PLACE DES RHUMS AU SEIN DES EAUX-DE-VIE
P. de SMEDT et P. LIDDLE
Centre de Recherches de la S. A. F. Martini et Rossi,
19, avenue Michelet, 93404 Saint-Ouen
[This is an extremely interesting paper and splits its time between the esters and rum oil. The data tables on esters is quite fascinating. We also see the profiling a few different Grand Arôme rums and a very intriguing operation in Madagascar. My latest theory is that rum oil (now simply a marketing term) is primarily the rose ketones. This paper however, supports the idea that rum oil likely also encompasses TDN and TTN (TDN is the infamous petrol of Reisling wines). Additional components of rum oil may be: methyl-2 butanol-1 and the methyl-3 butanol which I don’t know enough about at the moment. All of these compounds have very low thresholds of perception.
The paper curiously describes Jamaican rum as lacking Damascenone relative to Martinique’s Galion. Keep in mind, this is an early paper. They may have been missing something else important that Jamaican rums have.]
The purpose of this study is to differentiate between spirits either by the presence of certain characteristic compounds, or by the ratio of certain constituents.
Using a column filled to 2 p. 100 of glycerol and 2 p. 100 hexanetriol-1,2,6 on Chromosorb R 100-120 mesh allows the separation of the two isomers 2-methyl-1-butanol and 3-methyl-1-butanol. The study of the ratio of these two isomers was carried out on rums, Cognacs, whiskeys, Armagnacs, brandies and Calvados.
The liquid-liquid extraction technique in a Mascré type apparatus, with methylene chloride as solvent, allows a semi-quantitative extraction of esters, furfural and beta-phenylethyl alcohol.
The gas chromatography of these extracts makes it possible to highlight certain specific characteristics of the following eaux-de-vie: rums, Cognacs, Armagnacs, Calvados, whiskeys and bourbons. This method is characterized by its reproducibility and rapidity, the duration of the analysis being reduced to that of chromatography.
The gas chromatography-mass spectrometry coupling makes it possible to identify, in certain rums, particular compounds, such as trimethyltetrahydronaphthalenes (TTN), trimethyldihydronaphthalenes (TDN), damascenone, etc.
The official analysis of an eau-de-vie with a total quantification of the different chemical functions gives little information on the type and quality of the spirit. On the other hand, if tasting can define the type of a product and say that it is good or bad, it does not identify the causes.
It therefore seemed interesting to try to define criteria to better characterize the various types of spirits, including rums.
As far as rum is concerned, a more detailed study of its olfactory components has been made to try to find some specific characteristics of the different types of this spirit.
It is obvious that the results of this study will be significant only if the analyzes were made on a sufficiently large number of samples.
1. – STUDY OF HIGHER ALCOHOLS
1. – Choice of the technique
Since the calculation of the impurity content of a brandy involves the quantity of higher alcohols it contains, it is important to be able to determine the latter with good precision. It should also be noted that higher alcohols are often the most important part of impurities.
In recent years, analysis by gas chromatography has become the method best suited to the determination of higher alcohols. However, among all the methods proposed, it is sometimes difficult to make a choice,
The BARAUD (1961) technique, which used ethyl alcohol as an internal standard, should be mentioned as a reminder. This method was rather imprecise because of the excessive amounts of ethanol causing saturation at the detector.
The method of BOUTHILET and LOWREY (1959) by external calibration is also to be rejected because it requires a perfect reproducibility of the injection and the analytical conditions, which is difficult to obtain.
The methods used by WEBB (1961), FLANZY-JOURET (1963), and BOIDRON (1966) are rather impractical because they require prior extraction of the higher alcohols.
BRUNELLE (1967) uses an internal calibration, which clearly improves the value of the results obtained. However, the chosen standard, n-butanol, sometimes exists in a desirable quantity in certain spirits. The use of n-pentanol as an internal standard by Singer (1966) has made it possible to avoid this disadvantage since this compound has never been reported in spirits. [they are putting in something specific of a known value along side their mystery stuff so if they count the known value correctly, they are counting the other stuff correctly.]
Apart from the problem posed by calibration, there is the choice of the column and the chromatographic conditions. As regards the choice of the column, a large quantity of supports and stationary phases have been proposed:
— Carbowax 1500 on chrom W, SCHOENEMAN, DYER (1968), (1973), BRUNELLE (1968), JOURET and MOUTOUNET (1968).
— Carbowax 20-M, GUYMON (1970).
— Carbowax 1540, French official method (1973).
These columns all have the disadvantage of not separating 2-methyl-1-butanol (active pentanol) from 3-methyl-butanol-1 (isopentanol).
Among the stationary phases allowing to separate more or less well the 2 isomeric pentanes, we can mention:
— glycerol, VAN DER KLOOT, TENNEY, BAUISSOTTO (1958);
— Tide, PORCARO, JOHNSON (1961);
— triethanolamine, SIHTO, NYKANEN, SUOMA LAINEN (1964);
— diethyl tartrate, SINGER, STILES (1965), SINGER (1966);
which give an incomplete separation.
Complete separation between the two isomers was obtained by SINGER (1965) (1966); with a polyethylene glycol 200 column, and by KAHN and BLESSINGER (1972), with a 2 p. 100 glycerol, 2 p. 100 hexanetriol 1,2,6.
In the end, our choice fell on the method proposed by KAHN and BLESSINGER recommended by the AOAC, because it is fast (18 mn) and gives a very good separation the methyl-2 butanol-1 and the methyl-3 butanol -1, important compounds for the characterization of certain spirits. The use of pentanol-3 as an internal standard makes it possible, if present, to determine n-butanol. Ethyl acetate can also be determined with this compound. It should be pointed out, however, that this column does not quite separate propanol from secondary butanol.
Description of the column used:
Column 2 p. 100 glycerol + 2 p. 100 hexanetriol 1,2,6 on gas chrom R (100-120 mesh, not washed with acid) – 10 feet Ø 1/8. [Ø = diameter, I assume]
Temperature of use: isothermal at 80 ° C.
Nitrogen flow: 24 ml / min.
2. — Résults
Figure I gives an example of separation obtained with this technique. Table I shows the distribution of the ratio r; r = 2-methyl-1-butanol / 3-methyl-1-butanol, in various alcohols. These results agree very well with those found by SINGER (1966) and JOURET and MOUTOUNET (1968).
II. – SEMI-QUANTITATIVE STUDY OF ETHYL ESTERS OF FATTY ACIDS, ETHYL LACTATE,
DIETHYL SUCCINATE AND FURFURAL
I. – Extraction technique
1. I. Description of the extraction apparatus.
The extraction of the sample is carried out in a MASCRE machine (1957). This device is a type of countercurrent liquid-liquid extractor, with up to 6 ampoules of the type shown in Figure 2. The bulbs in vertical position undergo a rather slow rotation of 180° and then remain stationary for a defined time . This time that is displayed is that necessary for the complete reversal of the two phases.
The number of complete tours can also be displayed, the extraction step is thus made automatic and can be performed at night. This apparatus makes it possible to carry out very reproducible extractions and also to avoid the formation of emulsions. Thus it is not necessary to make several extractions of the same sample to recover all the compounds, just add an internal standard before extraction and calibrate the method still working under the same conditions.
I. 2. Choice of the report:
Volume of alcohol * / Volume of solvent = R
The term alcohol means in fact any brandy at 40 ° GL.
For the determination of the esters, it is preferable not to have to concentrate the extract, so as not to introduce an additional step likely to reduce the reproducibility of the method.
If we consider any ester with:
[the “of o, of a, etc., all exist as substrates wordpress could not handle]
Q of o: Quantity of this ester in a volume x of alcohol before extraction.
C of o: Concentration of the ester in the initial alcohol.
Q of a: Quantity of ester in x volume of alcohol after extraction.
Q of a: Amount of ester in 300-x volume of solvent (CH2Cl2 [methylene chloride]) after extraction (with the volume of one ampoule equal to 300 ml).
C of s: Concentration of the ester in the solvent after extraction.
K: Coefficient of distribution of the compound between the solvent and the alcohol.
[I think what is happening here is they are trying to determine how good their method is. They want to be able to extrapolate where any known loses are when they move from slower wet chemistry techniques to faster spectroscopy based techniques.]
So, to have C, maximum, one needs K / R minimum, that is to say R the highest possible.
In practice, in this case, an R = 10 ratio is sufficient for the analysis.
After extraction for 300 revolutions, the methylene chloride is recovered and 5 μl are injected on a DEGS column 4 m 1/8 to 10 P. 100 on chrom W 60-80 mesh.
Oven temperature: programming from 60 to 170° C at 3° C / min.
2. — Résults
Figure 3 shows a chromatogram obtained with a Martinique rum.
Tables 2 and 3 show the results of analyzes carried out by this method on various samples of spirits and rums.
Ethyl and furfural esters of different spirits
(Average value in mg per liter of pure alcohol)
Ethyl and furfural esters of different rums
(Average value in mg per liter of pure alcohol)
2. 1. Cognac brandy counts 0. [new make?]
This study was carried out on 26 spirits from different regions and years.
Despite significant variations in quantitative terms, a more detailed study shows that the proportions between the different esters of the same Cognac are particularly constant from one region or from one year to the next (unpublished personal work), that is to say that the aromatic “profile” has a fairly remarkable consistency (this certainly reflects a high standardization of the elaboration process and a constancy of the yeast flora).
2. 2. Cognac marketed. [this describes aroma breakage?]
A decrease in the amount of ethyl esters of high carbon number fatty acids is noted, which certainly reflects the influence of the decrease in alcohol content at 40 ° GL followed by stabilizing refrigeration treatment.
2. 3. Armagnac.
The most important fatty acid ethyl ester of this type of spirit is caprylate, while caprate predominates in cognacs. The quantities are lower in armagnacs than in cognacs.
2. 4. Brandies.
In this type of spirit, there is no particular character compared to cognacs, if not smaller quantities of esters and a significant dispersion of the results.
2. 5. Calvados.
These spirits are characterized by significant quantities of hexanol, triethoxypropane (greater than 20 mg per liter of pure alcohol) and amounts of 3-phenethyl acetate (of the order of 2 to 3 mg per liter pure alcohol).
2. 6. Whiskeys (blended and malt).
The essential characteristic of whiskeys is the presence of ethyl palmitoleate in quantities greater than ethyl palmitate. It should also be noted the presence of ethyl 9 decenoate (probable identification) and β-phenethyl acetate.
2. 7. Bourbons.
These spirits, unlike whiskeys, do not have significant amounts of ethyl palmitoleate.
2. 8. Rums. [Is the unique character of Reunion rums due to Saprochaete Suaveolones and ethyl tiglate?]
In Table 3, we can see that the Martinique, Madagascar and Guadeloupe rums (industrial or agricultural) have a very similar “aromatic profile”. As regards Reunion rums, the differences recorded, in particular with regard to ethyl caproate, are due to two rums of rather particular character, possessing rather large quantities of ethyl caproate. Jamaican rums are essentially characterized by very large quantities of ethyl caproate. “Grande Fond Galion” rums have lower amounts of ethyl esters of fatty acids, but, on the other hand, fairly large amounts of ethyllactate.
The differences recorded between French light rums and foreign light rums certainly reflect methods of dissimilar elaboration.
III. – QUALITATIVE STUDY OF RHUMS: PRESENCE OF PARTICULAR COMPOUNDS
1. – Extraction technique
1. 1. Apparatus.
The apparatus used is the MASCRE extractor, already described.
1. 2. Solvent.
The solvent used in this case is n-pentane, chosen for its very good volatility and the fact that it extracts little alcohol (BOIDRON, 1966).
1. 3. Choice of the report.
Volume of Alcohol \ Solvent Volume = R
In the present case, since the extract will be concentrated to the maximum, it will be necessary to choose optimal R in such a way that Q of s is maximal (for a given compound).
Q of s will be maximal for R = √K. In this case, the choice of R will therefore be a function of the distribution coefficient of the compound to be extracted. Since for the compounds to be extracted K will be close to or greater than 20, the extraction will be carried out with R = 5, ie 250 ml of 40°GL and 50 ml of pentane.
1. 4. Méthod. [sample preparation for chromatography is pretty much as involved as birectifier operation.]
Two hundred and fifty ml of rum are extracted with 50 ml of pentane on the MASCRE apparatus for 300 rounds. The pentanic extract recovered is dried over anhydrous sodium sulphate and then concentrated to about 2 ml in a flask with an air-cooled condenser on a water bath at 60 ° C. The concentrated pentane is taken up with a Pasteur pipette and placed in a tapered tube on which the vacuum is made, so as to completely eliminate the pentane. The oily extract is thus obtained at a volume of the order of 0.5 to 5 μl.
The analysis is performed on a SCOT FFAP column with a length of 100 feet, and internal diameter of 0.02 inches (stainless steel column). This column is placed in a chromatography-mass spectrometer coupling Perkin-Elmer 270, the quantity injected is 0.15 μl [100 pieds, is that really feet?]
Conditions: oven temperature: programming from 80 to 210 ° C at 4 ° C / min.
2. — Résults
2. 1. -Jamaican rum (8 samples analyzed).
Figure 4 shows a typical chromatogram of Jamaican rum.
These very typical rums are characterized mainly by very large quantities of volatile compounds (ethyl esters of acids with carbon number of between 2 and 6, and acetals, in particular diethoxyethane, diethoxyisobutane and diethoxyiso pentane). These rums also contain significant amounts of ethyl esters of fatty acids with an odd number of carbons (C7, C9 and C13). [These acetals are what I think we need to learn a lot more about.]
It should be noted in this type of rum, the presence of two trimethyl tetrahydronaphthalenes (TTN) and a trimethyl dihydronaphthalene (TDN). These compounds were identified by their characteristic mass spectra, published by LIEBICH, KENIG and BAYER (1970), who had also found them in Jamaican rum. [dont’ forget TDN was also found in Australia by D.A. Allen a few years earlier.]
Two Jamaican rums had low amounts of menthol and 2 sesquiterpenes, one of mass M = 204 (peak no. 8) with mass spectrum: 41 (100) 119 (69) 93 (66) 41 (58) 55 (53) 27 (49) 95 (42) 111 (31), the other of mass M = 202, corresponding to ar-curcumene. [a component of many botanical essential oils]
The other six Jamaica rums studied contained significant amounts of propyl caproate, isobutyl caproate, isoamyl caproate and hexyl caproate, as well as significant amounts (greater than 5 mg / liter of pure alcohol) of methyl chavicol and α-terpineol.
It should also be noted the presence, in all Jamaican rums, of a particular acetyl triethoxypropane (1), at a rate of 1 to 10 mg / liter of pure alcohol approximately The mass spectrum of this compound is given in FIG. The presence of this acetal has been reported in “peppered” whiskeys by KAHN, LARDE, CONNER (1968), and, recently, by DUBOIS, PERFECT, DEKIMPE (1973) in a rum with an abnormal taste. In fact, the presence of this acetal, far from being exceptional, is on the contrary common, since we have found it, in a very variable quantity it is true, in different rums and whiskeys. In addition, it is an important constituent of Calvados. [solving this fault was figured out by the INRA in Martinique]
(1) commercial product of Aldrich Europe.
Three compounds present in all Jamaica rums have not been identified: their mass spectra are as follows:
peak 2 = 43 (100) 41 (60) 27 (55) 29 (50) 45 (43) 99 (43) 117 (30) 60 (25);
peak 3 = 97 (100) 125 (69) 39 (60) 29 (59);
peak 4 = 43 (100) 41 (83) 55 (43) 117 (41) 84 (39) 56 (39) 99 (38).
2. 2. “Grand Fond Galion” rum (5 samples analyzed). [Galion is the last Grand Arôme of Martinique]
Figure 6 shows a typical chromatogram of a “Grand Fond Galion” rum.
The aromatic profile of this type of rum is characteristic.
In the five rums studied, triethoxypropane is present in varying amounts. Five trimethyl tetrahydronaphthalenes (base peak 159 and peak M + = 174 important), as well as trimethyl dihydronaphthalene [reisling “petrol” TDN] (peak base at 157 and peak M + = 172 important) were identified. We also note the presence of various special compounds such as benzaldehyde, menthol, ethyl benzoate, diethylsuccinate, ethyl-3-phenylpropionate, the latter compound being quite characteristic of rum “Grand Fond Galion”.
Phenolic compounds, such as methyl salicylate, methyl guaiacol (possible), ethyl-4-guaiacol and prophyl-guaiacol have been identified (HRUZA and VAN PRAAC, 1974).
These rums contain a very particular compound: the β-damascenone (1), which we will find in appreciable quantity in all the other types of rum, except the “Jamaica” where it exists only in the state of traces. A precise dosage on a Grand Fond Galion rum gives a damascenone content of 1.4 mg / L of pure alcohol. The identification of this compound was carried out by mass spectrometry (Fig. 7) and by the measurement of its retention time. Until now, this compound has only been reported as constituting Bulgarian rose (Rosa damascena, MILL.) (DEMOLE, ENGGIST and SAUBERLI, 1970), tobacco essential oil (DEMOLE and BERTHET, 1971, 1972), raspberry (WINTER and ENGGIST, 1971), the volatile portion of cooked apples and wines (SCHREIER and DRAWERT, 1974).
[Wow, Galion is really kicking ass but what of Jamaica? Did they have poor samples? Did Jamaica have a lull in technique? What do we make of this? Did they miss other rose ketones or isomers that were more likely?]
On the other hand, this type of rum contains a number of compounds that we have not yet identified, including the following peaks:pic 2 = 55(100) 41(91) 111(76) 29(53) 27(48) 43(47) 39(44);
pic 3 = 43(100) 87(72) 41(68) 59(64) 55(56);
pic 5 = 43(100) 41(70) 27(60) 55(54) 29(49) 71(40);
pic 9 = 165(100) 137(71) 135(60) 91(47) 134(29) 43(27) M+ = 180;
pic 13 = 43(100) 41(79) 91(68) 105(62) 29(57) 55(55) M+ = 190
pic 15 = 43(100) 41(95) 91(77) 39(65) 105(53) 55(59) M+ = 192.
(1) We thank Firmenich for providing us with a sample of this product (trade name: Doricenone).
2. 3. Rhums agricoles (6 samples analyzed).
Figure 8 shows a typical chromatogram of agricultural rum. This type of rum has less marked characters than the previous ones. We can even consider that the chromatographic profile of this type of rum is closer to that of a cognac or brandy than a “Jamaica” or a “Grand Fond Galion”. The analysis of these six agricultural rums showed a perfect constancy of this type of product, in which only average amounts of triethoxypropane exist, traces of TTN and minimal amounts of menthol and damascenone. The dosage of damascenone in an agricultural rum from Martinique gives a content of 0.25 mg / 1 of pure alcohol. [Galion had roughly six times more.]
2. 4. Industrial rums (8 samples analyzed).
These rums have a chromatographic profile close to that of agricultural rums, but with a greater diversity, especially at the level of minor constituents. For example, for a rum from Madagascar, there are quantities of the order of one milligram per liter of pure alcohol for TTN and TDN. The precise dosage of damascenone in an industrial rum gives a content of 1.5 mg / liter of pure alcohol. [Um, I’m intrigued, but this was 1970. Is this the producer Dzama or a lost rum?]
We must also note the presence, in this rum as well as in a rum of Guadeloupe, of an isomer of eugenol, whose mass spectrum is as follows: 164 (100) 91 (75) 77 (68) 41 (67) 78 (66) 43 (52) 149 (49) 121 (48). The retention time of this compound is lower than that of eugenol on FFAP column.
In three rums from Puerto Rico and Guyana, the presence of a lactone with a peak base at m¦e = 99 was recorded. This is most likely β-methyl ϒ-octalactone; its mass spectrum is very close to that given by OTSUKA, ZENIBAYA and ITOH (1974). This compound has also been identified in the Bourbons.
2. 5. Rums from Réunion “Bois rouge” (3 samples) and “Quartier Français” (1 sample).
Figure 9 shows a typical chromatogram of a “Bois Rouge” rum.
The analysis of the three “Bois Rouge” rums showed that they had a rather similar composition, the main characteristics of these rums being significant quantities of triethoxypropane (greater than 10 mg per liter of pure alcohol), the presence of 9 peaks having a mass spectrum close to trimethyl tetrahydronaphthalene, two peaks of trimethyl dihydronaphthalene amounts (greater than 5 mg per liter of pure alcohol) of methyl salicylate and damascenone. The isomeric compound of eugenic, previously noted in the rums of Madagascar and Guadeloupe, is also present.
One of these three rums, very rich in TTN and TDN, also contained the two unknown compounds reported in the “Grand Fond Galion” (peaks 13 and 15 of masses 190 and 192 respectively).
The rum “Quartier Français” has the characteristics of an industrial rum, but with larger amounts of damascenone and methyl salicylate.
This study revealed the presence, in most rums, of very specific compounds: TTN, TDN and β-damascenone, probably resulting from the degradation of β-carotene.
Indeed, MULIK and ERDMAN (1963), DAY and ERDMAN (1963) showed that ionene, or 1,1,6-trimethyl-1,2,3,4-tetrahydronaphthalene (TTN) was formed by thermal degradation of the β-carotene. ROE and SHIPLEY (1969) found that in whiskeys the α and β-ionones came from the thermal degradation of β-carotene. BOGET and FOURMAN (1933), by catalytic dehydration, transformed the α and β-ionones into ionenes [never heard of ionene]. According to DEMOLE and BERTHET (1971), β-damascenone is believed to be a photoxidative degradation of β-carotene.
It would therefore be interesting to determine at what stage of the production of rums these compounds are formed and to know the factors driving the degradation of carotenes or vitamin A, either to TTN and TDN, or to damascenone.
ANONYME, 1973. Méthodes officielles d’analyse des alcools et eaux-de-vie. J. Off. République Française, 38, 2 oct., 1231-1266.
BARAUD J., 1961. Étude quantitative par chromatographie en phase vapeur des alcools et des esters de la fermentation alcoolique. Bull. Soc. Chim., 10, 1874-1875.
BOGERT M. I., FOURMAN V. G., 1933. J. amer. chem. Soc., 55, 4660.
BOIDRON J. N., 1966. Essai d’identification des constituants de l’arôme des vins de Vitis vinifera L. Thèse 3e cycle, Faculté des Sciences, Bordeaux.
BRUNELLE R. L., 1967. Evaluation of gas liquid chromatography for the determination of the fusel oil in distilled spirits. J. Ass. Off. Agr. Chemists, 46 (2), 294-297.
BRUNELLE R. L., 1968. Collaborative study of the quantitative determination of fusel oil and the ethyl acetate by gas liquid chromatography. J. Ass. Off. Agr. Chemists, 51 (4), 915-921.
DAY W. C., ERDMAN J. G., 1963. Science, 141, 808.
DEMOLE E., BERTHET D., 1971. Helv. Chim. Acta, 54, 681.
DEMOLE E., ENGGIST P., SAUBERLI U., 1970. Helv. Chim. Acta, 53, 541.
DEMOLE E., BERTHET D., 1972. Helv. Chim. Acta, 55, 1866.
Dubois P., PARFAIT A., DEKIMPE J., 1973. Présence de dérivés de l’acroléine dans un rhum à goût anormal. Ann. Technol. agric., 22 (2), 131-135.
FLANZY M., JOURET C., 1963. Contribution à l’étude des eaux-de-vie d’Armagnac par chromatographie en phase vapeur. Ann. Technol. agric., 12 (1), 39-50.
GUYMON H., 1970. Composition of California commercial brandy distillates. Annual meeting of the american Society of Enologists, Coronado, Calif, June 25-27.
HRUZA D. E., VAN PRAAC M., 1974. Isolation and identification of the components of the tar of hickory wood smoke. J. agric. Food Chem., 22 (1), 123-126.
JOURET C., MOUTOUNET M., 1968. Comparaison des alcools supérieurs des eaux-de-vie d’Armagnac et des whiskies. Ann. Technol, agric., 17 (2), 151-158.
KAHN J. H., BLESSINGER E. T., 1972. Collaborative study of the quantitative gas liquid chromatographic determination of the fusel oil and other components in whisky. J. Ass. Off. Agr. Chemists, 55 (3), 549-556.
LIEBICH H. M., KOENIG W. A., BAYER E., 1970. Analysis of the flavor of rum by gas liquid chromatography and mass spectrometry. J. Chromat. Sci., 8, 527-533.
LA ROE E. G., SHIPLEY P. A., 1970. J. Agr. Food Chem., 18 (1), 174-175.
MASCRE, 1957. Ann. pharm. franç., 15, 421.
MULIK J. D., ERDMAN J. G., 1963. Science, 141, 806.
OTSUKA K., ZENIBAYA Y., Iron M., 1974. Presence and significance of two diastereomers of B-methyl y-octalactone in aged distilled liquors. Agr. Biol. Chem., 38 (3), 485-490.
PORCARO P. J., JOHNSTON V. D., 1961. Analyst Chem., 33, 361.
SCHOENEMAN R. L., DYER R. H., 1973. Analytical profile of scotch whiskies. J. Ass. Off. Agr. Chemists, 56 (1), 1-10.
SCHREIER P., DRAWERT F., 1974. Gaschromatographisch-Massenspektrometrische Untersuchung fluch tiger Inhaltsstoffe des Weines. Unpolare Verbindungen des Weinaromas. Z. Lebensm.-Unters. Forsch, 154, 273-278.
SIHTO E., NYKANEN L., SUOMALAINEN H., 1964. Gas chromatography of the aroma compounds of alcoholic beverages. Qual. Plant. Mat. veg., 11 (2-4), 211-228. SINGER D. D., STILES J. W., 1965. The determination of higher alcohols in potable spirits : comparison of colorimetric and gas chromatographic methods. Analyst, 90, 290-296.
SINGER D. D., 1966. The proportion of 2-methyl-butanol and 3-methyl-butanol in some brandies and whiskies as determined by direct gas chromatography. Analyst, 91, 790-794.
VAN DER KLOOT A. P., TENNEY R. T., BAVISSOTTO V., 1958. Proc. Amer. Brew., 96.
WEBB A. D., KEPNER R. E., 1961. Fusel oil analysis by means of liquid partition chromatography.
Amer. J. Enol. Vit., 12, 51-69.
WINTER M., ENGGIST P., 1971. Helv. Chim. Acta, 1891.