Carbonation in Lambic

Carbonation forms one of the most important parts of the lambic-drinking experience. Unlike most other styles for which carbonation only exists in a narrow concentration band, lambics span the range from entirely still to some of the most carbonated beers brewed anywhere in the world[REF]. Further, unusual carbonation states, such as the "Loerik" or "Lazy" and "Doesjel" or "snoozer" guezes (examples from Cantillon, Lindemans, and two from 3 Fonteinen, 1 2) that fail to carbonate as normal, are often highly sought-after experiences, and some otherwise-still lambics may carbonate over long years of ageing. Carbonation may also decrease as CO2 can escape from a bottle, particularly over long timescales where the cork may become compromised.

Overview

Carbonation is formed by the dissolution of CO2 in a liquid. This carbonation may be the result of microbial action, or forced into the liquid from an external source. In lambic, only the former is traditionally responsible for carbonation in bottles, but the latter may be used, at least to some extent, to force the beer out of the keg or to augment the carbonation in the bottle.

The fermentation of sugars into ethanol releases CO2 and is the primary source of the gas in lambic. Thus, the production of alcohol correlates well with the production of CO2 and sacharomyces is the primary producer of carbonation in lambic with brettanomyces in a distant second [REF].

The Formation of Carbonation in Lambic

Carbonation in lambic, as in most beer, is primarily due to the fermentation of simple sugars by saccharomyces [REF]. For glucose, this reaction's overall form is:

C6H12O6 --> CO2 + H2O + stuff + energy

Saccharomyces dominates the yeast flora in lambic between XXX and YYY months[REF], during which time most carbonation is formed. Unblended lambic bottled after this time is generally still as the CO2 will have escaped prior to bottling. Lambic bottled younger than this, such as the use of jonge lambic as a blending component in Gueuze, will carbonate in the bottle.

The addition of simple sugars at the time of bottling, either in fresh wort or in priming sugar, has been shown to re-start fermentation by saccharomyces and lead to carbonation [REF]. This indicates that the end of sacc dominance in the lambic is due to a lack of sugars the yeast can metabolize and not due to inhibition from ethanol or other compounds.

Additionally, the brettanomyces may produce some carbonation, as its fermentation may produce ethanol through the same metabolic pathway as saccharomyces [REF]. Due to the presence of other metabolic pathways and the dearth of simple sugars by the time the yeast-niche in lambic has been ceded by sacc, brett is usually not implicated in the production of much carbonation in lambic.

The Concentration of Carbon Dioxide in Lambic and Other Beers

Measurements of carbonation can be reported in volumes of CO2 dissolved in the beer. By dividing both by the volume of the liquid, we arrive at a dimensionless number called "volumes of CO2". So if one liter of carbon dioxide at cellar temperature and pressure ("CTP", 55 F, 1 atm) is dissolved in one liter of lambic, we may say that this beer contains "one volume of CO2". As the molar volume of CO2 at CTP is 0.043 mol/l [NIST WEBBOOK], we can convert from "volumes of CO2" to molarity by multiplying the former by 0.043. Note that this measures the volume of CO2 applied, and thus the total carbon in the system irrespective of whether it's in the form of aqueous CO2, carbonic acid, or any of its deprotonations.

Labic ranges from still lambic with 0 volumes of CO2 (0 molar) to upwards of 5 volumes of CO2 (0.2 molar) in the case of some highly-carboned guezes [REF].

Measurements of carbonation in lambic are shown below along with other beers and carbonated beverages for comparison:

TABLE [REFS].

The Chemistry of Dissolved Carbon Dioxide

When dissolved in water, carbon dioxide forms carbonic acid with its solvent via:

CO2 + H2O -> H2CO3

at a ratio of about one H2CO3 molecule per 590 dissolved CO2 molecules [REF, CRC handbook?]. Carbonic acid has pKas of ~3.49 and ~10.32 [PINES 2016], which are defined as:

Ka_1 = [H+][HCO3-]/[H2CO3] Ka_2 = [H+][CO3 2-]/[HCO3-],

Where the brackets, [], indicate the molar concentration of the species they contain and

pKa = -log(Ka).

Thus, we can see that the first deprotonation completely dominates the acidity of carbonic acid at low pH, being seven orders of magnitude larger than the second, and we are justified in ignoring the second deprotonation's contribution henceforth. Combining the coefficient of hydration from above with the first deprotonation gives an overall equilibrium constant defined as:

K_total = 4.47*10^-7 = [H+][HCO3-]/([H2CO3] + [CO2 (aq)]) [REF].

Note that the denominator is the total concentration of CO2 and products to an approximation better than one part in two million. Combining this with the total carbon from above and re-arranging, we can write

[CO2 (aq)] = (4.3*10^-2(pH+1))V_CO2/((1+Ka_0)*10^-2pH + Ka_0 * Ka_1 * 10^-pH + Ka_0 * Ka_1 * Ka_2) [H2CO3] = Ka_0 * [CO2 (aq)] [HCO3 -] = Ka_1 * [H2CO3] * 10^-pH [CO3 2-] = Ka_2 * [HCO3 -] * 10^-pH

Where

Ka_0 = [H2CO3]/[CO2 (aq)],

which uses numbers a brewer is likely to know, pH and volumes of CO2 as independent variables.

As the pH (pH = -log[H+]) of lambic is lower than most other beers (~3.4 for lambic vs. ~4.5 for non-acidic beers [Embrace the Funk, etc.]), and the volume of CO2 can be comparatively larger (~5 volumes for some geuze vs. ~2.5 volumes for your average American Pale Lager), the chemical environment due to carbonation is markedly different:

N.B. these calculations were performed assuming a temperature of near enough to 25C to make no meaningful difference, and no account was taken for ionic strength, etc.

The effects of Carbonation on Mouthfeel and Physiology

The prickly mouthfeel of carbonation is partially attributed to the human body's pain sensing system, specifically those pain sensors that express the TRPA-1 protein, which responds to the local pH drop due to the presence of carbonation (as well as mustard and cinnamon) on the tongue with a sharp, tingling sensation. The mechanical sensation of a large amount of gas nucleating and coming out of solution quickly as it warms is responsible for the rest of the sensation (note that the solubility of gasses in liquids generally decreases with temperature) [WANG, CHANG, and LIMAN, 2010].

The human body maintains a blood pH in a very narrow range between 7.35 and 7.45 pH, and is very good at buffering itself. Dissolved CO2 in a beer has no significant effect on the body's pH [The Effects of Carbonated Beverages on Arterial Oxygen Saturation, Serum Hemoglobin Concentration and Maximal Oxygen Consumption].

Other Gasses in Lambic

In contrast to CO2, most other common gasses like nitrogen are much less soluble in water, leading to a much different drinking experience for nitro beers. Also, the helium beer thing is a prank. Helium isn't very soluble in water and most of it would be drunk, not inhaled.

Oxygen has a reasonable solubility in water (https://water.usgs.gov/owq/FieldManual/Chapter6/table6.2_6.pdf), and its presence or lack thereof is very important to the ecosystem in the bottle and the final lambic, as well as the slow oxidation of the product as it ages. These effects are outside the scope of this article.

Unusual Carbonation States of Lambic

Sometimes a lambic will fail to carbonate as expected [LINKS to Loeriks, Doesjels], which are usually reffered to as being some form of "lazy" or "snoozing" in Flemish. Many of these lazy lambics are highly sought-after due both to their rarity and the quality of some batches. They may eventually carbonate slowly over many months or years.

Because lazy lambic is a rare and unpredictable event, there is no research and little evidence of why they occur, though we can speculate as to the general cause. As most carbonation in lambic is due to the action of saccharomyces, we can surmise that this group of yeasts is inhibited in lazy lambic for some reason. It may be due to differing temperature conditions early in fermentation, strain variation, outcompetition or inhibition by another organism due to inoculation variation and environmental differences, or inhibition or lack of simple sugars due to batch variation in starting ingredients. It may also be the result of extremely rapid saccharomyces fermentation in the youngest lambic prior to the introduction to the bottle, leaving little sacc-fermentable sugars to induce carbonation.

Because brettanomyces may produce carbonation, but has much slower growth rates than saccharomyces, it is likely that the slow months-to-years carbonation that builds in lazy lambic is due to the action of brett which may work on similar time frames.

It is possible, though less likely, that more complex sugars are being slowly broken down into more simple sugars which saccharomyces may then consume. Some bacteria may excrete simple sugars in the presence of more complex sugars, or may simply synthesize them as a storage compound which they release upon lysing[REF]. Enzymes released by brett or other organisms during lysis may also break down complex sugars [REF]. As priming sugar may re-start saccharomyces fermentation in the bottle [REF], it would stand to reason that the slow re-introduction of simple sugars either by their synthesis or by the cleaving of sugar monomers off of higher-order carbohydrates would lead to carbonation by saccharomyces fermentation with the former cleaving step being rate-limiting.

In some rare cases, intentionally-still lambic has been reported to have carbonated after years of ageing [REF]. Reasons for this are also unknown, but are likely similar to how lazy lambic carbonates over long time frames.

Pines, D.; et al. How Acidic is Carbonic Acid? J. Phys. Chem. B, 2016, 120 (9), pp 2440–2451 http://pubs.acs.org/doi/abs/10.1021/acs.jpcb.5b12428?journalCode=jpcbfk

WANG, CHANG, and LIMAN, 2010. TRPA1 Is a Component of the Nociceptive Response to CO2. The Journal of Neuroscience, September 29, 2010. 30(39):12958 –12963 http://dornsife.usc.edu/assets/sites/381/docs/documents/WangChangandLimanCO2sensingJN.pdf

Waibler, M., 1991. The Effects of Carbonated Beverages on Arterial Oxygen Saturation, Serum Hemoglobin Concentration and Maximal Oxygen Consumption. Master's thesis, Oregon State University. https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=2&cad=rja&uact=8&ved=0ahUKEwimp92O5oPYAhWD5IMKHa6LCb8QFggyMAE&url=https%3A%2F%2Fir.library.oregonstate.edu%2Fdownloads%2Ftt44pp885&usg=AOvVaw1Hx-mDg9llHl_QzQeeWZcf