Follow along: IG @birectifier
The formol number is an approximate but useful index of nutritional status for a ferment and measures yeast assimilable nitrogen (YAN) which is the total of free amino nitrogen (FAN) and ammonia as well as ammonium. It is arrived at via a simple two part titration. Very conveniently the first part is an acid titration which means the formol number becomes extra information added on to an already valuable measurement and requires only one extra easy to obtain reagent.
Some of the best ideas on this topic come from a spectacular paper, Prediction of Prefermentation Nutritional Status of Grape Juice – The Formal Method by B.H. Gump, B.W. Zoecklein, and K.C. Fugelsang. All that I write that follows is encouragement to pursue the original document, possible applications in rum fermentation as well as notes to help me learn the process. Hanna Instruments has a lot of familiarity with this titration and supplied their method. However, it may need slight modifications for working with both higher acid ferments and smaller sample sizes.
This titration is most common to winemakers, but not all winemakers use it. It may be most important to winemakers challenged with working with substandard fruit or in stressful conditions such as dessert wines.
The formol number may become especially relevant to heavy rum as we pursue slightly unconventional nutrition. For high pH Arroyo style ferments, we are using ammonia hydroxide (as opposed to ammonium sulfate or diammonium phosphate. The idea for high pH ferments is to keep constant pH as high as 5.8 to maximize enzyme activity, and when successful with a suitable yeast, this can produce spectacular levels of rum oil. For these ferments, ammonia is incrementally dosed which has a multitude of beneficial effects as described by Arroyo.
High pH ferments are quite sensitive to cleanliness and LAB bacteria may eventually take hold dramatically slowing down fermentation if not making it seem stuck. Fission yeast may become inhibited by competition among bacteria for YAN as well as other nutrients like vitamins. There are some open questions which will hopefully evolve into strategies for both avoiding and playing through an infection. Sanitation can only help so much and another area to control infection is walking a tight line on nutrition. Keep in mind, before you get excited about a bacterial infection in this type of heavy rum, most all this acid is fixed lactic acid and therefore not immediately valuable to aroma creation. A formal number may give clues as to what is going in during an infection in this unique context.
High acid heavy rum ferments historically saw no added mineral acids such as sulfuric or phosphoric and no added ammonia as emphasized by Jean Guillaume who developed the grand arôme rum for Galion in Martinque. YAN was derived primarily from dead yeast in fermenter bottoms and additionally, in the Jamaica process, from broth in the muck cistern. For Guillaume, the lack of additives was also a marketing argument. These were the most natural of rums.
A formol number may help standardize YAN in high acid Jamaica style ferments and may also help observe progress in the muck cistern which is essentially peptonized yeast facilitated by lactic acid bacteria. Fermenter bottoms are highly variable, typically managed by depth measurements and are often drawn down because of the accumulation of molasses sludge from using unclarified molasses. My understanding is that some Jamaica producers currently add ammonia salts to high acid ferments, but that could possibly be avoided by more exploration of the nutritional value of the muck cistern.
There is also a lot to learn about competition for nutrients between yeast and bacteria in a high acid environment. How much YAN does bacteria absorb? Does mixed culture bacteria in saturated environments behave differently than in high pH environments? Could slow fermentation be slightly sped up by a better understanding of nutrition? Could we squeeze far more utility from muck and come up with measures for its progress other than acidity? (Remember, the formol number is a simple addition to measuring TA)
Wine literature is a great place to start drawing ideas for managing heavy rum ferments and basic winery lab skills not currently practiced in rum may hold a lot of value. We may find both numbers to relate two as well as strategies for abnormal situations. Fermentation duration is also a factor that can make heavy rum fermentation very relatable to wine.
Formaldehyde, the reagent for the formol number, is a mildly dangerous toxic chemical which will be new to most and we will have to learn to handle it safely. Formaldehyde reagent used is 37% which may also include a % of methanol. My understanding is that the formol number method require very small amounts, further diluted by the process, and are safe for the city sewer. Many official regulations deem formaldehyde non hazardous when the concentration is below 10% and the pH is between 6-9. There are products available that can be added to make the result even safer and fit for the landfill. Concentrated formaldehyde is mildly flammable, but that decreases substantially when diluted. This reagent may be about as dangerous as handling sulfuric acid in the lab. Education and preparationand may make using it no big deal.
The titration starts with neutralizing both the analyte (the sample) and the reagent (formalin) to pH 8.0 (or 8.2). Then an excess of formaldehyde is added. This releases acids from the amino acids and ammonia which can be further titrated to calculate YAN.
An important note on formaldehyde:
A new bottle of formaldehyde may have a pH as low as ~3.5. This will require about one-half milliliter of 1N sodium hydroxide to neutralize. If the formaldehyde is not neutralized, significant over-titrations may result yielding high values for fermentable nitrogen. The pH of the formaldehyde will begin to drop with time and should be periodically re-adjusted to pH 8.0.
The most accurate end point for the formol number is 8.0 but others can be used with a slight trade off in accuracy. I am more interested in the accuracy of titrateable acidity so I am going to use pH 8.2 as an endpoint. If you are batching the procedures, you just have to choose where you want your trade off in accuracy.
This is a two part titration. The first thing you do is titrate the TA of the sample and this simultaneously prepares the sample for measuring the formal number.
For wine, they start with 100 ml of analyte, but we will likely use 25 ml. Your titrator is setup with 0.1N NaOH, but you can also use 1.0N NaOH to get you to the equivalence point quicker. For wine, they typically start with 6.0 ml of 1.0N (100 ml analyte), but for heavy rum with over 15g/L TA it could be substantially higher. Experimentation will reveal a starting point. For every 2.5 ml of 1.0N NaOH added to a 25 ml sample, you can add 6.02 g/L to the TA.
The formol number result is expressed as ppm (mg/L) of fermentable nitrogen. Don’t forget upon completion, the analyte is now toxic and not even fit to smell before it is dumped.
For wine, a 100 ml sample from the first titration which is now larger and transferred to a 200 ml volumetric flask and top with water. From this, they take 100 ml of sample using a volumetric pipet. The original sample is essentially cut in half, no doubt to limit the use of formaldehyde.
After that, 25 ml of 37% (pH adjusted formaldehyde using same end point as the previous titration) is added. That is a scaling of 25 ml per 50 ml of the original sample size. Keep in mind, the formaldehyde measure is also a surplus so it does not have to be measured perfectly and is likely best done with disposable pipets. Hanna Instruments gives the advice:
NOTE: The formaldehyde should be adjusted or readjusted to pH 8.00, or to the specified endpoint before it is added to the sample. If no drop in pH is observed after the formaldehyde is added, the sample does not contain nitrogen in a quantifiable amount.
It may be common to come across samples, no doubt at the end of fermentation, that have no measurable YAN.
This paper comparing various methods mentions a critical step to titration accuracy is readjusting the pH of the formol solution every time it is used. It trends downward over time, but I’m not yet aware by what magnitude. I’m also not aware if the reagent becomes spoiled over time. Is it best to separate a volume to use in the short term and adjust to pH 8.0 or 8.2 (depending on method)?
Additional notes on what can be gained:
Yeast strain differences may result in different nitrogen needs and a number may help us optimize. Fission yeasts may exhibit different needs than budding yeasts and then their may be big differences among specific strains.
Winery ferments can be as extreme as high acid heavy rum ferments. Wineries are recommend to pitch yeasts at 2-5 million/mL when brix is below 24 and pH above 3.1. A brix of 25 is considered grand arome territory and even with a massive TA, the pH would be well above 3.1. The stressor for high acid ferments is not pH so much as titratable acidity. TA also needs to be subdivided among fixed acids like lactic and volatile acids like acetic, butyric propionic, etc. You can easily have a high acid ferment that produces a light rum if the only acid is lactic.
Nutrient numbers for wine: minimum 150 -250 mg/L
Recommendation exist to go as high as 500-900 mg/L
Yeast with lower concentrations of N may perform well under optimum, but not adverse, conditions. Concentrations of 500-900 mg N/L give yeast the ability to produce cellular proteins needed to meet the worst environmental conditions.
Phosphate deficiency may also have a direct impact on yeast cell growth and fermentative performance (11). Inorganic phosphate is required for synthesis of ATP and ADP and nucleic acids. Supplementation should be carried out using a balanced source of diammonium phosphate, DAP (25.8% ammonia, 74.2% phosphate), amino acids, minerals, and vitamins. Diammonium phosphate additions of 1 g/L (8.3 lb/1,000 gal) provide 258 mg/L fermentable nitrogen which exceeds the suppliers’ recommended level. In the U.S. the legal limit of DAP is 960 mg/L which corresponds to 203 mg N/L.
Amino acids are not incorporated equally by yeast and may vary significantly among yeast strains (14). Some are utilized at the beginning of the growth cycle, some later, and some not at all. Ammonia, on the other hand, is consumed preferentially to amino acids in growing populations. Stationary phase yeast also vary significantly in terms of the order of amino acid incorporation, but do not always show the preference for ammonia over amino acids (14). Therefore, timing of DAP additions is important. A single large addition of DAP at the beginning may lead to an excessive fermentation rate and an imbalance in the uptake and usage of amino acids. To avoid this problem, multiple additions at 16° Brix and 10° Brix are preferred.
While addition of ammonium salts may not significantly benefit stationary phase yeast (15), the addition of specific amino acids may have a stimulatory effect and extend fermentative activity (16). Single amino acids may be quickly utilized to resynthesize transporter proteins that
are rapidly “turned over” during accelerated growth. Supplements added after about half the fermentation is completed may not be used by the yeast because alcohol prevents their uptake. For the same reason, adding nutrients to a stuck fermentation is seldom effective.
Juice/must can be vitamin deficient as well as deficient in assimilable nitrogen when there is a high incidence of microorganisms (mold, yeast and/or bacteria).
Yeast hulls are byproducts of the commercial manufacture of yeast extract. Consisting of cell walls and membranes, hulls are added to enhance fermentation rates and to restart stuck fermentations. Their mode of action has been described as lowering the concentration of inhibitory C8-10 fatty acids. Ingledew (3) reported that yeast hulls stimulate yeast populations by providing a source of C16 and C18 unsaturated fatty acids which act as oxygen substitutes under long-term fermentative conditions. Additionally, hulls may provide a source for some amino acids as well as surface area to facilitate release of potentially inhibitory levels of saturated CO2.
Yeast propagated aerobically contain a higher proportion of unsaturated fatty acids and up to three times the steroid level of anaerobic yeast. Without initial oxygen, replication is usually restricted to 4-5 generations as each yeast budding cycle reduces the sterol content of the membrane by approximately half. When the level reaches a critical point, replication stops and fermentation must continue with the population present at that point.
Something I learned asking around is that an unconventional distillery I think is probably among the most important in the country uses the NOPA test instead of the formal number. The data for them also justifies a sophisticated autosampler whose price may make you faint. We are going to increasingly find, fine spirits are born in the lab.
My formol starting pH: 4.2. I went dropwise with 1.0N NaOH and then finished with 0.1N NaOH to easily hit 8.2. It did not take long.
The first test I did was of a ferment that started at 25 brix, TA 13 g/L (1/3 acetic). Only a small amount of liquid ammonia was added for nitorgen. This also reused dunder that was likely high in nitrogen. I used a 25 ml sample size. I however did not use the method where the TA was calculated on a large volume and the formol number was calculated on a half sized volume using volumetric flask to cut the volume in half.
TA 15.92 (using 5 ml 1.0N as a starter to hit 12.01 g/L) The TA on heavy rums is quite substantial and uses a lot of 0.1N NaOH so I’ve started using a pretitration dose of 1.0N administered by micro pipet. For a 10ml analyte size, 1.0 ml of 1.0N titrant would be 6.005 g/L so for a 25 ml analyte size it takes 2.5 ml of titrant. You can double that when you are over known to be over 12 g/L. This may make working a lot faster.
When it came time to add the excess of formol, I added 5.0 ml and the pH dropped, then 5.0 more and the pH dropped, then 2.5 ml (to hit the recommended scaling) and the pH stopped dropping eluding to an excess. pH has settled at 6.498. For things like muck, which don’t have data yet, we may have to keep exploring what it takes to generate an excess based on pH drop.
Programming the formol number into the Hanna Instruments automatic titrator was easy and I had all their parameters on a sheet. It had me changing a few settings I did not understand which I think may relate to sensitivity of measurement, but it was easy to follow their program.
Now that I can measure this, I am curious what the different parts of the rum wash look like and how they transform. What do dunder infections do to nitrogen? What happens to muck? For cleaner ferments, how does this number related to escalating risk of infection? How do fission yeast needs differ from those of budding yeasts? How does this number correlate to the risk of ethyl carbamate?
Because there are so many parts to measure, it is going to take a lot of work to frame all the data and have patterns emerge.
(page 48 SOR) Arroyo’s ideal Nitrogen (mg /100 ml) 75.0 to 100.0 to 750 to 1000 in PPM (mg/L)
Ammonia hydroxide 1 ml/L or 0.1% 0.25g/L
Data from Alcohol Textbook sixth edition page 283 examining FAN in grain backset
Notice that FAN is not YAN so I don’t know how that would scale the numbers for a formol number by titration, but they do appear to be pretty close to 100. I believe they are also from fuel ethanol plants and not beverage ethanol which has different objectives.
Formol data test ferment
YAN before ammonia: F# 75.75 pH 6.2
Added 0.5 g/L Ammonia hydroxide as 10% solution pH 7.28 F# 199.69
Added second 0.5 g/L Ammonia hydroxide as 10% solution pH pH 8.14 F#286
I’m not quite sure what has gone wrong here. I’m gong to try again after I adjust the pH back to 5.0
I adjusted the pH down to 5.1 and got an F# of 290. This a little bit surprising so I need to learn more about this calculation relative to Arroyo 750-1000 note. The final formol number for this ferment a few days later was 63.
Calculator which will help with increasing YAN using DAP