Aroma Breakage and Rum Design

[If you are a distiller, make sure you know about the birectifier]


Arroyo spends a lot of time discussing the topic of of aroma breakage due to diluting spirits with water. This is a curious topic I’ve never seen discussed in other research papers. I suspect this is because it is so hard to quantify. Apparently, the sudden shock of adding water to spirits, particularly rum, because much of its character is dependent on esters and rum oil, is capable of splitting the esters via hydrolysis and possibly precipitating aromatic substances as salts.

The ideal raw distillate, however, would be one needing no dilution or treatment of any kind before proceeding with its curing, either slow or natural; or accelerated or artificial. These conditions are attainable through proper yeast and raw material selection, appropriate methods of fermentation and post fermentation treatment of the beers; followed by carefully controlled distillation technic based on the principle of selective extraction. Present manufacturing methods with few exceptions, have not come to this high state of development as yet. Hence, mention must be made of the important changes occurring during the diluting, or the diluting and carbon treatment process during rectification.

I was first casually introduced to the topic many years ago by Ed Hamilton who explained to me the phenomenon as a past problem with Martinique rhums shipped to France for bottling. According to Ed, these rhums would be improperly diluted, rushing the process, and the aroma would be detrimentally effected. The science on the topic so far is not as easy to fully explain as I’d like, but there is one overall solution pointed out above and that is to distill at lower proofs so much of the water is already integrated. Easier said than done, lets let Arroyo drop a little more science:

When a freshly distilled rum is diluted, its chemical composition as well as its physical characteristics are affected. The action of the diluent is felt to a greater or less extent, in accordance with: (1) original proof at which distillation took place; (2) chemical composition of the original distillate; (3) nature of the diluent, and manner in which it is applied. The deleterious action of rum dilution is twofold: chemical and physical. The first consists in a dissociation of part of the ester content through the hydrolytic action of the diluent acting in an acid medium; the second operates by salting out or separating certain essential oils through shock, and by the lowering effect on the alcoholic concentration of the raw distillate through the addition of the diluent. Among these oils we have some of the most valuable natural constituents of a genuine rum. When the raw rum has been distilled at a very high proof, say between 170-180 degrees, a very large amount of diluent must be added in order to bring the proof down, say to 110 degrees. Now, the more diluent added the stronger will its hydrolytic effect be on the esters present in the raw distillate, and the stronger will be the tendency toward separation of essential oils. Hence, the caution previously given in the chapter on rum distillation that rum should be distilled at the lowest possible proof compatible with high quality of distillate. When the original chemical composition of the raw rum in its relation to the process of dilution is considered, it will readily be seen that the higher the free acidity and the ester content of the raw distillate, the faster and the more intense will become the hydrolyzing effect of the diluent on the ester content of the raw spirit. The same holds true on the salting out of valuable essential oils.

Lets stop him for a minute. It appears that the fullest flavored rums have the most to lose. It also seems like little inhomogenous zones get briefly created in the spirit where there isn’t enough ethanol to keep certain things in solution like the rum oil and, though brief, it is enough to shock stuff out. With the esters it is a little different and my favorite explanation of ester hydrolysis comes from Peter Atkin’s Reactions: The Private Life of Atoms. What I would think is that everything would eventually come back to equilibrium and that would easily re-dissolve the rum oil. I guess it doesn’t happen fast enough and there is a danger of stuff getting left behind when spirits are being only temporarily held in a vat that needs to be turned over for another process.

[Large droplet of rum oil out of solution in fraction 5 of birectifier fractioning.]

The nature of, and manner of adding the diluent, will also become a factor of importance during the process of dilution. There are at least six classes of diluents commonly used for this process: (1) ordinary tap water from the city mains; (2) well water; (3) rain water; (4) distilled water; (5) chemically treated water; (6) alcoholic solution in distilled or rain water, previously cured by either natural or artificial means. This last mentioned diluent is the least harmful and best recommended for this purpose, but is the one less used on account of the trouble of preparing and storing under suitable conditions large quantities of these weakly alcoholic solutions. Good practice should restrain the number of diluents to be three of the six mentioned above, and these should include; (1) distilled aerated water, (2) rain water, and (3) alcoholic solutions. Of the two plain water diluents, rain water is to be preferred, as it contains plenty of air and, therefore, lacks the flatness of taste peculiar to distilled water. Thoroughly aerated distilled water has the advantage of being readily available at all time. Diluting with aged mixtures of alcohol and water, or rum and water, will be the best method to follow; but as explained before, this method has its shortcomings especially for the large producer.

I remember reading this and wondering if Arroyo’s ideas like the ethanol-water mixture were arm chair ideas, dreamed up but never tried, or if anyone got it off the ground. Could it be worked into a strategy that would give a small producer an advantage when they typically lack any advantage due to scale? I could see column distilled relatively light rums being easily diluted much under 40% and then averaged up with stocks of aged, extra heavy, pot still spirit (to protect the pot still spirit which has more to lose). Sadly, I suspect not many small new Continental producers are making anything at risk of aroma breakage or using that template.

Now, whatever the nature of the diluent used, the manner of its application will have considerable influence in the extent of ester hydrolysis, and especially as to the degree of separation of valuable essential oils constituents. Cold diluent, suddenly added in bulk or added in a very short period of time, will prove the most harmful. The opposite conditions of applying the diluent, that is, slightly warm, slowly and in an atomized form, will prove the least harmful.

The water so many distilleries use is at the mercy of the seasons. They aren’t really in a position to raise a large volume of distilled water much above room temperature. I suspect the cheap alternative would be to use the still’s condenser as a heat exchanger and warm diluting water during distilling runs. I can’t imagine rigging up a atomizer with an agitator would be too difficult.

In the course of these studies on rum we have observed that the harmful effect of the diluent in rum diluting last over an appreciable period of time, never less than for three moths, and extending up to six months. It is really a very important matter that has been lightly, too lightly, considered up to the present time.

Whatever vats you have, you’d definitely need back sooner than three months and whatever precipitated would not be evenly distributed between the barrels or bottles. Whatever is precipitated could also be left as an oil clinging to the side of the vat. If the tail end barrel did get an appreciable extra quantity to try and re-dissolve it may not do it properly. Many blending practices where spirits are briefly re-barreled after marrying may be related to the concepts Arroyo is presenting. Whiskey’s may have different congener sets, but they may also be subject to similar forces.

Its will be noticed that strong dissociation of the ester molecules occur in every instance after the raw rums have been diluted, and that this state of original esters hydrolysis is not ameliorated even after three months of curing in an oak keg. The ester loss through hydrolysis is heavier in those rums originally high in ester content. Organoleptic tests on the undiluted and the diluted sample are consistently in favor of the former. Hence, rum dilution is really a complicated problem, and the ill effects of dilution interfere also with the proper and expected progress in quality during the first few months of the curing period. If to obviate the ill effect on the curing of the raw rum during the ageing period occasioned by the diluting process, the raw rums are barrelled at the high proofs of distillation, nothing is gained in the long run; for if it true that it will mature quicker during aging in this case, it is also true that its proof must be lowered before it is bottled up. When diluting the cured rum, the effect of the diluent will be more disasterous even than in the case when the beverage was diluted as a raw product. These considerations make the problem of dilution an arduous one. The solution must come during the distillation stage of rum manufacture. Raw distillates much be distilled at low enough proof so that no dilution, or at least very little dilution should be required during rectification and curing. And this must take effect in such a manner that the quality of the commercial rum shall remain unimpaired.

I have a feeling Arroyo was not always popular because he was usually advocating taking the long road and being thorough. To distill rums at a lower proof you’d have to design fermentations worthy of a lower proof. Or you might not have to if you’ve got the time. Arroyo mentions that rum, unlike a whiskey, can be enjoyed right from the still and will have no objectionable roughness like a white dog. On the other hand, Peter Valaer, in 1937 for an IRS survey, describes the legendary extra aged New England rums as being rough right from the still and distilled at 60% which is considered fairly low. Valaer also briefly mentions the use of “pure culture rum yeasts prepared from a single cell in some of the most scientifically controlled rum distilleries” in the U.S. Who knows if U.S. producer’s like Felton & Son’s conducted their own Olympiad to select a champion rum yeast, but it does look like they had the technology. I’m not sure if they used a Pombe yeast because I have not been able to find hints of their fermentation durations which would imply an answer.

Many spirits out there distill at curiously low proofs such as cachaça and many tequilas. They raise the question, are these cruder spirits like the New England rum Valaer describes or are they produced from unique ferments with less objectionable congeners than the average whiskey ferment? The answer is probably the latter. If one tried to optimize a fermentation to distill at a lower proof, what would it look like?

The answer lies in limiting ordinary esters and short chain fatty acids typically separated from the hearts fraction by the heads cut as well as limiting fusel oil which is hard to eliminate during distillation because it risks reducing rum oil. Fusel oil in rums is best reduced by careful yeast selection and then careful optimization of other variables like fermentation temperature with a lower temperature typically correlating to less. Fusel oil production tests were part of the yeast Olympiad I described recently.

Excessive ordinary esters get reduced in the ferment by salting them out. Short chain fatty acids like acetic and formic have a higher affinity for alkaline substances than longer chain fatty acids and thus combine to become non-volatile salts. If this is done with precision, which basically requires analysis and systematic trials, there will be less objectionable congeners in the ferment and distillation can proceed at a lower proof incorporating more water. Extraordinary esters of a lower frequency of occurrence, that form in both the ferment and the still, will suffer less occasion to split apart. New England rums in 1937 may not have been benefiting from the careful pH buffering that Arroyo was using to distill at lower proof. Harris Eastman Sawyer, who I identified as the architect of modern New England rum, sadly died in 1911. Though he was progressive, he was not likely to be that far ahead of anyone else.

By this point, most of you are probably thinking: why don’t you just super fractionate as explained previously, over dilute the central neutral fraction which is taking Arroyo’s advice of diluting with an ethanol water solution and see if that keeps more of the esters together? I didn’t realize it was that obvious, but what is left is to simply put it to the test and see what happens. And to hope the logistics aren’t a deal breaking nightmare.

Even more of you are probably thinking: where is Lost Spirits during these recent revelations on rum design? Lost Spirits has put most all their emphasis on post distillation curing ideas thinking that that is what makes a rum great. Arroyo would probably laugh them out the door. He time and time again explains that you need a distillate worth maturation to even bother, and there is little emphasis on that (so far) in the Lost Spirits camp. You also need to get a well crafted ferment to get a distillate worth a damn. But if you could bring all those things together it may well be more worth accelerated maturation and I hope they do. On the other hand, if you really nailed all those processes, you’d probably not feel the need to accelerate anything.

Next up will be Arroyo’s thoughts on curing, he didn’t like the term aging, and the ideas are more relevant now than ever.

Follow @b_apothecary

14 thoughts on “Aroma Breakage and Rum Design

  1. Very much enjoy reading your work and take on Arroyo’s experimentation.
    On this dilution question, would you say that there is a finite and lower proof you can take a distillate before all acids/esters precipitate as salts as you do with non chill filtered spirits and Louching? Usually I take my rum down to 43% vol and use water that has been though a Reverse Osmosis setup but I’ve yet to try a proper heavy rum through the pot so that will be the next test.

  2. Hi Bradley.

    There is just so much arm chair speculation here that our next step has got to learning the analysis techniques arroyo used and setting up some experiments. there are some known techniques to isolate rum oil and artificially concentrated rums could be created and toyed with. all the fatty acids/esters could be measured with simple titration. we could also find barrel proof examples from the Caribbean and try and mess with them.

    I suspect the salting problem applies more to the rum oils than the fatty acids. if you take it to 43%, what do you start with? and you can also do trials and take ten 100mL samples of your distillate and dilute them progressively further to see what happens. you could then perform titration analysis on each one.

    Cachaca and tequila are good places to explore in this territory. they are distilled extraordinarily low and see very little dilution and yet can be amazing. the big punchline is that by every degree that you want to get closer to this ideal, you’ll need more involvement in your fermentation and yeast selection. you’ll also probably need a thorough analysis lab.

    cheers! -Stephen

  3. Thanks for your reply Stephen!

    Yes there’s plenty of experimentation still needing to be done. Currently have a little dunder pit being started and attempting to infect using some cane I have grown myself (yes I have actually grown cane in Scotland…) since Arroyo talks about isolating the bacteria from the cane itself.

    I can’t remember where I read it but it was along the lines of the fatty acids go through the esterification process but can also break down and precipitate as salts but yes I know what you mean about the rum oils as well.

    For my my rum we start taking the hearts cut about 76% and stop around 60% and the total spirit collected sits around 66-68%. This is using a copper still from Hoga.

  4. I should also mentioned that the molasses I use are the catering grade that have invert cane syrup added to up the sugar content and the yeast is lallemand rum yeast and nutrient which isn’t actually that great for yield but it is good for flavour.

  5. Hi Bradley. very cool projects. you’re in Scotland making rum? cane in the green house?

    esters form and break at every stage of the process. all these compounds are also ordinary or extraordinary relating to frequency of occurrence and role in more systemic perceptions. in the Arroyo method, buffering became particularly important and he would salt some ordinary esters out to lock them up. this means they didn’t have to be sorted during distillation and they didn’t have to be worn off in the aging process. building all these sculpting tools requires analysis, but I think the ideas were fruitful enough to become part of other spirits productions like American whiskeys.

  6. Yeah, cane plant grown at home and in the greenhouse. Just felt like trying it and it worked but it doesn’t like the frost so it won’t last long over here! I’ve recently upped the scale from mucking about at home to small commercial (craft-scale) and launched product in June, our white rum is un-aged and very drinkable so hopefully going in the right direction thus far. I found out a good bit of history with rum and Scotland as it used to be made in the old Sugar Houses of Glasgow in the 1670s and was banned for a brief spell in 1695 as it was becoming more popular than Whisky and they tried labelling it as a Poison… I’ve been speaking with another local guy (isle of Islay) who’s starting up and we’re both interested in researching old methods to make proper rum as it should be.
    I think I really need to read Arroyos stuff a few times to let a lot of the stuff sink in. Might have to go old school with paper copies and the ink highlighters out to get a thorough grasp of it all! I’ve read that Hampden Distillery adjust the ph level of the muck pits to put them into long term storage during production shut downs etc so it’ll be interesting to find out what can be done over here but. I might get in contact with the local University to see if they’ll be interested in analysing samples of fresh rums vs aged Caribbean etc so it’d be interesting to see what we can do.

  7. Great article Stephen, thank you !
    Unfortunately this problem of agricole rums being diluted (destroyed) in France is not one from the past…

  8. Is there an references anywhere on using Stillage (Spent Still Wash) to proof back the finished spirit.

  9. Great to hear from you Bruce,
    I don’t think there is a tradition of using stillage from a second distillation because it likely has a lot of dissolved copper. You many not be complaint with regulations that would limit copper content. Very heavy rum stillage can even be green from copper and have long chain acids or esters in colloidal suspension appearing as flocs that don’t really settle. However, there is mention in the 19th century Jamaican literature of proofing down finished spirit with low wines. Many spirits were collected at 70 or 80% ABV which were customary proofs of commerce. It was apparently novel advice that if spirits were required for local consumption, such as on an estate, they could simply be proofed down with more low wines instead of water and this may also improve their flavor. How widely the advice was circulated or if it was context specific that it really improved a spirit is anyone’s guess.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Discover more from Boston Apothecary

Subscribe now to keep reading and get access to the full archive.

Continue reading

search previous next tag category expand menu location phone mail time cart zoom edit close