AGING OF RUMS
Rums that have just been made have a burning flavor, due to the high alcoholic degree; a poorly developed, coarse, often unpleasant, repulsive bouquet even in the case of certain grand arôme products. They leave the palate feeling dry and of more or less marked bitterness. When the manufacture has left to be desired, they also have defective tastes: tastes of the boiler, putrid odors, are pungent, etc.
Despite their shortcomings, fresh rums are sometimes delivered directly to the consumer, after a simple dilution to the marketable level (45-50 °). This is particularly the case for certain rums of vesou or syrup (grappe blanche of the French West Indies, clairin of Haiti) and some very light molasses rums (rums of Cuba).
Most often rums undergo a prior aging. During this, bad tastes fade or disappear. The primitive bouquet is modified, increases in power and finesse; the flavor becomes softer and more mellow. This “improvement”, which greatly increases the value of the product, is unfortunately accompanied by significant losses in alcohol. In order to avoid these, as well as the cost of warehousing, which makes it very expensive to preserve spirits in barrels, attempts have been made to artificially provoke the reactions which occur over time. Rapid aging processes, which are widely used in some countries, particularly in the United States, never give as good a result as natural aging.
Physical-chemical phenomena of aging
During the conservation of spirits in barrels, on the one hand, very apparent physical phenomena occur: reduction of the volume, modification of the alcoholic degree, dissolution of the tannic principles of the wood; and on the other hand, chemical phenomena occur: concentration of secondary products, oxidation, esterifications, polymerizations, etc.
[Hopefully we will learn more about polymerizations.]
If spirits are placed in well-sealed glass containers, the nature of the phenomena is quite different. The volume of the liquid remains unchanged and the alcoholic degree sees little or no change. The oxidation actions are reduced, but the esterification continues normally. At the same time, there is a slight attack of the basic silicates of the glass, which increases the ash content and reduces that of the acids and esters of the eau-de-vie (Schidrowitz and Kaye). When the containers are imperfectly porous, the oxidation continues as in the wood, but it is more violent and less sparing.
Reduction of volume.
In wooden barrels, there is a continuous reduction in the volume of the liquid, resulting from the slow evaporation that occurs through the walls and the incomplete closure of the bung. The importance of loss depends on many factors, some internal: nature, dimensions and fill level of the barrels; the others external: temperature, aeration, humidity of the local.
Losses are stronger in smaller rather than larger casks. This is because the ratio of the surface area to the volume of the liquid is higher in the first (1), and also that large containers have thicker staves, making it easier to pass liquids through the pores of the wood. The losses are about the same whether charred, new or used. Loses diminish when the containers are internally waxed or varnished externally, the coatings counteracting the osmosis of the liquids. Loses are even higher if the emptying state of the container is more advanced. As a practical consequence, it is advisable to fill the casks as the volume of the liquid decreases, by using eau-de-vie of the same age. It will however be important to leave a vacuum of a few liters, to allow the expansion of the liquid.
[My understanding is that waxing was only ever common to grappa and possibly only the type used for fortified wine.]
(1) The surface area, or ratio of surface area to volume (in liters), is 76 cmq for a barrel of 550 liters, 98 cmg for a barrel of 270 liters, 184 cmq for a barrel of 50 liters, and 245 cmq for a barrel of 25 liters.
The factors that influence the intensity of evaporation: temperature, humidity, ventilation, also affect the loss of alcohol. This will be maxima, all conditions being equal, in a warm, dry and well ventilated spaces, and minima in a cold and humid atmosphere. If the barrels are placed in a medium where the air, saturated with moisture, is always in motion, the eaux-de-vie can even, while losing degree, increase in volume (Mathieu).
[Remind me to find Mathieus paper.]
Finally, there may be abnormal losses by leaking, due to the presence of slits, produced during the assembly of barrels or from the “work” of wood. Some xylophagous insects also dig galleries in the wood of the barrels, when it has some moisture, and can cause significant losses by leaking. It is therefore important to arrange the barrels in the aging warehouses, so that it is possible to visit them easily and to quench leaks quickly.
Due to the variety of factors involved, the losses experienced during the conservation of eaux-de-vie are highly variable. In France, for spirits housed in barrels of 500-600 liters and placed in cellars subject to variations in ambient temperature and humidity, an average volume reduction of 25 to 30% is allowed after 10 to 15 years. In some cases, however, this may be as high as 50%.
In the United States, where barrels of 180-195 liters are used and stores are frequently heated during the winter, the losses are more susceptible. Crampton and Tolman found that “rye whiskeys” placed in heated cellars (average temperatures: 20-30° in winter, 26-30° in summer), at the end of 8 years, the volume was reduced by an average of 46%, but for some samples it reached 62%. In the case of “Bourbon whiskeys”, kept in rooms subject to variations in ambient temperature (4-12° in winter, 26-30° in summer), the loss was only 35%. According to the authors above, the volume reduction would be done in a very regular manner and would be substantially proportional to the number of years, except in the first year, when it is relatively larger. The following variations in volume and alcohol content were observed, for example, in the case of two samples:
Valaer observed, in the case of the rums placed in heated warehouses, losses varying after 2 years between 10 and 22% of the original volume, according to the conditions of storage (temperature, humidity, etc).
In Martinique, the Régie tolerates, in the justification of volume reduction, 10% on average per year, in view of the accidental leakages which, due to defective storage conditions, sometimes reach a very high figure. In good storage conditions, the losses appear to be on average 10% the first year and 25-30% after 5 years.
[Safe to assume “Régie” is their IRS/TTB]
Modification of the alcoholic degree.
It is admitted in France and in the French colonies that the degree of spirits preserved in wood is constantly decreasing. The opposite phenomenon can, however, occur, as shown, for example, by the following experiment carried out by Bouffard (1).
(1) Ann. Brass. Dist. 314, 1912.
This author has placed small oak barrels, of 10 liters capacity, filled with alcohols of varying degrees, in three different premises: on a terrace protected from the direct action of the sun by a simple board; in a closed attic, sensitive by its roof to atmospheric variations; finally in an underground cellar saturated with humidity and low temperature (12-14°). Each series included 9 containers. After 14 months, the following findings were made.
The casks kept on the terrace showed an increase in the alcoholic degree: this was maximum (Δ6°8) for the low-strengh alcohols (18 to 25°), then it decreased to become almost zero (Δ0°2) with the richest alcohol (83°7). The weight loss of the liquid varied in the same direction: maxima for low alcohol, it diminished in the opposite direction of the richest in alcohol. The absolute loss of alcohol by weight on the contrary increased with the initial richness in alcohol.
In the case of casks housed in the attic, the losses in weight of liquid and alcohol were of the same order as for the previous series. As for the alcoholic strength, the maximum increase was a little lower, and the strongest alcohol already had a decrease of Δ0°5. [I added the deltas to make it clearer and I’m not sure if he means 0°5 as a range of 0-5 degrees or 0.5 degrees. I’m sure the original paper may clarify it.]
For barrels stored in the cellar, the decrease in alcoholic strength was general: slight in the case of weak strengths, it increased with the concentration of alcohols to reach the maximum (17°9) with alcohol at 83°1. The losses of weight of the liquid were the most considerable with the weak alcohol and then diminished, to become null in the barrel at 44°; Beyond this measure, there was an increase in weight, increasing with the initial alcoholic strength.
Bouffard explains the modifications of the alcoholic degree by the dialysis phenomena combined with that of evaporation on the surface. Porous wood acts as a dialysing membrane, which would allow water to pass more easily than alcohol. In a hot and dry atmosphere, water and alcohol evaporate in the same proportion and, as the wood continues to dialyze, the alcoholic strength increases. On the contrary, when the air is saturated with moisture and devoid of alcoholic vapors, the alcohol continues to evaporate easily, but water, in the presence of a maximum vapor pressure, undergoes only low or no evaporation: the alcoholic degree is lowered.
[When you think about the surface evaporation a barrel stored on its side versus upright may be quite different. This could have a practical bearing on ethyl acetate that some Jamaican style rums are trying to shed.]
Sauvaire (2) formulates a slightly different hypothesis. It would occur, during the storage in barrels, a double evaporation: one, all the more slowly by the free surface of the liquid as the receptacle is hermetically closed; the other by the outer surface of the barrel. The first, which obeys the ordinary physical laws, results in a loss of alcohol and a diminution of the degree. The second is complicated by a sort of dialysis, in which the alcohol, which behaves like a colloid, is retained on the side of the inner surface of the barrel wall, while the water is absorbed and would evaporate on the outer surface, this all the faster as the atmosphere is hot and dry. According to the respective intensity of the two phenomena, the alcoholic degree will remain constant or will be modified in one direction or the other.
(2). Bull. Soc. Pharm. Bordeaux LXXIII, 111, 1935.
To verify this hypothesis, Sauvaire placed in two barrels made of the same wood and incompletely filled, a brandy at 56°. After 4 years of storage in the same space, he observed an increase of 4° in one and 6° in the other. One of the barrels was then varnished externally, with the exception of the vicinity of the bung. At the end of four years, the degree was reduced in the varnished drum, while it had continued to increase in the other.
In the ordinary conditions of aging in France, where the humidity of the cellars is relatively high, reduction of the alcoholic degree is the general rule. The Régie tends to accuse the owner or the merchant of fraud, when the alcoholic degree increases, and to hit them with rather high fines. Rocques (1) indicates the measure listed below in two examples from the Charentes:
(1) J. Pharm. V, 177, 1897.
According to Lapparent (2), the measure would fall in the Charentes from 6 to 8° on average after 15 to 20 years. The first year, under the influence of a new barrel, the loss would be 2 to 3°.
(2) Le vin et l’eau-de-vie de vin. Paris
According to Lührig (3), there may be an apparent loss of alcohol, due to the increase in the specific gravity of the liquid, as a result of the extraction of the soluble principles of wood. In new barrels, the extraction of the water contained in the pores of the wood determines a degree decrease of 0.1 to 0.15. When the receptacles are incompletely filled and not stirred, the distribution of alcohol is inconsistent, and there are sometimes appreciable differences in alcoholic richness in the different parts of the barrel, the upper layers being the richest in alcohol. Thus, the author found the proportions of alcohol hereafter in the upper, middle and lower parts of a 4-meter-tall barrel, after 7 weeks of storage: 40.70, 37.70 and 36.33% by volume. The differences diminish over time, especially when the liquid is subjected to temperature variations: a barrel of 250 liters which suffered a volume loss of 20 liters after 33 months, showed in its upper, middle and lower parts respectively : 40.44, 40.35 and 40.06% alcohol by volume.
(3) Pharm. Zentralhalle LXVII. 49. 1926.
[The first think Kervegant is talking about is obscuration due to soluble barrel solids. Those last numbers are interesting because it is so challenging to measure ethanol concentration with that much exactatude before the era of the u-tube densitometer.]
In the tropics, where the humidity of the air is often close to saturation point for a good part of the year, the lowering of the alcoholic degree during storage is more accentuated than in temperate regions.
In Martinique, for example, an average loss of 2-3° is expected in the first year and 1° each of the following years, in the case of Rhums initially measuring 60-65° and stored in 250-liter barrels. The figure is however very variable depending on the humidity of the premises: so we found a brandy, put in barrel, according to the declarations of the owner, at a measure close to 70° and whose degree had fallen after 8 years to 48.2. (4).
(4) Arroyo suggests that rums should be kept in an elevated, hygrometric atmosphere over the period necessary to remove unwanted constituents of non-alcohol, and then reduce the moisture content to prevent excessive losses of alcohol.
[Absolutely wild numbers.]
It is quite different in the United States, where the eaux-de-vie subjected to aging have a lower alcoholic strength (generally 100 degrees proof or 50 ° G. L.). In this country, the elevation of the degree is the rule. Crampton and Tolman have observed that after 8 years of aging in 180-190 liter charred oak barrels, different batches of 50-51° initially strength whiskey have experienced average increases of 11°9 for “rye whiskeys” stored in heated warehouses and 5°5 for “Bourbon whiskeys” in unheated warehouses, with extremes of 0° and 19°5 G. L. Valaer noted, for rums kept under similar conditions, an increase in alcoholic strength of 1 to 2° annually after 2 years.
But in all cases, regardless of the variation of the alcoholic strength, during aging, due to the decrease in volume, a loss of alcohol occurs.
In the ordinary conditions in which the eaux-de-vie of the Charentes region are placed (housing in barrels of 500-600 liters, average temperature of the cellars 15 ° C, the following annual alcohol losses are allowed, according to Rocques:
500 liters of eau-de-vie put in barrels at 70° become 25 years later 350 liters at 50°.
In the United States, the losses for the first year reach 7 to 10% of the volume of the original alcohol and the following years 2 to 5%. Crampton and Tolman found that after 8 years, the overall loss averaged 34% for “Rye whiskeys” and 28% for “Bourbon whiskeys”. Valaer, in the case of rums, noted the average losses below:
[Its a little bit hard to follow in this conversation if we are talking about liquid volume or absolute alcohol.]
The secondary products of eaux-de-vie (higher alcohols, acids, aldehydes, furfurol, esters) do not appear to participate in the phenomena of osmosis and evaporation which affect ethyl alcohol and water.
[This is an extremely critical point I have never seen acknowledged yet runs counter to Arroyo’s notion of high humidity removing the volatile undesirable congeners.]
The quantity of higher alcohols, relative to the initial volume of the eau-de-vie, hardly varies during the storage in barrels. It is about the same for acids, aldehydes, furfurol and esters, at least after a few years of aging. The increase in the relative proportions of these impurities over time then seems to result from the reduction of the quantities of water and alcohol contained in the barrel, that is to say, the concentration of the eau-de-vie. The wall of the barrels would exert a selective action towards the various elements and would oppose the passage of higher alcohols, acids, aldehydes and esters (Crampton and Tolman).
[He is still not spelling out if this includes ethyl-acetate but that is starting to look likely.]
The non-variation of the absolute quantity of the impurities after a few years could probably also be explained by a simultaneous production and evaporation of these products. This explanation is, however, less satisfactory, especially in the case of higher alcohols, which could only be produced during aging by the hydrolysis of the esters.
[Ethyl-acetate could be produced and evaporated simultaneously still resulting in a typical net increase.]
Finally, it should be noted that the intensity of the chemical phenomena of aging, especially the formation of acids and esters, seems to be directly related to the quantities of water and alcohol passing through the walls of the barrels. The more these quantities are important, and the more the eau-de-vie ages rapidly.
[I don’t think this last line is translating too well. My understanding is that the lower your pH starts, via possessing noble acidity, the faster aging reactions will start. Otherwise you have to wait for wood to breakdown to drop your pH.]
Phenomena of dissolution.
During the stay in casks, the eau-de-vie extracts from the wood soluble principles, which color and give it body while improving the bouquet. However, if these principles are too significant, they communicate to the alcohol an unpleasant taste, called “boisé”.
[boisé translates as woody.]
According to Faure, the main substances yielded by oak wood are quercitic acid, quercitrin and quercine. Quercanic acid, or tannin of the oak, appears in the form of a reddish powder, astringent, giving by hydrolysis of pyrocase. According to Bottinger, the tannin of the bark of oak would have as formula C19H15O10, and that of the wood C16H12O9. Quercitrin, also called quercitric acid, C12H22O12, is a gold-colored dyestuff, giving by decomposition under the action of the dilute acids of quercelin and rhamnose. Finally, quercin or quercinite is a hexatomic alcohol, of formula C6H6(OH)6, endowed with a particular aroma.
[I’m going to leave this paragraph very speculative because I think a lot of these names have changed since then. Not sure if these molecule counts are even accurate. Quercitrin appears to be C21H20O11]
According to Reif (1), among the products extracted from the wood of barrels, small quantities of vanillin and other substances giving similar colored reactions are found. This author found in samples of wine brandy that he examined doses of 0.1 and 0.2 mgr of vanillin per liter of liquid. Haüssler (2) had previously reported the presence in oak of a substance giving the same reactions as vanilla, and also Reinke its presence in the spirits stored in casks.
(1) Z. Unters. Lebensm. LIV, 90, 1927.
(2) Z. Offentl. Chem. XX, 184, 1914.
The furfurol contained in the spirits also comes partly from the barrels, especially when these were previously charred, Crampton and Tolman observed that “Rye whiskeys” and “Bourbon whiskeys” originally containing respectively 1 mgr and 0.7 mgr of furfurol per liter of eau-de-vie at 50°, after 8 years of storage in charred barrels 3.4 and 2.1 mgr of furfurol. Compared to the original volume and alcoholic strength, the increases were 1.23 and 1.09 mgr of furfurol per liter of whiskey. In some cases, the increase may be much larger. Valaer reports two rum samples which originally measured 5.5 and 6.3 mgr and, after 24 months of storage in barrels, 12 mgr of furfurol per liter of eau-de-vie. According to this last author, the increase of furfurol would be especially sensitive during the first months of contact with wood; after 6 months, there would no longer be any increase in the level of this aldehyde, at least in absolute value.
The nature and quantity of the dissolved substances depends on the quality of the wood, special preparations suffered by the casks, the duration of the contact, etc.
Oak wood is very commonly used for the construction of barrels for housing eaux-de-vie, because of its hardness, its low porosity and the extractive principles it contains. Chestnut, which contains a tannin of coarse flavor is too heavy in soluble principles, is to be avoided. To store white spirits (kirsch, quetsch, etc.), ash barrels are sometimes used. These, especially when they have been parboiled and coated internally with a layer of melted wax, yield very few extractives to the eau-de-vie, and preserve its original color.
Oak wood itself has a highly variable composition of soluble products, depending on its origin (botanical species, climate and soil conditions of the producing region), the age and the part of the tree from which the the staves were drawn, the time during which the wood was kept in storage, etc.
The Danzig staves, for example, give the brandies a slight color, those of America an amber color, those of Riga a dark color, and those of Bosnia an almost black color.
The most valued wood in the Charentes for aging cognac, is the Limousin oak, from the forests of Limousin and Allier. White oak from North America, considered in France as a secondary quality, is almost the only one used for the manufacture of barrels used in storing rums in the West Indies and the United States.
Staves intended for the preparation of barrel staves must come from oak heart wood. The sapwood, too rich in soluble organic substances, is to be rejected, as well as wood too porous or too colored. “The staves,” writes Jacquet (1), “should come as much as possible from trees 40 to 50 years old at least, be cut in the direction of the wood grain, split with an ax and not with a saw; parts devoid of sapwood and knots, without rot or worm, the grain will be fine and tight, the coloring quite clear and the fibers crossed by veins clear and shiny; the staves must take the hair under the scraper; finally, they will only be made into barrels after drying in a stack for at least 5 years (2).
(1) La fabrication des eaux-de-vie, Paris 1895.
(2) Barrels of green wood should never be used, which not only delay maturation, but also impart an unpleasant taste to spirits.
[This paragraph actually uses three different words for stave. So Eskimo.]
The preparation suffered by the casks has a great influence on the quality and quantity of the extract yielded to the eaux-de-vie.
Crampton and Tolman highlighted the influence of barrel charring. They kept the same whiskey in two barrels of oak, one charred and the other uncharred, subject to identical conditions for a period of 8 years. At the end of this time, they analyzed the product and obtained the following results:
From the point of view of chemical composition, the two whiskeys were therefore very comparable, except for dry matter and coloring. From an organoleptic point of view, there were considerable differences. The uncharred barrel whiskey was entirely devoid of the flavor of American whiskeys and was more like “Scotch” or “Irish whiskey”. The other had a strong, characteristic aromatic bouquet, an oily appearance, and a more persistent and greasy foam; in addition to tasting, it was leaving the impression of being very old. The dyestuffs and extractives of the spirits having passed through an uncharred barrel were much more soluble in water and less resinous than those of the whiskey stored in charred barrels. The dry extract taken up by a few cubic centimeters of water possessed in the case of the latter a very strong odor and the particular aroma of the American whiskey; its flavor was also very accentuated, both resinous and aromatic. The uncharred cask whiskey had on the contrary a slightly acidic and astringent flavor, very different from that of the other product. The authors therefore consider that the differences between American whiskeys and English whiskeys come largely from the packaging used.
[The “greasy foam” park I believe refers to shaking the liquids and implies surface tension. The old spirits scientists were aware of Traub’s rule that long carbon chain molecules significantly reduce surface tension. Many rum scientists employed a surface tension test.]
According to Valaer, the chemical modifications are more pronounced in charred barrels. The increase in dry matter, acids and coloring in particular would be faster and more significant than in uncharred barrels.
New barrels give a pronounced woody taste to brandy that stays for a long time. Also, in the case of fine eau-de-vie, we take care of briefly storing the spirits in new barrels for only a few months, before transferring them to used containers. This is rarely practiced with rum casks, but in order to reduce the coloring and to soften the taste of wood, containers which have already been used are often used for aging.
[I think in this era rum maturation was really finding its way and there weren’t the uniform practices we see today.]
The increase of color, rate of the dry extract, acids and probably other impurities in the eau-de-vie are weaker in used casks. They are also more irregular and depend on the number of times and the length of time the containers have been used. Where these have previously contained dessert wines, sherry, etc., the extract may contain special aromatic substances deposited by the liquids which have previously been there. In paraffinized casks, there is only a slight coloration during aging and the dry extract only increases in small proportions.
The duration of storage obviously affects the quantity of dissolved solids. According to the observations of Crampton and Tolman, the dry matter, as well as other impurities in the eaux-de-vie, increases especially at the beginning of barrel storage: after the third year, the increase would be due almost entirely to the concentration of the liquid, at least under the conditions of aging spirits in the United States. It is the same for dyestuffs: there is a close parallelism between the curve of coloring and that of extractive materials. The above authors noted the following average variations in whiskeys kept in charred oak barrels.
Valaer found that after 2 years of storage the rate of the dry extract of rums initially having 48-52% of alcohol in volume, reached 100 to 200 gr per hl of eau-de-vie. We ourselves observed that rums, put in new charred oak barrels at 70-72°, measured, after 4 years and 8 years of aging in Martinique, 390 gr. and 1040 gr. dry extract per hectolitre.
Spirits stored in glass attack the alkaline silicates of the containers and can dissolve appreciable amounts of mineral matter. Strunk found in rums-kept for 30 years in bottles, up to 35 gr. of mineral matter per hectoliter, and Valaer, in cognac spirits, up to 28 grammes per hectolitre. However the amount of ash is usually much lower. Minerals extracted from dark glasses have a brown color and are rich in iron: they can influence the color of the eau-de-vie.
At the same time as fixed acids are extracted from the wood (tannic acids, etc.), the alcohol oxidizes little by little during storage, giving rise to aldehydes, then to volatile acids (mainly acetic acid).
Trillat (1), for example, has reported the oxidation of cold ethyl alcohol in a host of cases, with production of aldehyde acetic acid.
(1) Oxydation des alcools. Paris, 1901.
Duchemin and Dourlen (2) showed that rectified alcohol could slowly oxidize on contact with air until acetic acid appeared. They found that after neutralization of an alcohol with sodium hydroxide, the acidity then increased to a maximum, which varies with the alcohols and the impurities they contain. The nature of the containers also has a great influence (catalytic and saturation actions). Previous authors noted, for example, the following variations in a 90° alcohol sample, having as primary 0.014 acidity, aldehydes: traces, esters: 0.052, placed in different containers:
(2) C. R. OXL, 1466, 1905.
[I think there is a typo when he said neutralize alcohol with sodium hydroxide. I think he means the acid produced by the oxidation of the alcohol.]
The oxidation phenomena are favored by the porosity of wood, which plays a role comparable to that of platinum foam: the alcoholic liquid absorbed by the wood is in a finely divided state, which increases its affinity for oxygen. They are even more active as the contact surface of the liquid with the air is greater (that is to say that the barrels are smaller) and the external temperature is higher.
During aging, there is a continuous increase of acids, both apparent and real. In other words, not only the quantity of acids added to the ethyl alcohol contained in the eau-de-vie, but also that of the acids contained in the whole cask, increases with time. The increase is felt especially during the first years: after 3 or 4 years, there would be no appreciable increase in absolute value, but only apparent increase due to the concentration of spirits (Crampton and Tolman).
The increase relates to both fixed acidity and volatile acidity. These are affected very differently depending on the nature of the container used and the mode of aging. However, there is a tendency for the proportion of volatile acids to decrease with age compared to the fixed acids. Thus, Rocques (1) found in the spirits of Charentes the following quantities of fixed acids and volatile acids (expressed in gr per hectolitre of alcohol at 100°):
(1) Ann, Chim. Anal. II, 308, 1897.
Lusson (2) gave the name of oxidation coefficient to the proportion of acids and aldehydes contained in 100 parts of total impurities. He found that this coefficient rises as the spirits age, without however being proportional to the age of the spirits. He obtained, for a series of eaux-de-vie of various ages, the following figures:
(2) Mon. Sc. Quesneville XLVII, 785, 1896.
Lusson thought he could rely on the oxidation coefficient to assess the age of spirits. But it is not possible to compare, as Rocques pointed out, (1) that for eaux-de-vie of an identical nature, the rate of acids varies according to many factors: acidity of the fermented must, method of distillation, aging method, etc. With regard to distillation for example, the more exhaustion is pushed, the higher the content of the acid distillate.
(1) Ann, Chim, Anal. II, 84, 1897.
[The further you distill into the tales, the more you accumulate steam volatile acids. The birectifier reveals these lessons quite clearly.]
The aldehydes undergo a slight increase during the first years of storage. However, there may be a decrease in their absolute amount after some time, especially when the oxidation phenomena are energetic (storage conditions in the United States). The variations in the rate of these impurities are, moreover, always low.
Hewitt (2) attributes the greenness of freshly distilled whiskeys to the presence of aldehydes and considers that the improvement of the bouquet during aging is due to the disappearance of these substances by oxidation or evaporation. He reports that in a few minutes he has been able to remove the greenness of young eaux-de-vie by treating them with a solution of sodium phenylhydrazine sulphonate, a reagent that affects only the aldehydes.
(2) J. Soc. Chem. Ind. XXI, 96, 1902.
Finally, several authors consider that the oxidation of tannins plays a key role in the improvement of spirits. “The principal phenomenon of aging,” writes Pacottet, “is a transformation of the tannins of wood, dissolved in the eau-de-vie. In the case of wine, the tannins are oxidized and give fragrant products. In eaux-de-vie, there is a similar phenomenon: tannins and neighboring bodies oxidize slowly and give aromatic products which, together with the ethers, constitute the bouquet of eaux-de-vie “.
The laws of chemical esterification have been established by Berthelot and Péan de Saint-Gilles (3).
(3) Ann. Chim. Phys. (3). LXV, 335, 1862; LXVI, 5, III, 1862; LXVIII, 225, 1863.
These chemists have shown that the combination of acid and alcohol is never complete and that it is limited by the opposite phenomenon: the saponification of the ester, with elimination of water. In the general case where there is a mixture in any proportions of one or more acids, alcohols, esters and water, the limit of the reaction depends almost exclusively on the ratios existing in the mixture between the number equivalents of these various bodies; it is almost independent of their nature.
When one equivalent of acids is mixed with several equivalents of alcohol, the amount of ester formed increases approximately in proportion to the number of equivalents of alcohol added. In mixtures containing one equivalent of alcohol and several equivalents of acid, ester production is approximately proportional to the amount of acid. The presence of an excess of ester or water on the contrary decreases the esterified quantities. The decrease in the amount of ester formed, when mixing one equivalent of acid and one equivalent of alcohol with increasing amounts of water, varies much more slowly than the amounts of water. In a system where the proportion of alcohol and water is constant and water dominates, the acid being in proportion less than the alcohol, the quantities of acid esterified will be proportional to the total quantities of acid.
Esterification takes place slowly, with a speed depending on the various circumstances of the system in which it is located. One of the most important factors is the temperature: the elevation of it significantly accelerates the phenomenon. The speed also depends on the nature of the acid: it is much greater for mineral acids than for organic acids, which combine more slowly with alcohol if their molecular weight is higher. Finally, it is influenced by the relative proportions of the reactants: it is accelerated by the presence of an excess of acid or alcohol. In the presence of a strong mineral acid, the esterification of the alcohol with an organic acid takes place much more rapidly (1).
(1) This phenomenon is explained today by the action of pH on esterification.
Berthelot expressed the results relating to a system initially formed of an equivalent of acid and an equivalent of alcohol, by means of the formula:
where y is the proportion (equivalent fraction) of acid or alcohol combined after time x, and 1 is the limit value of y. 1 — y is then the proportion of acid or free alcohol. K is a constant, depending on the nature of the liquid and the temperature of the experiment.
By integrating the differential equation and determining the constant, by the condition that the combined quantity is zero at the beginning of the experiment, we find:
equation which represents an equilateral hyperbole related to axes parallel to its asymptotes.
The action of water on esters is exactly the same as that of an acid on the alcohol. Increasing amounts of water decompose increasing amounts of esters, but can not cause complete decomposition. The limit is almost independent of the temperature and the special nature of the acid, but the speed of action is variable, as in the case of esterification. The influence of water is diminished by the presence of an excess of alcohol or an excess of acid.
In the special case of eaux-de-vie, which contains only very small proportions of acids, Berthelot (2) has found that the quantity of acids which are esterified is an almost constant fraction of the total quantity, and depends only on the relationship between alcohol and water. The following limit values are indicated, to which the ratio of esterified acid to total acid tends, when the weight of the acid decreases indefinitely:
(2) Chimie végétale et agricole IV, 360, Paris, 1899.
[I suspect that first alcohol numbers should be 100 and not 10. This is fascinating implies there is a direction relationship between volatile acids in a spirit and esters at a given proof.]
To confine oneself to the mixtures of alcohol and water among which spirits are understood, one obtains, with a sufficient approximation, the proportion of acid esterifiable, adding 1/10 to the proportion of alcohol in the mixture. If x is the quantity of alcohol % by weight, and that of esterifiable acid (relative to total acid), we will have the very simple relation:
Thus, in a brandy containing 60% alcohol by weight, the amount of combined acid will be 66% and that of the free acid 36%; in 45% brandy, the first will be 55% and the second 45%.
[Does this mean if we measure the free acids in a spirit (and the ABW), we can therefore assume the esters or is that not true because the spirit is likely not at equilibrium?]
If, in a recently distilled spirit, the proportion between the esters and the free acids corresponds to the limit of esterification, there will be no modification as regards the esters, as long as the elements of the mixture will not undergo variations. But if, on the contrary, the acids or esters predominate, there will be progressive formation or decomposition of the esters, so as to bring the system back to its steady state of equilibrium.
[This is fascinating. A freshly distilled spirit could have more esters than equilibrium can support because of reactive distillation. Most all spirits are also diluted later. This may mean, like I’ve hypothesized elsewhere, that many esters at the point of distillation are temporary place holders. These temporary esters either bring more noble volatile acidity into the spirit (making these acids more volatile) or push fusel oil out of the way (reducing its volatility).]
Ribereau-Gayon and Peynaud (1) showed that the theoretical limit of esterification could only be reached with difficulty and that the speed of the phenomenon was strongly influenced by the pH of the medium. Solutions of different acids, buffered at various pHs, containing 70 to 80 milliequivalents of free acid per liter and 10% of ethyl alcohol by volume, were placed in sealed tubes in a 100° heated oven. We give below some of the results obtained:
(1) Bull. Soc. Chim. (5), III, 2325, 1936.
The limit esterification coefficient, calculated according to the mass action law formula taking K = 4, is 12.2% and according to the Berthelot empirical formula of 10.8%. It was not affected by any of the acids. The final figure is all the closer to the limit, as the rate of esterification has been greater. The reaction becomes very slow as it approaches the limit. The above authors found that in wines the esterification limit was not reached for either the fixed acids or the volatile acids, even after 36 years of storage.
The same is true of spirits, as shown by the following analyzes of very old eaux-de-vie.
(2) Formule de Berthelot.
These figures show that, except in the case of spirits preserved in bottles (samples 4 to 9) the proportion of esterified acids remains relatively far from the limit coefficient.
It is because during the storage in barrels, the phenomena of esterification are strongly influenced by the reaction of the substances ceded to the liquid by the wood, by the oxidation which can be exerted on these last substances or on the alcohol itself, finally by the endosmosis that modifies the relationship between water and alcohol.
These influences are especially felt during the first years of aging. Crampton and Tolman have observed in the case of whiskeys, whose ratio of Acids: Esters is less than 1, that the acids are first formed in greater quantity than the esters, but that later they are produced more rapidly. After the fourth year, an equilibrium of 1:1 is established between the esters and the acids, which no longer varies appreciably. This equilibrium would correspond approximately, according to the analyzes given by the authors above, to the transformation into esters of 70-75% of the esterifiable acids, a stage which would then be difficult to exceed, even after 8 years of barrel aging.
Valaer and Frazier (1), however, could not confirm Crampton’s conclusions. They observed that the acids exceeded the esters after 6 months and that the difference between these substances tends to increase rather than decrease thereafter. In the very old eaux-de-vie preserved in wood (samples 14 and 15 of the previous table), Lusson and Rocques also found a strong predominance of free acids.
Crampton and Tolman observed for the whiskeys an increase in the actual amount of the esters during barrel storage. Valaer made similar findings for whiskey, rum and wine spirits in the United States. On the other hand, Rocques has noted that, in the aging of cognac eaux-de-vie, during the first years there is a relative increase in esters, as a result of the concentration, it actually is a decrease in their actual amount, a decrease that can be very noticeable in very old eaux-de-vie. These divergences seem to be due on the one hand to the composition of the spirits examined, and on the other hand to the applied aging method. In France, cognac spirits, which are high in esters and are high in degree (70%), suffer a decrease in alcoholic strength which lowers the limit of esterification, whereas in the American Spirits, which are poorer in esters, are diluted before aging and present during this period an increase in the alcohol content.
In glass containers, the reactions that occur are less complex and the ratio of Acids: Esters is closer to normal equilibrium. However, according to Schidrowitz and Kaye, the alkalis yielded by glass have a marked influence on the rate of acids and esters, which are diminished. Valaer and Frazier found that young whiskey kept in bottles did not show, after 5 years, any significant change in composition; the unpleasant taste and smell of the young product had only disappeared. On the other hand, in the case of whiskeys already aged, they observed, after 4 years, a reduction in the acid content, up to 12 grams per hectolitre (5 grams on average), a gradual decrease of the aldehydes and especially of furfurol, finally, a slight increase of the esters and the coloring.
In the past, esterification was of primary importance in improving the bouquet of eaux-de-vie during aging. Some authors attributed the slow formation of esters to the fusion of the multiple tastes of recent brandies and the production of the so-called “fondue” flavor. Others considered that the tart and unpleasant taste of young spirits was due to fatty acids, and that the improvement of taste came from the disappearance of these bodies by esterification.
Subsequently, it was realized that the proportion of the esters did not increase appreciably during storage and that in many cases even there was a decrease in the absolute content. Various chemists have consequently been led to admit that these substances do not participate in aging, or at least play only a very secondary role in relation to the oxidation phenomena.
The fact that the overall amount of the esters does not undergo significant variations does not, however, exclude the possibility of profound changes in the nature of the esters. The observations of Bellet (1) even tend to show that this is what happens in reality.
(1) Bull. Soc. Chim. (5), 713, 1937.
This author has, several years apart, assayed the quantities of esters existing in the various parts of a brandy from the Charente, carefully divided into 4 equal parts. He obtained the following results:
These figures show that, while the total amount of the esters has remained almost constant, the nature of the esters has changed and this modification has tended to make them much more volatile.
[This is sort of a very simplified version of the birectifier methods. That first quarter fraction of “more volatile” esters represents an increase in ethyl acetate.]
In addition, the author found that the content of higher alcohols did not vary significantly: it was for each of the first quarters: 53 in 1913, 50 in 1922 and 51 in 1934, and for the totality of the liquid of: 205 in 1913, 195 in 1922 and 180 in 1934. Since the progression of the higher alcohols did not follow that of the esters in the first quarter, it would therefore be appropriate to admit a conversion of the esters of higher alcohols into esters of lower alcohols.
As regards the aldehydes, there has been a significant increase, in no way related to the small variation of the higher alcohols, the quantities found in 1913, 1922 and 1934 being respectively 5, 26 and 38.
Bellet explains the above results by the intervention of the phenomenon of alcoholysis. In weakly acidic alcoholic solutions, the esters are transformed into ethyl ester with release of the corresponding higher alcohol (Haller) (2). Then, the higher alcohols thus released oxidize gradually and turn into aldehydes. One of the essential reactions of aging, more important than the phenomenon of oxidation, would be the reaction of alcoholysis. This will occur all the faster as the acidic medium will be established more quickly. This is the explanation, according to the author above, why relatively neutral brandies age less rapidly than others and why rapid aging by artificial oxidation has not yielded so far the good results expected.
(2) C. R. CXLIII, 659, 1906.
[My understanding is that high alcohols could oxidize to and turn to aldehydes but in practice it is not significant. This idea from Arroyo of starting with a low pH to age quicker gets even more traction here.]
According to Trillat (3), the formation of acetals, by coupling between the aldehyde and alcohols, would play an important role in the constitution of the bouquet of wines and spirits:
(3) C. R. CXXXVI, 171, 1903.
The reaction, which takes place without the intervention of oxygen, occurs in the bottles and barrels, and would be favored by the presence of certain contact agents, such as tannin, and by the acidity of the medium. The presence of the acetal was characterized by the author in samples of old cognac.
Acetals have a great aromatic power. The scent of some of them is recognizable at the dilution of 1 / 10,000 (Trillat).
[I wish more time was spent on this. Supposedly acetals are important to defining Calvados.]
The observations made by the various authors who have studied the variations in the composition of spirits during aging (Rocques, Lusson, Crampton and Tolman, Valaer), tend to show that the higher alcohol content, if it increases relatively, as a result of the concentration of spirits in barrels varies little or slightly decreases in absolute value.
Crampton and Tolman found, in the case of “Rye Whiskeys”, a slight absolute increase until the second year (16 grams on average per hl of brandy at 50°), followed by a decrease thereafter, and for “Bourbon whiskeys”, a small, continuous decrease from first to eighth year. Valaer and Frazier saw higher losses, from 6.9 to 58.4 gr. per hl. of brandy (average 28.6 gr.), in whiskeys kept for 4 years.
According to Allen (1), the wood of the barrels would exert an absorbing action vis-à-vis the higher alcohols, which would explain the decrease of the rates of these substances in the aged spirits. Analyzing the liquid obtained by steaming old whiskey casks, the author found that it contained higher amounts of higher alcohols than commercial whiskeys. He also observed that debris of cork and oak chips, brought into contact with a solution of neutral alcohol containing amyl alcohol, retained a much higher proportion of it than ethyl alcohol.
(1) J. Soc. Chem. Ind. X. 305, 519, 1891.
According to Bellet, during the storage, new higher alcohols would be produced by alcoholysis of the higher esters and transformation into aldehydes of a part of these alcohols, the result of the two phenomena being the maintenance of alcohols greater than a roughly constant rate.
Some authors have considered higher alcohols as the cause of the unpleasant taste of young spirits (Allen).
Chemical composition of aged spirits.
Rocques, in France, studied variations in the composition of a Charente brandy placed in a new 25-liter oak barrel and treated as commercial eaux-de-vie. He obtained the following results:
Crampton and Tolman, in the United States, followed for 8 years the compositional variations of 31 barrels of 180-190 liters of whiskey, placed in heated storage during the winter (Rye whiskeys) or in unheated storage (Bourbon whiskeys). The barrels, made of white oak, were charred during assembly. The higher alcohols were determined by the Allen-Marquardt method. Coloration was determined in a half-inch cell using a Lovibond tintometer, and the results expressed in degrees of brewing scale. The figures below represent the averages provided by 14 samples of Rye whiskey and 17 of Bourbon whiskey, stored under conditions as varied as possible.
Valaer studied variations in the composition of rums kept for 2 years in new or used oak barrels, charred or uncharred. Coloration was determined as in the previous case and the higher alcohols measured by the official American method (modified Allen-Marquardt). We give below some of the analysis made by the author.
(1) Non déduites les quantités de liquide prélevées comme échantillions pour l’analyse,
(1) Used barrel. brick cellar, heated during the winter (average temperature 21°C).
(2) New barrel in charred white oak; heated cellar: temperature during the 1st year 26 – 30° and during the 2nd year 21 – 22°.
(3) New barrel in charred white oak; heated cellar, average annual temperature 21°.
(4) Used barrel; heated store and provided with a humidifier watering the barrels with hot steam. Average annual temperature 25° 5; relative humidity 85°. [not sure what that extra 5 implies or if the 85 should be 85%.]
(5) New charred oak barrel; heated store: average temperature 21°.
(6) New charred oak barrel; wooden store: average temperature 23°. [I think wood store implies a wooden rickhouse instead of something like a stone cellar.]
The same author has studied the changes that occur during aging of California wine spirits (1). These, unlike whiskey and rum, are housed in uncharred oak barrels and placed in unheated cellars during the winter.
(1) Uncharred barrel – Loss of liquid in 4 years: 29.23 l.
(2) Uncharred new barrel – Loss of liquid in 4 years: 38.36 l.
(3) Uncharred new barrel – Loss of liquid in 4 years: 20.32 l.
(4) Waxed barrel – Loss of liquid in 4 years: 5.02 l. (barrels of 180 – 190 liters).
The chemical changes are comparable to those noted for whiskeys and rums. However, they are more attenuated, slower and more regular. In the case of rums and whiskeys, the most significant changes take place during the first 6 months of storage.
Arroyo, in Puerto Rico, obtained, among others, the following results, with rums of molasses and vesou, aged in 5 gallon uncharred oak barrels, after dilution at about 50°:
These analyzes show the influence of the dilution on the hydrolysis of the esters, which continues during the first months of storage. There is a small reduction in alcoholic richness and a very strong increase in acidity, which contrasts with the results obtained by Valaer in the United States, which can be explained by the special conditions of storage (tropical climate of Porto-Rico, reduced capacity of containers).
Further analyzes of spirits of 1, 2 and 3 years led the same author to conclude that the acidity of rums of vesou and molasses continues to increase during the 2nd and 3rd years, but much more slowly. There is a steady increase of esters during storage, the rate of which is always higher in molasses rums. At the end of the first year of aging, the composition of the latter is much better balanced (particularly with regard to the Esters:Higher Alcs:Non-alcohol ratios) than those of Vesou. The equilibrium is reached almost in one year in the case of molasses rums, while it continues to improve in the case of rums de vesou, even after 3 years. This explains why the former mature much faster than the latter.
Arroyo, after numerous analyzes of rums of different ages, has found the following average ratios between the non-alcohol constituents. The spirits examined were obtained by the author in his fermentation work, diluted at 50° and kept in 5-gallon oak casks (which hastened the speed of aging).
The ratio of Esters : Higher Alcohols is significantly increased, especially in the case of molasses rums during the second year (which is partly due to the hydrolysis of the esters following dilution). The same is true of the Esters : Aldehydes ratio, which, however, is less significant. The Esters : Volatile Acids ratio decreases sharply as a result of the formation of the large amount of acids, but is approximately stabilized at the end of the first year: it remains above unity in molasses rums and goes down below 1 in rums de vesou.
Büttner and Miermeister (1) analyzed wine spirits stored in darkness for 18 months in half-filled glass bottles sealed with corks. Some samples were aged as they were and others were diluted with water to reduce the strength to 40°.
(1) Z. Unters. Lebensm. LXII, 317, 1931.
The phenomena of oxidation and esterification have been much more energetic than is habitually admitted for spirits stored in bottles. Also, note the elevation of the alcoholic strength for the samples brought back to 40° at bottling.
We obtained, at the laboratory of the Department of Agriculture of Martinique, very different results from the preceding ones, on rum samples stored during 4 to 12 years in white glass jars, of 250 cc., filled to about 4/5 and sealed with emery (1, 2, 3, 4, 6) or with corks (5, 7, 8). The flasks were placed in the light and the average room temperature was 30°C.
Observations — The analyzes were carried out by the same chemist using the official French methods, except for the dry extract, which was determined on the basis of the difference between the apparent alcoholic degree and the actual alcoholic degree (Blarez’s formula).
The rums examined had the following characteristics:
1: Rum of cane juice, aged in charred oak barrels, not containing caramel, aged 4 years at the time of the first analysis;
2, 3, 4: New rums of cane juice, not caramelized. The original white sample was yellowish at the time of the second analysis;
5, 6: New molasses rums, caramelized;
7 /: Grand arôme rum (Galion), not caramelized, aged in charred oak barrels for 3 years;
8: Grand arôme rum (Galion) new, non caramelized.
[Caramelized refers to the rum being colored or not.]
Regarding acidity, there is sometimes a more pronounced decrease for samples of uncolored young rums: 2, 3, 4, 8), due to the neutralizing action of the alkalis of the glass; and sometimes a slight increase. This behavior of the acids is very likely related to the low capacity of the storage containers and the relatively large surface area of the glass relative to the volume of the liquid.
[I think caramel may add a degree of buffering capacity. I’m also not sure if modern glass is different and we would get the same results.]
Aldehydes seem to tend towards a balance: they increase or decrease according to whether at the beginning they are in low or large quantity. Furfurol decreases. The esters increase each time the limit of esterification is not reached, they decrease if it is exceeded, following a neutralization of acids and a reduction of the alcoholic strength. There is no exception except in the case of sample 8, for which there must have been a significant saponification of the esters under the action of the alkalis of the glass. As for the higher alcohols, they have, with one exception (sample 5), an absolutely abnormal growth rate, which can only be explained by the imperfection of the measuring method used (irregularities in the composition of the liquor type).
[Aldehydes come to a roughly 50/50 equilibrium with acetals.]
Finally, we give below the composition, according to various authors, of very old eaux-de-vie:
Observations — 1 to 8 — According to Valaer – Higher Alcohols assayed by the US Official Method (Allen-Marquardt). 1 and 2 are molasses rums aged in the United States in used barrels for 19 years; 3 – 5 Jamaica rums of the “medium rum” type.
9 — Rum of cane juice, having received no addition of caramel, stored in the colony in charred oak barrels for 8 years. Analyzed at the Agricultural Laboratory of Martinique.
10 to 14 — According to Strunk – Rums kept in bottles for the durations indicated, after a certain stay in barrels. Higher alcohols measured by the Komarowsky-Fellenberg method and expressed as normal butyl alcohol. We translated into gr. per liter of eau-de-vie the figures given by the author in cc. % absolute alcohol.
15 to 19 — According to Lusson – Eaux-de-vie aged in casks in Charentes.
20, 21 — According to Rocques – Higher Alcohols measured, as in the case of samples 9, 15 to 19, by the colorimetric method Girard-Rocques.
In summary, aging is reflected, chemically, by an increase in the non-alcohol coefficient, when calculating it in relation to the hectolitre of pure alcohol contained in the eau-de-vie. However, this increase, mainly due to the concentration resulting from the loss of ethyl alcohol and water by evaporation, is often only relative. If we consider all the impurities contained in the barrel, we see that the non-alcohol coefficient varies little, the increases that some elements undergo being almost compensated by the decreases experienced by the others. The apparently dominant phenomenon is the production of acids by oxidation of the alcohol and dissolution of the extractive materials of the wood. The rate of higher alcohols changes little: it generally decreases in absolute value. It is the same with aldehydes, which increase or decrease in absolute value, according to the intensity of the oxidation phenomena. As for the esters, they increase or decrease according to the composition of the distillate and the conditions of storage. The balance between the various constituents of the non-alcohol is significantly modified during aging: the oxidation coefficient increases gradually, as well as the ratio Acids: Esters.
Practice of natural aging
Capacity. — American oak barrels, imported from the United States, are generally used for storing rums. The capacity of these containers varies according to the country: 250 liters (barrique) in the French West Indies, 180-195 liters (barrel) in the United States, 500 liters (puncheon) in Jamaica, 454 liters (puncheon) or 227 liters ( hogshead) in English Guiana, etc. The barrels usually used in France for eaux-de-vie are tierçon Charentais (1) (500-600 liters), barrique or pièce (260-280 liters) and quartaut (130-140 liters). Finally, we use for the storage of ready eau-de-vie, casks of varying sizes, all the more important that we want to avoid alcohol losses. They are usually made of wood, sometimes also of sheet metal, and often covered with a varnish or a white paint on the outside.
(1) In the past, the tierçon of 560 liters used to be used mainly in the Charentes, but as the aging took too long, it was replaced by the barrel of 270 liters.
Barrel treatment. — Rums are usually housed in new barrels. However, in Cuba, Puerto Rico and sometimes in the United States and Jamaica, casks are also used which have already contained spirits (whiskey, etc.) or liqueur wines (sherry, malaga , etc.) In these, aging is usually less rapid and especially less regular, probably because of their variable exhaustion in soluble principles. The losses in alcohol are also more irregular, being sometimes weaker sometimes stronger than those observed in the new barrels (Valaer).
The taste of spirits is unfavorably influenced by the quantity of soluble principles ceded by new barrels when the duration of the preservation is prolonged, it is preferable, in the case of brandies, to use containers which have already been used. In the Charentes, for example, brandy is only kept in new casks (previously stamped) for the time necessary for it to take a slight color (4 to 12 months), after which it is decanted into favorable used barrels. As the maturation is uneven depending on the quality of the barrels, those where it has been found to work best are carefully preserved. [SOS this could be translated much better. “previously stamped” is likely very wrong among other issues.]
Some authors admit that if the duration of aging should not be greater than 4 or 5 years, new oak is preferred. This is particularly the case for spirits from Armagnac, which are housed in new oak barrels and delivered for consumption after 3 or 4 years. This rapid aging takes place, however, at the expense of quality: the product obtained never reaches the finesse of cognac spirits, which are stored much longer in used barrels.
Most of the time (except in Cuba), the casks used to house the rums are charred at the time of assembly, placing them on a fire of chips or straw before putting on the hoops. This treatment gives the rum more body and color, and especially communicates a particular flavor, reminiscent of that of American whiskey. [SOS I took a risk that “les fonds” were hoops]
Finally, in some rare cases, when one wants to have a rum as little colored as possible and however having some aging, barrels are coated internally with a layer of paraffin or silicate are used. According to some authors, the eaux-de-vie would be apt to take paraffin wax in paraffin casks. [I think they are suspected of dissolving the wax.]
Brief storage in new barrels, which is the rule in the case of fine eau-de-vie, is practiced only exceptionally for rums. It is particularly useful when the quality of the wood leaves something to be desired.
Barrels are usually primed by steaming for about twenty minutes, until the condensed water flows colorless and without taste of any kind. For this purpose, tubular pressure steamers are used which supply steam at high temperature. In the absence of these devices, it is sufficient to scald the containers by pouring boiling water several times (4 or 5 times).
It is also possible, after scalding, to fill the barrels with a solution of soda ash (5 to 10 g per liter) for 2 or 3 days, then with cold water that is changed every 24 hours for a about week. Finally, the pure water is replaced by alcoholic water (20-25% alcohol), to which about 05% ordinary acetic acid (28%) has been added. This solution is left in the barrel until it is filled with the spirit.
[I’m not sure what 05% is or how it jumps to 28%. These directions seem pretty drastic like they are trying to re-use a barrel that held something weird.]
To clean used barrels with bad tastes, it is often necessary to treat them more vigorously. For example, they can be washed with the following solutions successively: a) boiling solution of 500 gr. caustic soda at 35° Baumé in 10 liters of water, per hl of barrel capacity; b) solution of 500 gr. of commercial hydrochloric acid (or 1 kg of sulfuric acid) in 10 l. of water per hl.; c) solution of 1 kg of quicklime, in small fragments, in a few liters of water. With each operation, stir and shake the barrel in all directions so that the walls get wet with the liquid. After resting for 2 or 3 hours, rinse vigorously with hot water, then with cold water, with about 20 liters of water per hectolitre of barrel capacity.
We can also, to remove bad tastes, rechar the inside of barrels (previously scraped), by placing them, the opening at the bottom, on paving stones of 15 cm. high and burning under the chips or small dry woods, until the staves warm up outside, about 10 minutes.
The type of cask to choose will depend on the quality of rum you want to get. For ordinary rums of molasses, use will be made of new charred oak barrels, which ensure faster aging, or, if you want to avoid the taste of whiskey, new barrels not charred and not seasoned. In the case of fine rums of vesou or syrup, it will be preferable to use new, well-seasoned barrels or barrels that have already been used. In any case, it will be important not to allow the spirits to remain in the new untreated or charred containers for a limited time (up to one year) and then to be transferred to used barrels.
Insects of the barrels. — During their storage in aging cellars and their stay in the hold of boats, the casks are frequently attacked by insects, which can cause significant losses of alcohol by leaking.
Cleare studied in English Guiana the conditions of development of these insects. He found that these were represented by several species of the genus Xyleborus (notably Xyleborus badius Eich.). These borers feed at the expense of certain fungi that develop in the galleries of the wood: they can therefore live only if it has a certain humidity. Barrels built with dry wood are not normally attacked. On the other hand, if the casks have been filled for a while with water or a weak alcoholic liquid, or if they have been in prolonged contact with the earth, they can absorb sufficient moisture to allow the development of insects. The use, for the stowage of the barrels in aging cellars or the holds of ships of planks belonging to species easily attackable by the Xyleborus, such as the Spondias monbin (monbin), the Cecropia (cannon wood), etc. , can also promote the attack of the barrels. It is the same for the presence of féntes or crevasses, produced during the assembly of the barrels. [can’t seem to translate “féntes”.]
When the insects enter the wood of the staves and the barrel is filled with brandy, they can not penetrate beyond a certain depth, the alcohol that impregnates the wood stopping their development. But if the attack occurs between the staves (especially with the help of the oakum used to seal the joints), in the bottom wood or in the groove that joins the bottom to the staves, leaks frequently occur through the “wormholes”. [I’m not aware of modern barrels using anything like oakum.]
The premises or aging cellars must be moderately ventilated and not too humid. Excessive moisture and lack of aeration promote the development of mold, and the musty smell that develops is easily absorbed by the brandy. As well, oxygen is difficult to renew, oxidation phenomena are slowed down and aging is delayed. Underground cellars, too humid and poorly ventilated, are not suitable. It is also important to avoid violent air currents, which unevenly advance the maturation of batches of casks, as well as the direct action of solar rays on the barrels, which would have the effect of increasing the temperature of the liquid and desiccation of the wood, resulting in abnormal evaporation.
In France, attics or stores are preferably used at ground level. These are usually built of stone. They are not usually ventilated and have, as far as possible, no other openings than the door, so that air and light do not enter. They are covered with a roof of tiles, so as to have a continuous aeration arranged and uniform. In the United States, warehouses are built of cement, tile or wood.
We do not seek in France, or obviously in hot countries, to guarantee the spirits against temperature variations. In the USA, on the contrary, the rum cellars are generally heated during the winter, so as to have an average temperature of 20-25°. It is the same for some whiskeys, while wine spirits are almost always kept at room temperature. The rise in temperature accelerates aging, but at the expense of the fineness of the product.
The casks are arranged on planks, rarely stacked. To save space, they are often available in large cellars on several floors. The rows are separated by narrow paths to allow aeration and inspection of the barrels. It is essential that the latter be frequent and easy, liquid losses may occur as a result of faulty woodworking or the attack of barrels by insects. Every effort must be made to regulate the aeration as well as possible, so that the drums, placed under the same conditions, age in a uniform manner. Finally, the cellars must be kept in a state of great cleanliness and sheltered from all bad smells, the latter being very easily absorbed by the alcohols.
Treatment of eau-de-vie.
In France, spirits, distilled at 65-70° (Cognacs) or 52° (Armagnacs), are put in casks as is. If the aging is long enough, the alcoholic strength is reduced to 45-50° at the end of the operation, and the spirits can be delivered directly to the consumer. In the French West Indies, a similar procedure is used for the aging of the rums, which, distilled at 65°, or more rarely at 70°, are kept in barrels until they are only 50-55°. It is the same in Jamaica, where however the initial measure is higher (about 80°).
In the United States, on the other hand, spirits are first diluted around the merchantable level (50°), before being subjected to aging: during this period the alcoholic strength increases by a few degrees.
To take into account the expansion of the liquid under the action of the temperature, the barrels must not be filled up to the bung: one leaves generally a vacuum of about 5 cm. The losses by evaporation being all the greater as the containers are less full, it is necessary to regularly fill the barrels during aging (ouillage). In order not to spoil the quality of the eau-de-vie, it is important to use for this operation a brandy of the same age and of the same origin as that found in the barrel. In some cases, in order to avoid the disadvantages of the mixing of spirits of different quality, we do not practice the ouillage: this is particularly the case for spirits of the Charentes (Lapparent). By not filling the barrels completely and by slightly blocking them, we get faster aging.
Duration of aging.
The duration of eaux-de-vie storage is very variable, depending on the method of aging applied and the nature of the product treated.
The light rums of the Cuba type and the “grappe blanche” of the French West Indies are not subject to aging, or only to an aging of very short duration (a few months).
In the United States, medium bodied rums, as well as whiskeys, are kept for a period ranging from a few months to 4 years, sometimes more in 180-190 liter oak barrels, placed in heated cellars.
In Jamaica, where rums are housed in tierçons of 500 liters, the minimum length of stay in government warehouses is 3 years, for products exported to England, and 5 years, for those destined for Ireland, New Zealand and Australia. But some rums can be kept for periods up to 40 years.
In the French Antilles, rums of vesou or syrup intended for the preparation of old rums remain 3 or 4 years in charred oak barrels of 250 liters then they are transferred to oak casks of 5,000 to 50,000 liters capacity where they stay before being bottled. Exceptionally, they are kept in 250-liter barrels for 5 or 6 years.
In France, for the eaux-de-vie of the Charentes, the duration of aging, in used tierçons of 500-600 liters, is generally 6 to 10 years, but some cognacs of quality remain in casks for 20 years and more. Armagnac spirits, on the contrary, are usually delivered for consumption after a stay of 4 or 5 years in new oak barrels.
Aging should not be pushed too far. Rocques wrote on this subject: “The stay of spirits in the wood should not be prolonged indefinitely, because their loss would eventually become too significant and we would only have worn out or faded eau-de-vie. Under no circumstances should the degree of a spirits be reduced to less than 45°. The 50° mark for old spirits may even be considered as a limit we must not go beyond. When the eaux-de-vie were produced at 65-70°, this reduction is observed after 30 to 40 years. It is usually considered that after 30 years the improvement becomes zero.”
This applies to the aging of wine spirits by the slow method used in the Charentes. In the case of rums, which are subjected to a much more energetic treatment, the duration is considerably reduced. In Martinique, the measure of 50° for rums put in cask at 65° is generally reached after about ten years. According to Arroyo, rum ages faster than whiskey, and after a period of more than 4 to 5 years, there would be little improvement in quality, at least in the conditions of aging in the tropics and the use of barrels of medium capacity.
The latter author also drew attention to the important role played by the composition of spirits in acquiring the character of maturation. Well-balanced rums, with no poor manufacturing taste, ripen much faster than others (1), and it can be said to a certain extent that aging occurs largely in the fermentation tanks and in the distillatory apparatus. The products resulting from a good fermentation and a well-conducted distillation can acquire their complete maturity in one year and even less. Arroyo declares that rums have been obtained in Puerto Rico which, after 6-7 months of aging, at the evaluation of Hamburg experts were found very comparable to the old rums of Jamaica kept in oak barrels for several years. A well-made rum may be ready for consumption after 6 months of storage. On the other hand, a defective eau-de-vie, poorly constituted, will never provide, even after an extended period in barrels, a product of poor quality, the defects of the young spirit are often amplified by aging.
(1) The elimination of bad tastes during aging generally occurs more slowly than the phenomena of dissolution and oxidation. On the other hand, alcohols that are very rich or insufficiently rich in impurities age more slowly than those with average richness and a well-balanced secondary product. [It is almost counter intuitive that “fuller bodied” spirits would age faster, but its due to the low starting pH from noble volatile acidity. Many of these bad tastes are also tufo and Arroyo gave us practical ideas to eliminate them.]
It is rare that very old spirits are delivered directly to the consumer. Most often it is used for the preparation of commercial marks, by cutting with young eaux-de-vie they improve.