bostonapothecary

I've been working with a brilliant micro biologist recently. He has isolated Arroyo's specific butyric acid producing bacteria and it may be available soon on a slant. Harnessing it productively, as @Silk City Distillers said, is another story. You may need to invest in more analysis than most small operations are currently using. The birectifier is likely useful and we're about to start using it on some model ferments. The delta acidity concept is likely also very useful, but you may need to invest in a nice automatic titrator.

Recently, I used the birectifier to look at the tails fraction of a full bodied rum. The results were very cool and quite insightful. Fusel oil was shown blocking a significant portion of high value congeners from entering the spirit. If fusel oil could be reduced, more HVC's could be captured and spirit quality improved. The birectifier made examining and making sense of it fairly easy. We generated great actionable insights with the minimum of hard science.

I think the other limiter is that temperature is hard to know exactly. It may give you a pretty big margin of error. You're making me feel a lot better about sticking with beverage distilling. Its a lot easier to wrap one's head around.

I know temperature is an unreliable thing to measure, but could you reflux it and try to measure the temp of the azeotrope to make sure it is in spec? Or are you concerned with measuring what is left behind in the oil? I thought the tax thing would not be that big a deal because you'd be able to reuse the solvent over and over?

I wrote a post on what to consider when making cuts. A lot of the ideas are based on using the birectifier as an analysis tool to ensure you are distilling at the peak potential of your ferment. This was brought on by a new distiller's questions and by a unique case study where we looked at a tails fraction and found a startling quantity of high value congeners that were just not making it into the hearts. Not many realize that what is high value is less volatile than fusel oil. Something I'm consistently seeing in new American spirits is that they are too light on ethyl acetate. Simple analysis, especially of role models, may help hone in on ideal amounts.

Oh wow, I would have thought heptane was from the fuel ethanol world and not the cannabis world. Doesn't it have properties that can help dry ethanol?

This is way over my head. I had to look up heptane. I don't have much formal science background besides reading obsessively and spending a ton of time playing with things. My interest in analysis has been learning the minimum to bring larger dreams to life. The big lesson from the Judgement of Paris in the wine world was that fine wine was born in the lab. What all the wine makers that won had in common (Mike Grgich, Warren Winierski, et. al.) was that they were routine lab work guys from larger operations like Mondavi. To execute their vision of fine wine, they knew it required the lab. They weren't exactly scientists and were only sitting in on night classes at UC Davis. The birectifier stuff tries to parallel that story. When it comes to something like a denaturant, it is way over my head, but I do think I know who to ask.

Productivity has been a hard lesson to learn. Accurately measuring ABV is such a time suck that the Anton Paar tools keep seeming more attractive. It's been slow progress coming up with an alternative. I've got my glass guy trying to make a jumbo Lang-Levy pipette to replace the really poor quality pycnometers on the market, but its been back to the drawing board a few times.

Hi Roger, There clearly isn't much interest here, but interest/sales are coming from the Caribbean and the largest new players. There is also interest from university programs and a few PhD projects. There definitely isn't much of a profit motive. Some of these congeners have extremely low thresholds of perception, so very small quantities are surprisingly very meaningful. Best bets are emerging on how to maximize them just through practical operation and nothing exotic. It is kind of like trying to cook the same dish as a chef with modest ingredients. You make something edible while they take the same ingredients, seem to understand a few more variables, and manage to create something far more appealing with no more effort than you. Ingenuity. During maturation, we know what is changing. During distillation we know what is going where and what is being formed. Garbage in, garbage out. Fermentation is the climax of production. The U.S. and other countries conducted decades of public research for entrepreneurs to build sound private enterprises upon. It blows my mind that it is so easily dismissed. The birectifier project aims to give one affordable and practical tool to use to compete against deep pocketed multi-nationals with million dollar labs. To have no tools but a pH pen is kind of crazy especially when people are investing millions. You can lead a horse to water, but you can't make it drink. Hopefully it should at least be comforting to know tools and ideas are readily available. Good luck.

I don't know where people are with exploring chromatography based analysis, but I just wrote a fun post about birectifier assisted chromatography. In the fall, I attended a conference in Jamaica where pretty much all the distilleries used chromatography (and ran 3 shifts around the clock!), but all wished they did more with organoleptic analysis and possibly olfactometry. A lot of the compounds producers are missing are derived from carotene and very hard to look at. These turn out to be the highest value components in a spirit, be it rum or bourbon. The birectifier becomes the cheapest way to evaluate them and the smartest way to contextualize them and tie them to decision making during production. This post was in response to three different university labs requesting birectifier samples. They wanted to look at what was in fraction 5 and possibly name the compounds that Arroyo called "rum oil". They keep insisting on 100 ml samples, but I had to explain that traditional birectifier samples are already concentrated 10x! If I sent one to further concentrate for chromatography, it would already be a better than average starting point. We'd also have more context than usual because we know the band of volatility we are looking at and we can repeat the birectifier work on the cheap. I'm making slow progress, but the birectifier is attracting some of the world's great labs and hopefully will be a part of research papers that will be especially useful to craft distillers.

This is quite exciting then. Any places I can read more about it? Awesome to hear your successes out West.

I think the precedent is Hendricks. Don't they blend two different distillates? Compounded gins are typically thought to be inferior, but I wouldn't worry about it if you were stuck in that category. I'd control the language and say "compounded from two distillates to fully realize the potential of the rare botanicals".

The higher you go above your fill level, the lower the temperature, relatively, because of the higher alcohol content in that zone. One of the differences between a suspended basket and boiling is also time under heat. Heat up can be quite long, dramatically increasing the time under heat until vapor makes it way to the column. Co-distillation of botanicals may reveal botanicals that are limiters as far as where cuts can be made. Traditional gin botanicals likely all have a lot in common as far as volatile aroma that would influence cutting decisions. If you add odd botanicals that require a unique cutting routine, you may need to produce the gin from at least two distillates which is doable, but increases production complexity and introduces consistency issues. I've developed a birectifier protocol so you can systematically learn where in the distillation run each botanical's characteristics are volatile and where a limitation may appear. Fractions can be faithfully compared across botanicals. This kind of homework can add a lot of insight before you scale up and start investing a lot of money per batch.

It is sort of like scalpel meets stethoscope meets magnifying glass. When you isolate the congners you can do so much more organoleptically. I'd be skeptical of anyone that could taste a typical spirit and make a hard pronouncement. Spirits have a lot of mystery. I think the first thing needed is to survey exemplary role models so it is known what is normal in each category. Most new American distillers I've surveyed are off on their cuts. When you compare yourself to a realistic role model, you can change distillation parameters to narrow the gap. The precision of the birectifier gives you confidence in the comparisons. Once you're distilling at the peak potential of your ferment, anchored by role models, its time to get into yeasts, nutrients, raw material etc which is some absolutely vast territory. Fine wine makers have pursued nuance and improvement to the ends of the earth and eventually craft distilling will start to make those explorations. Because of practicality, the birectifier is likely the tool that will make the biggest difference.

I need to take @Foreshot's advice and make it down. I'm about to move to Philadelphia so I'll be much closer. This project hasn't been easy. Reintroducing a still design takes many 100s of hours. I spent an insane amount of time on all the library work and translation. Developing automation has been quite intense and I have a big closet of failed equipment. Processing all the case studies was a big investment. I hoped the initial purchasers would be new American distillers into simple things like making sure they were executing cuts properly and distilling at the peak potential of their ferments or developing a gin from scratch. What has come first is wild boundary pushing projects like aspiring to run 200 ferments with automation for a PhD dissertation (this project is quite beautiful), and of course the infamous "ethanol processors". A major university lab just made contact and they want to do a study about rum oil and possibly restart some of Arroyo's other work. Internationally, a few new projects that look like they're 10+ barrels a day are showing interest because they want a comprehensive analysis solution cheaper than GCMS and that can possibly better target fine spirits. A few of the new blender/bottlers want a tool to reverse engineer blends and learn more blending grammar, especially for rum. I'm spending a lot of time working on yeast stuff with a microbiologist from Chicago. We are trying to crack the Pombe code and learn to collect more strains. It is very preliminary. I also found two historic rum strains theoretically available. One is from 1913 Jamaica, possibly even from Hampden and the other is from 1923 Batavia. Access at the moment is not the clearest, but I do have a plan. The birectifier would be key to testing yeasts.

Are you selling to all the new "ethanol processors" out west? One of them flew to visit because he wanted a birectifier for lab scale extraction research. I know synthetic ethanol has been known for quite a while, but there was a big industry wide agreement, I think due to prohibition, never to use it. I think they just didn't want to shake public trust in potable alcohol production. I've only ever seen little bits and pieces about it. It was a taboo.

What happens is vapor goes up the outside of the initial chamber into the air cooled dephlegmator. The liquid then condenses and crosses the vapor trap which looks sort of like a little bridge and then runs back into the inner rectifier. At the bottom of the inner rectifier is another vapor trap and this is important for equalizing the pressure. The vapor then ascends through the bulbs then exists into a condensor not shown. At the top are two thermometer ports which can show quite a differential implying a lot of refluxing. I typically start the voltage regulator at 41 (41% of 115V) and typically end at 76 in the last fractions. Going from 41 to 42 can make a big difference in throughput so high end voltage regulation is very key. The reflux is also incredible to see. A little vapor phantom takes quite a while to creep up the column and a different kind of phantom creeps up again when you exhaust the ethanol and higher boiling point compounds appear. Condensation can tell a lot about how heavy a spirit is in the last fractions. Automation will have a PID correlated to throughput (not temp) so it will be able to operate at a pace even smoother than manual. It will also have really good data logging. Slight pressure aberrations are more prevalent than you'd think. If you had too much energy input, such as when you're cleaning the still by running neutral spirits, you'll see bubbles being blown in the lower inner bulb. We used modern threaded thermometer fittings and the joint for the condensor is a ball-socket style to reduce breakage. It should be at a 45 degree when setup properly or you'll restrict the opening and influence reflux, but I wanted more freedom for inexperienced operators when setting it up so its less likely to break.

I think a way to sum this up more succinctly is that the design has especially high fractioning power, beyond most designs, but very importantly with a level of repeatable precision that decisions can soundly be based off of. You're not left wondering if there is really more of an isolated congener in sample A versus sample B, or rather, if it is because of how it was fractioned (inconsistent reflux, inconsistent throughput).

Thanks for thinking about this @indyspirits. This seems worthwhile, but is something that I've struggled with because I didn't know exactly how the still works relative to others (in terms of theoretical plates and all that...). I've used a lot of other fractioning stills over the years, but none with such power. The birectifier went extinct because other proprietary designs like "spinning band" could do it better, but those are startlingly expensive and only owned by pharma and petro-chemical companies so the birectifier becomes relevant again because it is affordable. For some reason, many distillers say why "wouldn't I just use GCMS?" The first answer is of course price and the second is interpreting the data. Distillers distill, they don't absorb (chromatography), so sorting congeners with intense reflux by distillation is eventually far easier to interpret and put the information to use. It is all about being actionable. GCMS also has a lot of blind spots and the Caribbean distillers bemoan that it misses a lot in rum. They want to do more old school organoleptic sensory analysis with tasting panels. A yeast company I talked to mentioned they relied on GCMS but wanted an "olfactometry" mask for their unit because they knew they were missing important things. I guess you wear the mask, smell the output stream, and then click a button to mark novel character you smell. The big takeaway is that expensive state of the art methods aren't exactly winning and there is a lot of merit in the clever ideas of the 1940's. They're relatively cheap and effective, but somehow got mostly banished from the text books. When I found the design, I didn't know it would work. I simply thought at least one should exist because it was historical. It looked wacky and counter intuitive. I was very skeptical of an air cooled dephlegmator. I had also done similar fractioning on other stills. I was blown away by the early results. Only by experiencing the fractions did I realize how much better it was at sorting congeners than other designs. Each fraction is well delineated so you can turn around and make big decisions with the information. Each time I operated it, I started to appreciate a little bit more about the design. A big one is control. You have to collect all your fractions at the same pace (throughput) or they won't be comparable. The design allows you do this with one variable, just the energy input (which I found a great controller to simplify), so the wacky air cooled dephlegmator started to make sense. Other designs that rely on water cooled dephlegmators don't do it so easily. When it is easy and consistent, you can delegate it to the interns... (I'm my own intern :). Other things started to make sense such as why you always put in 100 ml of absolute alcohol and how this number pushes everything to the right places. When you distill at the azeotrope, there is just a few milliliters of absolute alcohol left when you hit the 5th fraction and somehow that brief zone grabs all the high value congners and isolates them. Experiencing fraction 5 can be so absolutely wild. So the big difference between other designs is superior congener sorting power and then ease of consistent operation so faithful comparisons can be made. It is hard to appreciate these points without experiencing it. Performance is at a level you can confidently make decisions based on what you experience. The fractions are easy to interpret by smell and taste. Most small scale distillers, even with science backgrounds, don't want labs. They don't want "wet chemistry" and they don't want GCMS, but they still have to solve problems and figure things out. The birectifier allows that. You just sort stuff and cut stuff up then smell it and compare it to a role model. It is very much the way of our five year old selves. Sort, cut, smell. Sadly, it is homework. Dave Pickerell said, "For some reason I can close my eyes and see molecules running around in a still, and I know where they’re going and why." This the goal, right? But we're only going to learn little bits of it at a time until we get there. I think the birectifier method is one of the purest ways of getting there. Smell things, put a face to name, and know where they are, when, and how they relate to quality.

In a new post, I describe how you can integrate Birectifier Analysis Of A Single Gin Botanical into gin development. This homework takes time, but the results are very insightful. With this framework you can know the volatility and character of every ingredient, inside and out, before you run a pricey batch through your big rig. Much of this work can be performed before you even approach investors to start a distillery. It is good to get your homework out of the way before you're burning rent on a warehouse. Coming soon, I'm going to make available a modern version of the Clevenger apparatus for steam isolation of essential oils. For a simple principle, its been harder to put together than I thought. I'm going into the third variation after changing glass diameters and adding features to make the results more useful in a gin context.

To be honest, I was trying to be a little bit of a humorous troll. Quite a few years back many rum enthusiasts were discussing the myth of raw meat in Guyana rum noted by Peter Valaer in the 1937 IRS survey of rums. Raw meat was very likely responsible for the "micro organism of faulty rum". Quite the strange and improbable rum story. I don't completely believe the saponification idea because I keep surveying spirits with the birectifier that have very significant non acetic volatile acidity. Acidity is far more important to grander sipping spirits than is commonly thought. Turbidity that comes and goes however is very common. I was just given a bottle of Hawaiian rum that has something gnarly floating in it. My guess is that there was a well distributed extremely thin collection of plaque like stuff on the insides of the bottle that was not rinsed and it slowly de-laminated and gathered into a wispy mass. Of course I'll still drink it.

The Micro Organism of Faulty Rum, a cautionary tale.

The only thing I'd caution about using this technique is that it may mess up your ability to make optimal heads cuts in full flavored spirits. This is because the technique targets aldehydes but the limiter for most spirits is ethyl acetate. In the opening paragraph he does mention it can reduce ethyl acetate to a degree. You could totally tackle these challenges, but you'd absolutely need an analysis solution to get you there. Every new American spirit I analyze with the birectifier has wacky first fractions relative to established role models. Most are too light on ethyl acetate. The birectifier gives a pretty clear organoleptic way to assess ethyl acetate without performing a titration. You can even take the first fraction where the majority of it collects and perform an exhaustive test of systematic dilutions to weight it. If you could tackle that, the technique may be a meaningful piece of the puzzle of reducing ordinary congeners. A lot of the Arroyo advice adds up to making better spirits if ordinary congeners are reduced in the ferment so you can distill much lower and cut away less.

Its really awesome to see this brought to life. I'll be sure to get that second paper in case there is something helpful in it. MIT computers have quick access to anything hosted by the American Chemical Society so if anyone comes across a paper they need, him me up. My understanding is that UC Davis has special collection of Guymon's papers and I always wondered if there was anything especially cool in there. I have his 1970's course syllabus which includes a lot of hand outs. UC Davis uploaded videos of all his distilling course lectures on youtube and I have a bunch annotated somewhere but I never found time to write up a post. It appears the Davis distilling course got high jacked by the petroleum industry. They strayed from fine spirits production and went theory heavy so the petro industry could poach smart students. There are still some absolute gems in the series and I remember a guest lecturer gave a pretty amazing talk on how distillers malts differ from brewers malts. There are some great recollections of just after prohibition as well as the war years, red wood trunk wooden stills, etc.

James Guymon - Some results of processing heads by fermentation (PDF) This link is for his original pilot plant article, but there is also another later one published by the American Chemical Society that I'll grab next time I'm at MIT. The Elie Skofis interview is available here where I index the beverage industry parts of the California Oral History Project. Skofis also delivered a really informative James Guymon lecture in 1983 and I've digitized it here. With the birectifier, in the first fraction of finished spirits, you don't really notice acetaldehyde because in spirits that are cut well, ethyl acetate dominates on a sensory level. However, I just started analyzing vermouths and on Noilly Prat, wow was there some nasty acetaldehyde. This was probably because of how they torture their wine by madeirizing it. The interesting thing is how all the other features allow our attention to metabolize it and it is not dissonant and gnarly. Many dessert wines hid scary amounts of acetaldehyde. The oral history of Antonio Perelli-Minetti is a thriller with him even sitting down to have lunch with Poncho Villa. A technique he used as a hustling wine maker was to buy spoiled wines, presumably oxidized with high acetaldehyde, and then to referment them in volumes of new wine which somewhat rehabilitated them. I think that is where Guymon got the idea. There was a known folk wisdom about the concept.