Arroyo R. El Birectificador En La Justipreciacion De Los Rones Comerciales (1938) Revista de Agricultura de Puerto Rico, Volume 30, page 618-622.
The Birectifier In The Fair Evaluation Of Commercial Rums
Por RAFAEL ARROYO,
Jefe División de Quimica de la Estación Experimental Agricola de la Universidad
de Puerto Rico.
Dr. Curt Luckow of the Division of Liquors and Spirits of the Institute of Fermentation Industries of Berlin, has built and put into use a laboratory distillation apparatus which has been given the name Birectifier. It is the objective of this apparatus to enable the fractional distillation of the samples of commercial rums imported into Germany as an important part of the tests to which these rums are subjected for their classification and justification in terms of purity, quality and goodness of the product. It also serves to distinguish genuine rums from artificially prepared ones.
Our Experiment Station of Río Piedras, always alert to any innovation or discovery that can improve or clarify the technique of rum manufacturing, has imported one of these Birectifier devices, which have given us such valuable services, as they are giving in the already mentioned Fermentation Institute of Berlin.
We do not propose in this article to give details about the construction and mode of operation of the Birectifier; but all those interested people are cordially invited to visit us and can then see the device and how to use it.
What we do propose is to present the data obtained from a fractional distillation of one of our Jamaican-type experimental rums, and the conclusions we reached based on the analytical data and other observations made during the fractionation and subsequent manipulation of each of the fractions obtained.
The distillation in the Birectifier was carried out with a sample of Jamaican rum fermented and distilled in our laboratory. This rum had fifteen months of aging in oak barrels and the sample taken for this work was 240 milliliters, previously reduced to an alcohol concentration of 40 percent by volume (according to the instructions of Dr. Luckow). Seven fractions of 25 milliliters each were obtained in a time of one hour and forty-five minutes, that is, taking an average of fifteen minutes for the distillation of each fraction.
[Arroyo eventually modified Luckow’s 96 ml of absolute alcohol framework to be 100 ml (250 ml at 40%) and uses 8 fractions likely because we he is interested in investigating tufo that presents itself at the end.]
The following table presents the temperatures at which each fraction began and ended, as well as the analytical data obtained in the tests carried out on each fraction. In the eighth column, the physical appearance of each fraction is recorded.
The first two fractions passed at a constant temperature of 79° C. In the third fraction the temperature began to rise very slowly, starting at 79° C. and rising in degrees of degree until at the end of this fraction the mercury marked exactly 81° C. With the distillation of the fourth fraction, a more marked rise in temperature began, reaching 88° C. at the end. The fifth fraction started at 88° C., and ended when the temperature reached 97° C. The sixth fraction started at 97° C. and ended at 100° C., with a large turbidity in this fraction from the beginning. We collected a seventh fraction that passed at a constant temperature of 100° C, being completely constant all the time.
[These fractions align differently than my model because of less absolute alcohol and for some reason I believe this early model had less fractioning power.]
The different fractions of 25 milliliters each were collected in special graduated cups with glass lids, said cups being part of the accessory equipment that comes with the Birectifier.
The distillates Nos. 1 to 5 were then diluted to approximately the same alcohol concentration, about 30 percent by volume; then all the distillates were left at rest for half an hour, taking care to cover them with their corresponding glass lids.
After this interval of time the characteristic aroma of each fraction was examined, noting the following:
Distillate No. 1 offered an intense and pleasant aromatic smell; but that in no way reminded us of the characteristic smell of the original sample used in this fractionation. Rather it offered a compound aroma, as it was produced by a mixture of aldehydes and aromatic esters; this aroma being soft and pleasant, although of great intensity. The cup containing the sample of distillate No. 2 was less intensely aromatic, presenting a flavor which, although similar to that of cup No. 1, was, however, finer and less penetrating.
[I often encounter these spirits with concentrations I describe as non-culinary. Arroyo may be finding the sample pleasant (as I have in certain case studies) likely because the spirit does have much ethyl-acetate owing to its young age.]
By rubbing a drop of each of these first two fractions between the hands, an intensification of the aroma was felt in both cases; but this effect vanished relatively soon, leaving no residual aroma or odor in the hands. The cup containing the third fraction had an aroma that at first seemed very similar to that of the fraction No. 2. However, when doing the rub test of a drop between the hands, we could observe that after the first intense smell, and already apparently dry hands, we could observe a characteristic aroma of a mixture of higher alcohols; but whose character we could not specify. In the cup containing fraction No. 4 and the aroma changed greatly from those previously examined distillates. This aroma, being soft and delicate, began to remind us of the original bouquet of the distilled sample. These qualities of the aroma intensified greatly in the fraction No. 5, the smell being in this case very similar to that of the original sample distilled fractionally. The rubbing test between the hands of representative drops of the distillates Nos. 4 and 5, and especially of the latter, left a pleasant and persistent odor in the hands very similar to the aroma of the original sample of the fractioned rum. Cup No. 6 offered us a compound and complicated aroma, of difficult character to define; and the seventh cup that contained the last fraction presented us with a slight aromatic tone comparable to none of the previous ones.
We empty the respective contents of each cup in bottles prepared for that purpose, and without trying to remove the residual liquid left in each of the cup, we cover them (the cup) with their respective glass caps, leaving them for 24 hours. (The glass covers do not hermetically cover the cups but allow slow evaporation of their contents.) The re-examination of the cups after this time gave us the following data:
[I do this test with 1 ml of sample in a tasting glass with a watch glass cover. It is quite insightful for something so simple.]
No aroma was noticed in the cups containing the residues of fractions Nos. 1 and 2; weak aroma in the cup which contained fraction No. 3; more pronounced aroma in which contained the fraction No. 4; and very intense and characteristic aroma in the cup that contained fraction No. 5. The cup that contained fraction No. 6 also retained enough aroma although it did not have the purity of the cup, No. 5. In the cup that contained the last fraction, that is No. 7 we could not notice any aroma, although it kept a greater amount of aqueous residue than any of the others. We also observed that cups No. 1 and 2 were almost dry; Nos. 3 and 4 retained more moisture; in No. 5 we could notice a droplet of an oily character that ran slowly through the walls of the cup while rotating it between the hands; while in the No. 6 flask, a residue similar to a mixture of oil and water was noted and the smell was very similar to that of cup No. 5, although less intense and more complex.
Leaving those cups whose residues still offered aromas (that is, Nos. 3, 4, 5 and 6) completely uncovered for an additional 12 hours, we could observe, after this time, that there was almost no aroma in Nos. 3 and 6; but that Nos. 4 and 5, (and especially No. 5) still retained the characteristic aromas of their respective fractions. Returning to cover these two cups, and observed every 24 hours we could see that even after 7 additional days the aroma persisted although apparently the cups were dry. These cups are still under observation.
[I have been able to observe this with the most beautiful rums, especially those from Hampden. The aroma is never fruity, but smokey, leathery and quite complex. It seems to change as other components evaporate around it.]
Having made these preliminary observations, we then proceed to carry out the analytical work on each of the distillates stored in the aforementioned bottles.
Before commenting on the analytical results, we will mention some valuable observations made during this work: When diluting the samples with distilled water in their preparation for the determination of esters, those corresponding to fractions Nos. 1 and 2 did not suffer any notable alteration, while samples of fractions Nos. 3, 4 and 5 became noticeably cloudy, this turbidity disappearing, however, by incorporating the excess of alkali accustomed for the hydrolysis of the esters. After the esters were titrated in each case, we could notice that almost all the characteristic aroma had disappeared in the samples representative of fractions Nos. 1 and 2; that these aromas persisted with almost equal intensity and purity in the samples representative of fractions Nos. 4 and 5; and that in the numbers 3 and 6 one could also notice part of their respective aromas, although much less intense than those of Nos. 4 and 5. It was also noted that in this case the residual aromas of Nos. 3 and 6 resembled those of Nos. 4 and 5 more than before.
[I am on the edge of this in my own analytical work having finally started to tackle titration. The alkaline first neutralizes free acids and then starts to hydrolize the esters so basically aroma is gradually subtracted making rum oil more evident. I will restate that I think the fractioning of my birectifier is more powerful than Arroyo’s.]
Now studying the analytical data of the table above, we will see that:
1. Almost all ethyl alcohol passes into the first four fractions.
2. Most of the esters pass in the first two fractions; they greatly diminish during the distillation of the third and fourth fractions, and increase again in the fifth and sixth fractions. If we calculate this increase in esters based on the alcohol contained in each fraction, we would immediately see that it is very remarkable.
[Keep in mind ordinary ethyl acetate is significantly more volatile than other longer chain, higher value esters.]
3. Volatile acidity is low in the first four fractions, to increase rapidly during the last three.
4. Almost all the aldehydes are found in fraction No. 1, and they decrease notably in the fraction No. 2 to be preserved with almost equal values during the rest of the distillation.
[Keep in mind, acetaldehyde versus everthing else.]
5. It should be noted that the alcohol concentration hardly varies in the four first fractions.
From the analytical data presented above and from the other observations made during the course of this work, we have drawn the following deductions.
1. Most esters and aldehydes pass during the first two fractions probably in azeotropic mixtures of these bodies with ethyl alcohol, which is responsible for the low and constant temperature in which these first two fractions are obtained. These distillates, although intensely aromatic, do not present an odor similar to that of the original rum sample.
[The generic nature of the first two fractions holds true across nearly all spirit types.]
This information is important and must be kept in mind during the subsequent reading of these deductions.
2. The initiation of the upward change in the distillation temperature that takes place during the second half of that of the distillation of the third fraction indicates that bodies with higher boiling points begin to pass, probably higher alcohols and higher molecular weight esters than those of fractions Nos. 1 and 2.
3. Bodies of higher boiling points apparently pass during the distillation of fractions Nos. 4, 5 and 6; some of which are of little solubility in water, since from the beginning of the distillation of the fraction No. 6 there is remarkable turbidity in the distillate that acquires an emulsion aspect. There is no doubt that fractions Nos. 4 and 5 also pass these substances which are not very soluble in water, but as in these cases the alcohol concentration is still high, this prevents the separation of said bodies, which are more soluble in alcohol than in water. This deduction is corroborated by the fact that when the samples of these fractions are diluted for the determination of esters, they suffer the same turbidity phenomenon observed in distillate No. 6.
[emulsion settled after 24 hours.]
4. When the determination of ethyl alcohol in fraction No. 7 is negative, we assume that the residual aroma found here comes from very small amounts of esters and aldehydes with very high boiling points. Obviously, the opalescence-producing bodies in the distillate no longer exist here (at least in appreciable amounts) since fraction No. 7 remained clear and crystalline all the time.
5. The characteristic aromas of fractions Nos. 4, 5 and 6, and particularly that of fraction No. 5, makes us think that it is rather the aroma discharged by some body that passes in these fractions, to which is owed the peculiar smell of the original fractioned rum. There is no doubt that this aromatic substance does not belong to the family of esters, since this aroma subsists and persists through the determination of esters in which these are completely hydrolyzed, thus losing their original aroma.
[At this point, Arroyo does not appear to point to rum oil which we now know are various rose ketones.]
The fact that the fraction that occupied cup No. 5 kept its aroma long after those of all other distillates makes us think that this substance passes preferably during the distillation of this fifth fraction, that is between 88 and 97° C. and an alcohol concentration in the distillate of 59.3 percent on average.
The physical appearance of the droplet found at the bottom of cup No. 5 induces us to think that it is an essential oil that is slightly soluble in water, more soluble in alcohol, volatilizable with some difficulty, and whose boiling point is higher than that of ethyl alcohol or that of the ordinary esters found in rum, since it does not pass with these during the distillation of the first two fractions.
We propose in the future to use this Birectifier in the meticulous examination of our national and foreign rums for comparative purposes, as well as to compare them with the different types elaborated by us in our research work on rum manufacturing.
[As I mentioned earlier, Arroyo eventually advanced some of the German protocols such as working with 100 ml of absolute alcohol in a sample (instead of 96) and collecting 8 fractions like Karl Micko.]