everything I know about distillation and sensory science, it is quite obviously a well made rum. I could only fault it for being, well, light, and probably playing it safe and not representing any extra delicious risk. Tasting as I write, the color wants it to be heavier, but it is just not. I don’t love it, but I respect it.
The birectifier and its companion organoleptic tests are all about triage. Once you collect your standardized fractions you can go so many routes. I’m not set up for any titrations so I dove into a modified version of the German protocol I previously described. Instead of using the protocol’s 20 ml diluted for drinking, I used only 10 ml. This leaves 15 ml to ask various other questions like testing surface tension or performing the exhaustive test. Per the protocol, I diluted the first four high alcohol fractions 3x with water and the last four fractions only 2x. Using a separatory funnel as a water tower, makes quick work of setting it up.
Whether you enjoy it or not, this rum is a role model, so pay attention.
The first fraction, Arroyo teaches us, is all the low boiling point, low molecular weight, esters and aldehydes. Above the recognition threshold they can start to taste kind of chemical and non-culinary, but in a wide band above the absolute threshold, they can sort of subliminally act upon us, tonally modify and create unique percepts with other less ordinary congeners. When just under the recognition threshold, you may perceive the ghost of a very generic hollow sort of fruit. I experienced it in the first fraction of some of the junk rums. The first fraction is wildly important.
Fraction one can be modified by yeast selection (and that is probably more important for fermented undistilled beverages). At another level, it can be modified by changes in fermentation variables. Varying cutting parameters and tuning the still modify it again. Much changes chemically or evaporates off during maturation. Finally, it gets adjusted again by blending. At many points where decisions are made, these congeners will be experienced above the recognition threshold only to migrate downward again. You must know the trajectories!
If you have to correct first fraction congeners by blending, you risk diluting other variables that perhaps you nailed perfectly! Tasting this fraction via a diluted birectifier fraction will likely be above the recognition threshold. If it is unpleasant, it does not necessary perfectly parallel what is experienced in the 40% ABV spirit. If you want to play and get the lay of the lay of the land, you could add up the absolute alcohol in fractions one and two and scale it back to a 250 ml. Constructing sketches can be really helpful during concepting and product development.
Pampero’s first fraction is above the recognition threshold using the German protocol, but not by much. If I executed the exact same protocol on another rum, such as my own, I could make a comparison that would be faithful and could be actionable. I could inch my spirit closer to greatness by following the lead of a smart role model.
Fractions two and three don’t have much importance. They both have roughly the same amount of ethanol as fraction one, but they decline sharply in congeners. The third starts to pick up some fusel oil. Likely no sub analysis has to be performed on these two fractions (unless you were set up for titration or want to sketch!). However, they still have to be executed faithfully to the 25 ml / 15 minute standard so that all the other fractions align.
Fraction four is the fusel oil fraction where 75% of that congener category lives. So much can be gleaned from this fraction depending on where you want to focus your involvement. I’m not expert, but all the intuition I have tells me Pampero is kind of light. It makes me want to shift my triage and perform an exhaustive test on the fraction to try and generate an indexed number for my log book. Is it low because of how they tune their continuous still? The fraction was not too frontal olfactory aromatic so I had to take a sip. It was offensive in magnified form, but still probably would be considered light. This fraction from this rum will be an important bench mark for comparing other rums. Do we re-dub Pampero, the light rum with the dark color? More experience will tell.
Fraction five regulates quality according to Professor Arroyo. However, no oily droplets lept out at me as I watched it drip with anticipation. Literally, I watched for a full 15 minutes thinking: when’s it gonna happen, when’s it gonna happen. Mythic droplets did not happen, but I am thinking of getting a magnifying glass. Pampero may not be divine, but it is quality and this fraction smells great. It is so much more pleasant than the junk rums I previously analyzed. To further analyze this fraction, it would make sense to perform an exhaustive test and possibly the other tests related to surface tension. I could also take 1.0 ml and perform the evaporation test. There is no excuse not to.
For the exhaustive text on this fraction using the German protocol, I got as far as 0.3 ml ca. 1:700. The evaporation test on fraction 5 is turning out much different than the same test on an unfractioned spirit. It is not to the same degree yielding to sour aromas because they are for the most part pushed into 6,7, and 8.
Fractions six through eight are all very similar and mostly water. The aroma is pleasant relative to the rankness of the junk rums I referred to above. Arroyo notes that in these fractions fermentation faults will present themselves. Pampero is in the clear. Each subsequent fraction does declines a bit in intensity. Upon tasting them, there is a subtle acid tang which may be attributed to long chain, high boiling point, high molecular weight, free fatty acids that did not become esters. Fatty acids from these fractions are likely what is removed by chill filtration so they do not louche after bottling. Does the dark color hide a louche or is everything baseline, and right where it should be? Only further investigations will orient this rum among others.
Since these last three fractions do not beg sub analysis we could perform a quick test, such as the sulfuric acid, test to see what happens to the aroma. If the aroma was attributed to fatty acids, it would go silent (remember, sulfuric acid renders this test smellable, but undrinkable!). That test would also help us close in on where exactly the brimstone aroma comes from and if it can actually tell us anything useful. Arroyo didn’t quite figure that one out.
Fraction 1 inexplicably changes color and becomes yellowed. The traditional scaling of sulfuric acid is not enough to destroy all of the aroma and a subtle fruitiness persists which is a big surprise. I thought it would be the easiest fraction to remove the aroma from. Fraction 3 is the only fraction that retains any brimstone aroma which is interesting because it is typically the most neutral of all the fractions. Fraction 5 is the only one to retain any aroma reminiscent of rum which is classically the indicator of rum oil. Fraction 6,7,8 have only the very faintest hints of brimstone aroma. None of this will be normal or abnormal until we can test more samples.
What is remarkable is how dramatically all the fractions differ and how succinctly they can sum everything up. Equal volumes and 15 minute spacing from 250 ml was initially arbitrary but practical, yet it completely works. It makes you wonder how much empirical effort it took to settle all the variables and the serpentine proportions of the birectifier. We have unearthed the work of geniuses. Our future work stands atop their shoulders; Micko, Luckow, Wüstenfeld, and Arroyo. Eventually we can integrate the work of the microbiologists Parfait and Fahrasmane.
This tool will rapidly add weight to any development decisions a distiller may make. It will also be supremely helpful to troubleshoot problems and help pass the torch as more tasks get delegated to other staff and finally to a new generation of trained distillers.
Very cool!