The Modification of Certain Constituents of Flavourings after Addition to Alcoholic Beverages

Follow along: IG @birectifier

Lets go back in time 40 years and take a look at another chapter from the ultra rare text Flavour of Distilled Beverages edited by J.R. Piggott. This chapter is written by two scientists from Martini & Rossi and concerns changes to botanical flavor due to reactions with ethanol & water as well as acidity. The chemistry gets dizzyingly heavy and there is no candid discussion of why they are interested (within the context of vermouth production) or what they are going to do with the findings.

It was an era of new analysis tools like GCMS where they could finally examine some of this stuff. The chemistry was also going to relate to future opportunities for cleaning products related to botanicals; think citrus degreasers… Those products may have descended from these first chemical inquiries. At the same time, Martini & Rossi were achieving global distribution and likely had concerns about consistency and shelf stability of products they made across various portfolios (more than just vermouth!). And then there may be the potential to harness some of these modifications as a production feature.

I remembered an old note about gin production where certain botanicals could be treated with an acid pre-distillation to drive a transformation as a feature. So the botanical may be infused in 60% ethanol at pH 2 before distillation. What botanical would we do that to? And what brand am I thinking of that may have used it?

Other notes I’ve seen over the years use a similar strategy but they are transforming a botanical to be used for casing tobacco. The transformation is dramatic as far as it is described; I’ve never experienced it. Something not associated with smoke becomes smokey… I’ve intended to try it many times but have failed to arrange it.

On a practical level, I’m not sure what we can easily take away from this incredibly dense chapter. Infusion with pH as low as 3, such as Vermouth, may effect a botanical significantly and we can’t easily say whether positively or negatively. Some citrus products may encounter off aroma due to degradation over significant durations; I’m thinking decades old triple-secs I’ve encountered having an odd off character that likely developed over time, but concerns may be as simple as bottled citrus soda aged 6 months like a San Pellegrino Limonata.

If we wanted practical knowledge regarding the most acid or alcohol impacted botanicals what would be do? We could take 20-30 potential vermouth botanicals and perform infusions with neutral spirits at varying ABV’s and acidities. A pH like 2.0 could be chosen to hunt for dramatic transformations (you cannot drink it!), while pH 3.0 may represent something closer to wine and then finally something nearly neutral or unbuffered like 60% spirit. Don’t expect a miracle, but it may be interesting to add a face to all the weird chemistry below. Are Star anise and Angelica the most reactive? Have any of the new indie vermouth producers been around long enough to have opinions on this? Do any of the 3,000 new gin producers acid treat a botanical before distillation?

In the last decade, if we really added thousands of new alcoholic beverage producers, was every one of them really walled off from unique institutional knowledge about working with botanicals? Any prior careers? No one from the tobacco business bringing over a technique? No one from deep within the flavoring house world? Is there anything out there besides basic infusions and ripping botanicals through a still?

Chapter 12

The Modification of Certain Constituents of Flavourings after Addition to Alcoholic Beverages

P.A.P. Liddle, and A. Bossard
Martini & Rossi, B.P. 50, 19, avenue Michelet, 93401 Saint-Ouen Cedex, France

INTRODUCTION

The production of flavoured alcoholic beverages has traditionally involved the use of alcohol itself to extract the aroma and flavour compounds of plants, and the same principle continues to be used to a large extent today. This is in contrast to the rest of the food industry, where these compounds are more generally obtained by water-distillation, as essential oils, or by extraction with various solvents (including ethanol) and subsequent removal of the solvents to give oleoresins.

In the case of alcoholic beverages (Fig. 1), the extraction is usually carried out at ambient temperatures with 5 to 20 % (w/v) of the plant material in 50 to 90 % aqueous ethanol, sometimes for several weeks. The resulting extracts are filtered and then blended with other ingredients (wine, sugar, spirits, colouring, etc…), or sometimes distilled before being incorporated, depending on the type of final product.

These extracts generally have a certain acidity, and in addition are often used in wine-based products or those containing added, acids, so that the final pH may be as low as 3. Under these conditions, it is only to be expected that some of the essential-oil constituents of the original plant material might undergo modification by reaction with ethanol and/or water.

A great deal of work on this subject has been carried out by Taskinen [1], who studied the aqueous alcoholic distillates of marjoram [2], coriander [3], angelica root [4] and juniper berry [5], and in particular the formation of ethyl ethers of terpene alcohols in these distillates and in model solutions [6]. We have already reported some observations on these modifications [7,8] and a study has been carried out by Tangel [9] on the ageing of an aqueous-alcoholic extract used in the manufacture of ginger ale.

Figure 1 The production of flavoured alcoholic beverages.

In this work, we shall be considering the possible modifications which different classes of plant constituents, principally these considered as “volatile”, undergo in aqueous-alcoholic solutions.

EXPERIMENTAL

As a general method, analyses of plant extracts, model systems, or finished products were carried out after a simple one-step extraction [10].

After addition of a suitable internal standard, and if necessary dilution of the sample to an alcoholic strength of less than about 20 % v/v, 10 to 20 ml of the sample containing 1 g sodium chloride were extracted for 1 minute with 1 ml iso-octane (2,2,4-trimethylpentane) in a large test-tube on a “Vortex” type mixer. After separation of the organic phase, the latter was analysed directly by splitless injection of about 1 microlitre on to a fused silica column coated with Carbowax 20M, held at 60°C for 2 minutes, then programmed at 5°C/min up to 220°C. Identifications were carried out on a Hewlett-Packard 5992A GC/MS system, aided by calculations of linear retention indices (1.r.i.) relative to normal hydro-carbons [11].

Experiments with model solutions were carried out with 0.01-0.1% solutions of the pure compounds in 20-60% v/v aqueous ethanol, at room temperature, acidified to pH 3 with citric acid. As an aid to mass-spectral studies on reaction products of some monoterpene alcohols, solutions of hexadeuterated ethanol and deuterium oxide, acidified with deuterated hydrochloric acid, were used.

In general, gas-chromatographic (1.r.i.) and mass-spectral details of compounds described below have not been given, due to lack of space, but these may be obtained from the authors.

RESULTS AND DISCUSSION

(a) Acids

The formation of ethyl esters from fatty acids and aromatic acids in aqueous-ethanolic extracts is perhaps the most obvious and certainly the most important modification from a quantitative point of view. The ethyl esters of both volatile and non-volatile acids have been known for a long time in the chemistry of wines and spirits. Two papers in recent years [12,13] have discussed the kinetics of this reaction in wines.

The principal ethyl esters found in the aqueous-ethanolic flavourings commonly used are those derived from the plant triglycerides (saturated even fatty acids from C14 to C18, and the group of unsaturated C18 fatty acids), along with the aromatic acids benzoic, cinnamic and 3-phenylpropionic. An interesting study of the fatty-acid composition of the lipids of 18 common herbs and spices has been carried out by Salzer. [14]. The amounts of fatty-acid ethyl esters found in some alcoholic extracts after 10 days are shown in Table I, and as an example a chromatogram of an aqueous-alcohol extract of nutmeg is shown in Fig. 2. It can be seen that the levels of ethyl esters in such extracts can be at least of the same order as those of normal essential-oil constituents.

Table 1— Amounts (in mg/l) of ethyl esters formed in aqueous-alcoholic extracts (60% v/v alcohol) of some herbs and spices (10% w/v) after 10 days

and spices (10% w/v) after 10 days

As an interesting parallel, Heikes and Griffitt [15] have demonstrated the presence in spices and food products of the esters formed from fatty acids and 2-chloroethanol [ethylene chlorohydrin], the latter being produced by combination of the fumigant ethylene oxide with moisture and inorganic chloride.

(b) Carbonyl compounds

The formation of acetals from aldehydes and ethanol is a topic that has been surrounded in the past by a certain amount
of controversy. Acetals of lower aliphatic carbonyl compounds are found in spirits, and there has been much discussion as to whether these compounds exist as such, or are formed during solvent extraction and concentration. However, Williams and Strauss [16] have shown that although small but significant  conversion to acetals (of the order of a few percent) occurs when extractions are performed at pH 1.5, this was much less at pH 3.5, and at pH 5.0 no acetal formation was observed.

As far as plant materials are concerned, it is quite common to find that saturated aliphatic aldehydes such as those found in citrus fruits are partly converted to their diethyl acetals in alcoholic beverages. The amounts increase with the alcohol concentration, but are less at lower pH, where the equilibrium between the acetals and the original compounds is shifted in favour of the latter. The acetal/aldehyde ratios are also significant for benzaldehyde (0.1 to 0.15) in spirits based on stone containing fruits, such as “Kirsch”. Other aromatic acetals occasionally found are those of cinnamaldehyde and cuminaldehyde (for example, in extracts of cinnamon and cumin).

Figure 2—Chromatogram of an aqueous-alcoholic extract of nutmeg(10% w/v in 60% v/v ethanol/water), iso-octane extraction as in the text. 1: terpinen-4-ol ethyl ether, 2: terpinen-4-ol, 3: safrole, 4: ethyl tetradecanoate, 5: ethyl hexadecanoate, 6: myristicine, 7: ethyl 9-octadecenoate, 8: ethyl 9, 12-octadenadienoate.

Turning to terpene carbonyl compounds, one obviously thinks of citral. Extensive studies of the acid-catalysed cyclisation of this compound have been made by several authors [17-21], and particularly by McHale et al. [22]. Given the acidic nature of most citrus-based alcoholic beverages, it is hardly surprising that citral is rarely found in such finished products. It is, however, possible to find some of the compounds identified by the above authors, such as the menthadienols, menthenediols, p-cymen-8-ol, etc… as well as the ethyl ether of the latter. We have also found considerable amounts of 4-methylacetophenone in such model solutions stored for a long time.

The ethyl ether of p-cymen-8-ol has also been reported in an aqueous acidic solution of citral [19], to which had been added a small amount of ethanol, presumably to obtain a single-phase system. As the authors pointed out, this ether was not found when ethanol was omitted from the system. However, in  another study [20], other ethoxy compounds were identified as minor components in a similar system, and it was suggested that these compounds contributed to the off-flavour of deteriorated lemon, whereas it would appear that they may be artefacts of the experimental procedure.

The common terpene ketones appear to be relatively unreactive, although there is some evidence that pulegone forms small amounts of an ethoxy compound, perhaps due to the ease of enol formation in this compound via protons of the isopropylene group adjacent to the carbonyl function.

In parallel, the formation of acetals in flavour systems based on other alcohols is to be expected. Heydanek and Min [23] demonstrated such reactions in a propylene-glycol based cherry flavour and in model solutions based on the same diol. In wine and wine-based alcoholic beverages, the presence of glycerol and 2,3-butanediol gives futher possibilities for acetal formation, such as the reaction between glycerol and acetaldehyde, giving rise to isomeric 1,3-dioxanes and 1,3- dioxolanes often found in sherry-type wines [24-26]. Reactions between glycerol and aliphatic aldehydes such as octanal also give rise to such compounds in model solutions and these could be expected in citrus-flavoured wine products.

(c) Alcohols

As mentioned earlier, a great deal of work on the formation of ethyl ethers of terpene alcohols in alcoholic distillates and in model solutions has been carried out by Taskinen. The ethyl ethers of terpinen-4-ol and of the cis and trans isomers of 2-p-menthen-1-ol, piperitol, verbenol and sabinene hydrate (4-thujanol) were identified by the latter in distillates of various plants. In model solutions, the 2-p-menthen-1-ols and the piperitols were each found to isomerise, giving a mixture of both alcohols and their respective ethers, and the verbenols gave predominantly the corresponding trans ether. The sabinene hydrates, however, besides giving their respective ethers, also gave rise to terpinen-4-ol and its ether, along with alpha- and gamma-terpinene. Terpinen-4-ol itself, however, was found to be unreactive, as were the acylic allylic alcohols linalool and geraniol, although the reaction time was only 6 hours. The ethers of the latter were not found in flavour distillates, and it was suggested that the presence of terpinen- 4-ol ethyl ether could be explained by formation from the sabinene hydrates, and possibly from sabinene by analogy with the rather well-known acid-catalysed hydration of the latter [27-29].

Figure 3 – Mass spectra (EI) of linalool (A) and linalyl ethyl ether (B).

Our own interest in the formation of ethyl ethers arose from the identification of linalyl ethyl ether (1.r.i. 1300 on Carbowax 20M) in a number of finished products. The mass spectrum of this compound (Fig. 3), which we have synthesized, resembles that of the parent alcohol, apart from the ion at m/z = 99, arising from the replacement of the hydroxyl group by an ethoxyl in the m/z = 71 ion of the free alcohol. The m/z = 99 ether ion undergoes alpha-cleavage and transfer of a beta-hydrogen to give the “original” m/z = 71 ion (which became m/z = 72 in the case of the ether prepared in a deuterated system). The rate of formation of this ether is slow, about 5% conversion after two months in 60 % v/v ethanol at pH 3, but it can be found in finished products containing linalool at low pH, including wines produced from Muscat grapes, which contain appreciable amounts of this alcohol.

Figure 4- Chromatograms of thyme oil (4-thujanol/terpinen-4-ol type) before (A) and after (B) several months in 60% v/v ethanol at pH 3. 1 linalyl ethyl ether, 2 terpinen-4-ol ethyl ether, 3 alpha-terpineol ethyl ether, 4 trans-4-thujanol, 5 geranyl ethyl ether, 6 lynalool, 7 cis-4-thujanol, 8 terpinen-4-ol, 9 alpha-terpineol, 10 thymol, 11 new compounds, as yet unidentified.

Our model experiments with trans-sabinene hydrate gave similar results to those of Taskinen, after the same reaction time (6 hours). However, after a longer period, many of the ethers disappeared, giving rise almost exclusively to terpinen- 4-01. The trans-sabinene hydrate used was isolated from a sample of a type of thyme oil (kindly provided by Dr. J. Passet of Montpellier) containing a high percentage of this compound [30]. Fig. 4 shows chromatograms of this oil, before and after several months in 60 % v/v ethanol at pH 3. The sabinene hydrates have virtually disappeared, with a notable increase in the proportion of terpinen-4-ol and of its ether (1.r.i. 1370). Terpinen-4-ol itself, as observed by Taskinen, was found to be unreactive even after several months in 60 % v/v ethanol at pH 3, and alpha-terpineol gave only minor amounts of an unidentified ether under these conditions (1.r.i. on Carbowax 20M greater than that of the original alcohol), presumably by intermediate rearrangement reactions.

Figure 5 – Mass spectra (EI) of alpha-terpineol (A) and its ethyl ether (B).

In contrast to these non-allylic alcohols, the allylic alcohols nerol and geraniol were found to react, albeit slowly, in the same way as linalool, to give ethers with 1.r.i. of 1440 and 1470 respectively. The mass spectra of these ethers (MW 182) closely resembled those of the starting alcohols, which is to be expected, as it has been shown that the majority of the fragments in the spectra of these compounds consist of the hydrocarbon skeleton only. The identification is thus tentative, but is supported by the fact that each alcohol gives both ethers, along with linalool and its ether, plus alpha-terpineol and an ether which, again by analogy, would seem to be that of the latter alcohol (Fig. 5).

All these products could be envisaged by the sort of scheme shown in Fig.6, although it has been shown that the
reactions may not be governed by the classical charge-localized ions shown [27]. Monoterpene hydrocarbons such as myrcene, limonene and gamma-terpinene were also among the products. A better idea of the probable mechanisms involved may be gained from an extensive study by Baxter et al. [31] of the reactions of linalool, geraniol, nerol and their acetates in aqueous citric acid.

Figure 6 – Monoterpene alcohols and their ethers.

The high reactivity of sabinene hydrate prompted us to look at the behaviour of another alcohol with a cyclopropane ring in the beta-position to the hydroxyl group, chrysanthemyl alcohol. Although the ether of this alcohol can be readily prepared by Williamson synthesis (unlike those of alpha-terpineol and terpinen-4-ol), it was not found among the products of reaction of chrysanthemyl alcohol in a model solution. Instead about 10-15 % of the alcohol was converted to other “irregular” monoterpenes : artemisia triene, yomogi and artemisia alcohols and their ethers. These irregular mono-terpenes are of interest in flavoured alcoholic beverages, as they occur in several of the plants traditionally used by the vermouth and liqueur industry.

Other cyclic allylic alcohols, e.g. carveol, reacted fairly rapidly, while other acyclic allylic alcohols also reacted, including sesquiterpenes such as farnesol and nerolidol, but at a much slower rate, nevertheless giving yields of 10-20 % after a period of six months. As observed by Taskinen, other non-allylic alcohols, e.g. citronellol, menthol, did not react, or gave very small amounts of products other than their corresponding ethers.

Some aromatic alcohols also appear to be reactive. Cinnamy] alcohol, for example yielded 25 % of an ether after 6 months, and p-cymen-8-ol ethyl ether can be found in citrus-based products, as mentioned earlier; however, we were unable to check whether it is formed from the alcohol directly, due to the lack of a pure sample of the latter. The ethyl ethers of anisyl, vanillyl, and 4-hydroxybenzyl alcohols can be found in vanilla extracts [32-33], which represent perhaps one of the main uses of ethanol as a solvent outside the beverage industry.

Before leaving the subject of ether formation, mention should be made of a recent study of the volatile constituents of fresh lemon juice [34]. Among the minor compounds identified were several ethyl and methyl ethers of terpene alcohols. Although small amounts of ethanol and, to a lesser extent, methanol, have been reported in citrus juices, it is probable that the addition of ethanol to the fresh juice (to avoid the enzymatic formation of off-flavours during the extraction of the volatiles) contributed to the ether formation under the acidic conditions.

The study also gives some flavour descriptors for the ethers identified, along with their threshold values in water. These range from 0.2 ppm (p-cymen-8-yl ethyl ether) to 5 ppm (linalyl ethyl ether), although the presence of an ether group was found to alter the organoleptic character significantly only in the case of the monocyclic terpenes.

(d) Esters

The behaviour of the common monoterpene esters in aqueous ethanol at low pH would be expected to be analagous to that described in the study already mentioned by Baxter et al.[31]. Indeed these esters are rarely found in finished alcoholic beverages. As an example, a solution of neryl acetate in 60% v/v aqueous ethanol at pH 3 contained only traces of the initial compound after several months, the main products being limonene (2 %), linalool (7 %), nerol (60 %), geraniol (1%) and their respective ethers (4,7 and 3 %), and alpha-terpineol (8%).

(e) Hydrocarbons

The hydration of terpene hydrocarbons, accompanied by cyclisation and acyclisation reactions, has been studied fairly extensively. The acid-catalysed hydration of sabinene, for example, was studied at the beginning of the century by Wallach [28, 29], the main oxygenated compounds formed being the sabinene hydrates and terpinen-4-ol, the latter predominating [27]. Production of ethers as well as alcohols in aqueous-ethanol is to be expected, along with simple isomerisation and oxidation, e.g. to p-cymene, the latter reactions again having been studied for sabinene [35]. The disappearance of beta-pinene from solutions of lemon oil in acidic aqueous ethanol was observed by Tangel [9], while Watanabe [36] has studied the reaction products of pure beta-pinene under similar conditions. The main products were alpha-terpineol and its ethyl ether, the latter being, as expected, absent in acidic solutions of beta-pinene which did not contain ethanol.

Of the monoterpene hydrocarbons, the terpinenes are often found in the types of plants used in alcoholic beverages, along with the product of their dehydrogenation, p-cymene. In model solutions at pH 3, gamma-terpinene was found to be fairly stable over a period of several weeks, showing only slight formation of p-cymene. The alpha-isomer appeared to be slightly less stable, as reported already [35], giving small amounts of p-cymen-8-ol and its ether among the reaction products, and a few percent of two isomeric compounds of molecular weight 198 [with a base peak in the mass spectrum at m/z = 140], which may arise from the addition of ethanol to an intermediate epoxide.

(f) Epoxides

The acid catalysed solvolysis of epoxides in aqueous ethanol might be expected to occur at the pH levels encountered in alcoholic beverages, with the possibility of formation of adjacent ethoxy and hydroxy groups as well as the expected glycols. Three examples of such reactions are briefly given below.

The first concerns a pair of compounds identified in Artemisia absinthium L., the cis and trans epoxyocimenes [37]. After a few months in 60 % v/v ethanol, at pH 3, these compounds were found to have disappeared, giving a fairly wide range of products, of which the most important were two isomeric compounds of molecular weight 180, which could be accounted for by loss of water from the ethoxyl/hydroxyl compound formed by opening of the epoxide ring.

The second example concerns a compound which has been the subject of considerable study, caryophyllene oxide. After several months at pH 3 in 60 % v/v ethanol, only a few percent of the original compound remained, the main products being an isomer (MW 220), and two compounds of molecular weight 266 having very distinctive mass spectra with base peaks at m/z = 113 and 207, respectively. These last two compounds presumably arise from addition of a molecule of ethanol to the original compound, but given the cyclization-rearrangement reactions which this compound is known to undergo in aqueous acid [38, 39], further work would be necessary to establish the structures of the reaction products.

The final example concerns a compound widely used for the production of anise-flavoured aperitifs, trans-anethole. Samples of this compound contain small amounts of the cis-isomer (normally less than 0.5 %), together with similar levels of the oxidation products p-anisaldehyde and the so-called anis-ketone (4-methoxyphenyl-acetone). The latter is assumed to arise via epoxidation of the double bond of anethole [40]. Routine analyses are necessary to check on the levels of these compounds, and also on the trace amounts of sesquiterpenes, which may give an indication of the origin of the anethole.

When using ethanol as a solvent in routine GLC examination of anethole samples, it was noticed that solutions analysed a few days after having been made up exhibited two additional peaks in the latter part of the chromatogram (Fig.7). The same two peaks could be found in finished products, which generally have a pH of around 7. The amounts of the additional compounds were found to increase with decreasing pH and alcoholic strength.

Figure 7— Iso-octane extract of a sample of pastis. 1 methyl undecanoate (internal standard 10 mg/1), 2 cis-anethole, 3 trans-anethole, 4 p-anisaldehyde.

The mass spectra of the two compounds were almost identical (base peak at m/z = 165, with a peak of 80 % at m/z = 137), and indicated a molecular weight of 210, which would correspond to the addition of a molecule of ethanol to an intermediate epoxide, to give the products shown in Fig.7. This is supported by the infra-red spectrum of the isolated products, and by the formation of a TMS ether. Studies of anethole in methanol and in deuterated ethanol have shown that the compounds are enantiomers of the form containing the ethoxy group on the carbon alpha to the aromatic ring, although trace amounts of the other isomers, as well as of the corresponding glycol, were also found as reaction products.

CONCLUSIONS

From the analytical chemist’s point of view, it is important to bear in mind that the analysis of flavoured alcoholic beverages may reveal the presence of compounds which were absent in the raw materials used in their production and the parallel decrease or even disappearance of compounds initially present. The effect of these reactions on the quality of the finished product is difficult to assess, as flavoured alcoholic beverages are usually even more complex than the initial wines and spirits.

Such modifications are also to be expected when ethanol (or, for that matter, other alcohols such as methanol, isopropanol, etc..) is used as a solvent in an analytical procedure, or on an industrial level in the production of plant extracts.

ACKNOWLEDGMENTS

We would like to acknowledge the contribution to much of this work by our late colleague, Pierre De Smedt.

REFERENCES

1. Taskinen, J. Thesis, Helsinki University of Technology, (1977).

2. Taskinen, J. Composition of the essential oil of sweet marjoram obtained by distillation with steam and by extraction and distillation with alcohol-water mixture. Acta Chem. Scand. 1974, 28, 1121-28.

3. Taskinen, J.; Nykanen, L. Volatile constituents obtained by the extraction with alcohol-water mixture and by steam distillation of coriander fruit. Acta Chem. Scand. 1975, 29, 425-29.

4. Taskinen, J.; Nykanen, L. Chemical composition of angelica root oil. Ibid, 757-64.

5. Taskinen, J.; Nykanen, L. Volatile constituents of an alcoholic extract of juniper berry. Int. Flav. Food Add. 1976, 7 (5), 228 & 233.

6. Taskinen, J. The acid catalysed reaction of some monoterpene alcohols in aqueous ethanol. Ibid, 235-36.

7. De Smedt, P.; Liddle, P.A.P. Modification of certain compounds of aromatic plants in an aqueous-alcohol medium. Ann. Fals. Exp. Chim. 1976, 69 (747), 965-73.

8.  Liddle, P.A.P.; De Smedt, P. Modification of certain essential-oil constituents in aqueous ethanol media. 8th International Congress of Essential Oils, Cannes, France, October 1980.

9. Tangel, F.P. Determination of the aging mechanism of ginger extract used in the manufacture of ginger ale. Thesis, Rutgers University, (1979).

10. Liddle, P.A.P.; Bossard, A. Glass capillary gas chromatography in the wine and spirit industry. Instrumental analysis of Food and Beverages (Charalambous, G., Ed.), Academic Press, 1983 (in press).

11. Van den Dool, H.; Kratz, P.D. A generalization of the retention index system including linear temperature programmed gas-liquid partition chromatography. J. Chromatogr. 1963, 11, 463-471.

12. Shinohara, T.; Shimizu, J.; Shimazu, Y. Esterification rates of main organic acids in wines. Agric. Biol. Chem. 1979, 43, 2351-2358.

13. Ramey, D.D.; Ough, C.S. Volatile ester hydrolysis or formation during storage of model solutions and wines. J. Agric. Food Chem. 1980, 28, 928-934.

14. Salzer, U.-J. Fatty acid composition of lipids of some spices. Fette, Seifen, Anstrichmittel, 1975, 77 (11), 446-50.

15. Heikes, D.L.; Griffitt, K.R. Mass spectrometric identification and gas-liquid chromatographic determination of 2-chloroethyl esters of fatty acids in spices and foods. J. Assoc. Off. Anal. Chem. 1979, 62, 786-791.

16. Williams, P.J.; Strauss, R. 3,3-Diethoxybutan-2-one and 1,1,3-triethoxypropane: acetals in spirits distilled from Vitis vinifera grape wines. J. Sci. Food Agric. 1975, 26, 1127-1136.

17. Baines, D.A.; Jones, R.A.; Webb, T. C.; Campion-Smith, I.M. The chemistry of terpenes-I. Tetrahedron, 1970, 26,  4901-4913.

18. Clark, B.C.; Powell, C.C.; Radford, T. The acid catalysed cyclization of citral. Tetrahedron, 1977, 33, 2187-91.

19. Iwata, I.; Yamamoto, H. Deteriorated products of citral in citric acid solution. J. Jap. Soc. Food Nutr. 1978, 31, 389-94.

20. Kimura, K.; Doi, E.; Nishimura, H.; Iwata, I. Degradation products of citral by acid. Nippon Nogeikagaku Kaishi 1981, 55, 1073-1079.

21. Kimura, K.; Iwata, I.; Nishimura, H. Relationship between acid-catalysed cyclisation of citral and deterioration of lemon flavour. Agric. Biol. Chem. 1982, 46, 1387-1389.

22. McHale, D.; Laurie, W.A.; Baxter, R.L. A. reappraisal of the acid-catalysed cyclization of citral. 7th International Congress of Essential Oils, Kyoto, Japan, October 1977.

23. Heydanek, M.J.; Min, D.B.S. Carbonyl-propylene glycol interactions in flavor systems. J. Food Sci. 1976, 41, 145-147.

24. Williams, P.J.; Strauss, C.R. Studies of volatile components in the dichloromethane extracts of Australian flor sherries: the identification of the isomeric ethylidene glycerols. J. Inst. Brew. 1978, 84, 144-147.

25. Muller, C.J.; Kepner, R.E.; Webb, A.D. 1,3-dioxanes and 1,3-dioxolanes as constituents of the acetal fraction of Spanish fino sherry. Am. J. Enol. Vitic. 1978, 29, 207-212.

26. Etievant, P. Volatile constituents of a Sherry wine (vin jaune du Jura). Identification of acetals arising from glycerol. Lebens.-Wiss.u.-Technol. 1979, 12, 115-120.

27. Norin, T.; Smedman, L.A. The cyclopropylcarbinyl rearrangements in sabinene and alpha-thujene. Acta Chem. Scand. 1971, 25, 2010-2014.

28. Wallach, O. Terpenes and essential oils. Justus Lieb. Ann. Chem. 1906, 350, 141-179.

29. Wallach, O. Ibid, 1908, 360, 82-100.

30. Granger, R.; Passet, J.; Pinede, M.C. Trans-4-thujanol and terpinen-4-ol in Thymus vulgaris L. C.R. Acad. Sc. Paris, série D 1968, 267, 1886-1889.

31. Baxter, R.L.; Laurie, W.A.; McHale, D. Transformations of monoterpenoids in aqueous acids. Tetrahedron 1978, 34, 2195-99.

32. Shiota, H.; Itoga, K. Study on the aromatic components of vanilla beans. Koryo 1975, 113, 65-71.

33. Galetto, W.G.; Hoffman, P.G. Some benzyl ethers present in the extract of vanilla. J. Agric. Food Chem. 1978, 26, 195-97.

34. Mussinan, C.J.; Mookherjee, B.D.; Malcolm, G.I. Isolation and identification of the volatile constituents of fresh lemon juice. Essential Oils (Mookherjee, B.D.; Mussinan, C.J., Eds), Allured Publishing Corp., Wheaton, 1981, pp. 199-228.

35. Wrolstad, R.E.; Jennings, W.G. Chromatostrip isomerization of terpenes. J. Chromatog. 1965, 18, 318-324.

36. Watanabe, A.; Iwata, I. Deteriorated products of beta-pinene in citric acid solution. J. Jap. Soc. Food Nutr. 1980, 33, 305-308.

37. Chialva, F.; Doglia, G.; Gabri, G.; Aime, S.; Milone, L. Isolation and identification of cis- and trans-epoxy- ocimenes from the essential oil of Artemisia absinthium. Riv. Italiana E.P.P.O.S. 1976, 58, 522-525.

38. Aebi, A.; Barton, D.H.R.; Lindsey, A.S. J. Chem. Soc. 1953, 3124

39. Warnhoff, E.W. New rearrangements of the caryophyllene skeleton. Can. J. Chem. 1964, 42, 1664-1675.

40. Kraus, A.; Hammerschmidt, F.J. Research on fennel oils. Dragoco Report 1980, 2, 31-41.