During batch distillation, why do we collect heads, hearts, and tails fractions? How best do we process the heads and tails? Are there any alternatives frameworks? and what the hell is actually in all the fractions? Sadly, the literature is not the best about bringing us up to speed.
This is challenging to organize because there are a lot of balls up in the air. One of the first things to know is that we reprocess heads and tails fractions to do far more than just scavenge ethanol. Another thing to know is that for the most part, we don’t exactly cut things because we don’t want a particular congener in a spirit. Instead, we merely don’t want the surplus. In wholesome ideally fermented spirits, the two main congeners that exist in surplus are ethyl acetate and fusel oil.
[This point about ethyl acetate is incorrect and was made because I was mainly studying aged spirits with appreciable ethyl acetate as well as heavy vinegar process rums that had surpluses of ethyl acetate. Something that has slowly dawned on me from doing practical work is that the heads cut is far smaller than most people think. Arroyo’s heads cuts for a batch column spirit that saw an amount of full reflux before collection was 1 oz. per gallon of wash. The most major compounds in the heads cut are high boiling point congeners rinsed out of the condenser after the last distillation which makes the demisting test extremely valid.]
Surplus ethyl acetate is in the heads and surplus fusel oil is in the tails. The level of these congeners also vary somewhat batch to batch. Recycling these fractions to the next distillation run helps to average things out and ensure the hearts fraction has a consistent level. In pot still double distillation, because the heads and tails are higher ABV than the ferment, recycling averages up the alcohol content and helps increase the ability to fraction.
[When we consider this averaging up, keep in mind the heads volume is quite low relative to the tails volume so most of that averaging up comes from the latter. Kervegant provides some remarkable data and notes a heads cut for double retort pot still distillation can be as small as 1 gallon per 1000 gallons of wash and there is actually two different tails cuts! One thing this means is that they want all their ethyl acetate!]
Another important idea to introduce is that most all high value congeners (HVC’s) are less volatile than fusel oil. If you max out your fusel oil quota and therefore need to change from collecting hearts to tails, your tails may be ladden with gorgeous HVC’s (esters and high value terpenes/ketones). The birectifier can easily analyze a tails fraction and it is extremely insightful to see what of high value didn’t make the cut.
Heads and tails cisterns shouldn’t be mysterious cauldrons of despair. There is very little that is truly nasty in them, it is simply surplus and the surplus will be less than what ends up in your final spirit. In many cases it is also labile and thus prone to chemical transformation. A large percentage of your acetaldehyde is labile and over time will change to acetal. Acetaldehyde is perceived as prickly and rough while acetal is perceived much differently. In a short amount of time, that roughness will erode away. It is important to have an intuition for the trajectories.
We fall for thinking many ordinary congeners are nasty and toxic instead of merely surplus and manageable, because they are perceived one way above a certain threshold and significantly differently below. We also often see them out of context as they cross the condenser without synergistic relationships they rely upon to be beautiful.
The most common fault of new American spirits is that they don’t allow enough ethyl acetate. Above the recognition threshold, ethyl acetate smells solventy and non-culinary, but below it smells generically fruity and it facilitates many synergistic relationships. New distillers are often over eager to cut it away because they experience it in one big burst where it is above the recognition threshold. Ethyl acetate however, must be in proportion to the longer chain esters they need to support. You can only establish the relationship once they are both measured.
[This is slightly valid, but the thing to note from surveying more spirits is that the vast majority of ethyl acetate (besides vinegar process rum) comes from wood maturation. Ethyl acetate becomes one of the most significant markers of maturation and this is reinforced by many newly discovered Bourbon papers that have come to light over the years. Heavy rum turns out to be a very odd duck concerning ethyl acetate and not a good place to extrapolate any lessons to other spirit types.]
The birectifier is a great way to learn how much ethyl acetate is permissible relative to a high value congener content because we can deconstruct great rolemodels and follow their lead. In the first fraction from the birectifier of a mature spirit, ethyl acetate is near always concentrated enough to feel solventy, non-culinary and inharmonious, but a role model proves that to be normal. It must be considered that birectifier fractions are roughly concentrated 10x and then the first fraction is traditionally diluted 3x for nosing. Its take experience to get the hang of evaluating congeners in a isolated and magnified state.
Ethyl acetate also evaporates during maturation so often the desired amount must be overshot so that a degree of surplus will dissipate in the angel’s share. You cannot make a mature bourbon your role model for a new make spirit intending to spend time in a barrel. You also cannot make a full bodied rum your role model if your own distillate does not also posses a significant HVC fraction.
[The opposite of this last statement may be true in normal spirits. When we start to get deeper into the data, possibly only rum (and some other odd flawed ferments) may have overly surplus ethyl acetate. Ethyl acetate accumulates during maturation, but possibly first because it is broken by hydrolysis when spirits are watered for barrel entry proof. After that, the oxidation of ethanol may yield most acetic acid to slowly form more ethyl acetate. I just got some great papers from the IRS with more detailed data what will be helpful here. Many new American spirits may be too light on ethyl acetate due to over cutting as well as lack of maturity.]
Surplus congener accumulation needs to be tracked because a percentage is going to need to be removed from the recycling loop if the cuts are going to be held constant (by either volume or proof/temperature). This unrecycled surplus can be processed in special allotments. Sometimes heads and tails surplus allotments can grow to the point where the surplus has just as much ethyl acetate and corresponding fusel oil as a typical spirits run so the distiller is tempted to create a special run they call a queen’s share. In theory, the fusel oil fraction is dragging along an abnormal amount of HVC’s to make it all worthwhile.
If the stars align, these odd batches have the potential to be above average in quality. In practice, the surplus scraps are somewhat miss matched and a spirit is created with an uneven congener profile. Notice, the goal clearly isn’t about scavenging ethanol, but rather is about capturing HVC’s mixed in with the surplus fusel oil. It is hard to create a stand alone queen’s share without analysis keeping track of the scraps, but something worthwhile easily can be created as a special blending stock or to be averaged into a hearts cistern.
[I don’t want to delete this but would caution what follows because it is probably not valid for most all distillates because they likely do no have excessive ethyl acetate from the ferment.]
These E’s represent all the ethyl acetate in the charge to be distilled. The dash- represents the point the cut is made to remove the surplus.
If ethyl acetate accumulates, the cut either has to be moved or the distillation proof increased to keep what makes it to the final spirit constant. The bold E represents where most extra ethyl acetate shows itself. The cut has to be moved to keep a constant level in the spirit.
Another option is to reflux the still higher for a duration (if you have a column still), before dropping to what you would normally distill at. In practice this last idea may be hard to execute consistently.
This attempt at graphical representation has horrible limits. One thing to note is that as ordinary congeners like ethyl acetate increase, due to the nature of distillation and phenomena like entrainment, they don’t only get denser in a zone, but rather trail off stretching further down the run. With birectifier fractions, when a distillate has a lot of ethyl acetate, organoleptic judgement comes mainly from the second fraction and not the first where most has accumulated.
Managing all of this to create consistent spirits with even congener profiles becomes tricky without analysis. The birectifier is a useful tool to help understand the flow of your surpluses so they can be managed.
[Even though a large degree of this is wrong, the birectifier is the tool that has helped us to understand it more clearly.]
Ethyl acetate is not too hard to understand because there is little that is high value around it, but managing fusel oil is make or break to the quality of a spirit. You may find that there is tons of HVC’s that just aren’t making their way into your hearts. Soon the limits of distillation to create an even congener profile are realized.
Because a tails cut jeopardizes HVC’s, it is wise to do everything possible to reduce fusel oil at the point of fermentation. This is the big lesson from Rafael Arroyo and why he referred to fermentation as the climax of production and not distillation. In most cases, the aim is not about producing extraordinary aroma, but rather reducing ordinary congeners that block HVC’s. Fine beverage distillation is incredibly holistic.
The less ordinary surplus congeners you must cut away, the lower proof you can distill at. This has many benefits because of the potential for aroma breakage when spirits are cut with water to drinking proof. Esters have the potential to break and the rarest & highest value congeners like the Ionones and Damascenones are known to come out of solution.
An important thing to pause and note is that direct consumption ferments have far different relationships with ethyl acetate and fusel oil than fermentation intended to be distilled. Many ciders actually benefit from more ethyl acetate because it can make them seem fruitier. Producers often seek out yeasts that produce more. Fusel oil has a similar story. Everything is very different with ferments intended to be distilled.
[Some of these ideas come from Arroyo who was not too enthusiastic about ethyl acetate, but he may have been wrong.]
As ethanol increases (after distillation), and there is a decrease in non-volatile sensory features, the needs of the sensory matrix change and the surplus concept becomes important. Yeast strains aimed at distilling are often selected for being low producers of fusel oil. Besides a yeast’s genetic propensity, surplus congeners can also be reduced by optimizing yeast pitching concentration, nutrients regimen, and fermentation temperature. The birectifier is likely the most economical analysis tool to help monitor distillates while optimizing.
There are other frameworks besides heads, hearts, and tails, but they are poorly understood and really stretch the concepts. The main alternative comes from Rafael Arroyo and is derived from birectifier lessons which really have to be experienced first hand to be believed. The framework relies on long distillation runs with significant reflux to distill at very high proofs (against the azeotrope) paralleling birectifier operation. The fractions you end up with could be called: ethyl acetate, neutral, fusel, and high value. The final spirit ends up being a blend of percentages of each fraction. Such as 90% ethyl acetate fraction, 100% neutral fractions, 75% fusel fraction and 100% high value fraction. Arroyo even postulated that two spirits, a light bodied and full bodied, could be produced from one distillate. The more you play with the birectifier, the more plausible it seems.
[Arroyo actually describes a fraction before what I named the ethyl acetate fraction. It was collected between 67-72° and was completely discarded as well as likely quite small. This may represent congeners that are not surplus but completely undesirable. However, their volatility makes it likely they completely blow off in the angel’s share. If we knew their composition, we may find they are also labile and transform. Don’t forget at this point, we are also still demisting and collecting long chain fatty acids that were stuck in the condenser from the previous spirits run.]
The Arroyo framework may have advantages for processing special unmatched allotments as an alternative to a queen’s share. It may also have benefits when working with single batches with no opportunity for fraction recycling. Salvaging blindly inherited ferments, such as from wineries and breweries, may benefit as well.
I have long sought to make single tree fruit brandies, but it gets complicated by the notion that most distillates as we know them are the product of interconnected batches. How small scale should your equipment be if you are only getting 30 bushels per tree, and could you rival the traditional product of the HHT framework?
The Arroyo framework stretches a few more concepts and keeps it speculative. Do we cut what is only surplus or is there anything that we truly do not want? Analyzing a tails fraction with the birectifier from a full flavored rum, I encountered many soapy barely soluble components (that precipitated to form a soap ring) in the HVC fraction that were not in the corresponding fraction of the normal distillate. Could they simply be chill filtered or do they pose a problem to the method?
Another limitation, that may be easily solvable, is that fine spirits typically contain significant noble volatile acidity. They are basically tart. The higher the proof you distill at, the more these get pushed to the back. To capture them, you would have to continue distilling only water beyond the typical high value congener fraction. The importance of noble volatile acidity is not properly recognized in spirits quality and will require a lot more exploration.
Yet another limitation is Arroyo’s idea of aroma breakage that was already briefly mentioned. Broken esters are thought to reform after durations as brief as six months. We do not know enough about broken components such as Ionone and Damascenone (coming out of solution). Today, we have better tools to investigate this kind of thing than ever before. Resolving issues of the Arroyo framework would make an incredible PhD dissertation and greatly benefit the industry.
These ideas are for more than just distillers. Microbiologists and fermentation chemists need to be aware of these concepts and basically the limitations of the still itself to understand the importance of reducing ordinary surplus congeners at fermentation. Analytic chemists with expertise in chromatography that are tracking exotic extremely high value congeners like the Ionones and Damascenone must be aware that their work is bound to ordinary surplus congeners like fusel oil. We may maximize Damascenone, not by producing more, but rather by reducing what blocks its entry into a spirit.
Optimal cutting routines require a lot of involvement. Many new American distillates are currently limited by uneven congener composition, but it would take only very basic analytic work to be distilling at the top potential of a ferment. The birectifier is an economical tool to assist the new distiller while requiring the minimum of science relative to other methods.