1 (eingereicht JSFA, bitte momentan vertraulich behandeln)
2 Behaviour of thujone during distillation and
3 possible concentration ranges in pre-ban
4 absinthe
5 Running title: Thujone distillation
6
7
8 Dirk W Lachenmeier* and Thomas Kuballa
9 Chemisches und Veterinäruntersuchungsamt (CVUA) Karlsruhe, Weißenburger Str. 3,
10 D-76187 Karlsruhe, Germany
11 *Corresponding author. E-mail: [e-mail]
12
1
12 ABSTRACT
13 A discrepancy in the magnitude of thujone concentrations in pre-ban absinthe has
14 existed until now. Concentrations of 260 mg L-1 were derived at by theoretical
15 calculations. Tests of authentic pre-ban absinthes and studies about absinthes produced
16 according to historic recipes found concentrations below 10 mg L-1. In this study, the
17 behaviour of thujone during distillation was studied and a significant discrimination was
18 determined (80% yield in water-cum-steam distillation). The thujone concentrations in
19 pre-ban absinthe were then calculated under regard of the composition of wormwood oil
20 derived by a literature review. A typical Absinthe Suisse de Pontarlier might have
21 contained 22Ä…26 mg L-1 of thujone. It was proven that the previous calculations
22 overestimated the thujone content and the discrepancy was resolved as our new
23 calculations are in good accordance to the experimental findings.
24 Key words: Thujone; absinthe; absinthism; Artemisia absinthium L.; wormwood;
25 distillation
26
27 INTRODUCTION
28 Recent review of absinthism syndrome shows a discrepancy in the thujone contents of
29 pre-ban absinthes.1 One evidence shows that absinthes produced according to historic
30 recipes contained only relatively low concentrations of thujone (mean: 1.3 Ä… 1.6 mg L-1,
31 range: 0 - 4.3 mg L-1).2 Concentrations below 10 mg L-1 were also found in a number of
32 tests of vintage absinthes.2-4
33 On the other hand, another theory brought up in 1992 by Arnold shows the thujone
34 content of pre-ban absinthe as high as 260 mg L-1.5 This theory was based on the
2
35 assumption that 100 l of pre-ban absinthe employed 2.5 kg of dried Artemisia
36 absinthium L. (1.5% oil, of which 67% is thujone; corresponding to 251 mg L-1 of
37 thujone in the final product) and 1 kg of dried Artemisia pontica L. for coloration
38 (0.34% oil, of which 25% is thujone, corresponding to 9 mg L-1 of thujone in the final
39 product). The value of 260 mg L-1 is often cited in the newer literature, e.g. Refs. 3,6-10.
40 Some badly researched articles misleadingly refer to the value of 260 mg L-1 as if it was
41 derived by analyses of the 19th century and not by theoretical calculations in the
42 1990s.8,11 For example, Strang et al.6 erroneously cite Duplais 1855 book12 giving the
43 thujone content of absinthe. In fact, Duplais recipes merely give the wormwood content
44 used for absinthe making, and there is no mention of thujone (or any other terpene) in
45 both volumes of Duplais work. Moreover, the exact composition of wormwood oil was
46 unknown in Duplais time.
47 The mentioned calculations of a 260 mg L-1 content assumed that the total amount of
48 thujone would be recovered in the final product. The following three points were given
49 by Arnold to support his calculation of relatively high concentrations. First, by adding
50 water to the first decoction before heating, a type of "steam-distillation" was achieved
51 wherein the amount of any constituent distilled over depends on both its vapour pressure
52 and molecular weight. In this way the effect of a low vapour pressure for a particular
53 compound may be counteracted to some extent by its high molecular weight relative to
54 that of water. Second, the distillation head of the industrial apparatus was simple and
55 little attempt was made to restrict carry-over by aerosol entrainment. And third, the
56 purpose of the secondary extraction at moderate temperature was two-fold  to achieve a
57 green coloration and to add additional flavour.5 However, experimental proof to validate
58 these considerations is lacking so far. It cannot be totally disregarded that during
59 distillation a discrimination of thujone occurs.
3
60 In this study, the behaviour of thujone during distillation is determined for the first time.
61 The results are used to critically judge the possible thujone contents of pre-ban absinthes
62 under regard of a number of historical recipes and the discrepancy between theoretical
63 and experimental values is tried to resolve.
64 EXPERIMENTAL
65 1-Thujone and cyclodecanone were purchased from Fluka (Buchs, Switzerland).
66 Ethanol (>99.9% vol) was obtained from Merck (Darmstadt, Germany). The standard
67 distillation still was purchased at TGP (Karlsruhe, Germany). The direct steam
68 distillation was accomplished with the Vapodest 30 (C. Gerhardt, Fabrik und Lager
69 chemischer Apparate, Bonn, Germany).
70 For the distillation experiments, a model solution containing 260 mg L-1 2341-thujone in
71 ethanol (85% vol) was prepared. 500 mL aliquots of the model solution were distilled
72 directly (ethanolic distillation) or after dilution with 237 mL of water (water-cum-steam
73 distillation). Additional 500 mL aliquots were distilled using direct steam distillation.
74 During all distillations, 50 mL fractions were collected. All experiments were conducted
75 in duplicate.
76 The alcoholic strength of the distillates was determined using the reference method
77 applying oscillation-type densimetry as prescribed in Commission Regulation (EC) No.
78 2870/2000.13 The thujone content was determined using liquid-liquid-extraction with
79 1,1,2-trichloro-1,2,2-trifluoroethane after adding 350 µL of cyclodecanone as internal
80 standard and subsequent GC/MS-determination using the validated method of
81 Lachenmeier et al.2 without modifications.
4
82 Analysis of variance (ANOVA) was done using the software Design Expert V6 (Stat-
83 Ease Inc., Minneapolis, USA). Statistical significance was assumed at below the 0.05
84 probability level.
85 RESULTS AND DISCUSSION
86 Behaviour of thujone during distillation
87 The results of the distillation experiments are shown in Figs. 1-3. The three different
88 kinds of distillation show large variations in thujone recovery (p<0.0001). Highly
89 significant differences (p<0.0001) in thujone concentration can be determined between
90 the different fractions in all experiments. The ethanolic distillation shows an exponential
91 growth (r2=0.995). Enrichment occurs in the last fractions (tailings). In contrast, an
92 exponential growth could only be determined until the 8th fraction for both the water-
93 cum-steam distillation (r2=0.998) and the direct steam distillation (r2=0.999), whereas
94 the thujone concentration declined in the 9th and 10th fraction. The total recovery was
95 52% for the ethanolic distillation and 80 82% for the steam distillations.
96 Our results confirm Arnold s previously mentioned theory that the addition of water
97 significantly influences thujone concentration. However, our results disprove the
98 previous view of a 100%-recovery. As absinthe was usually distilled using the water-
99 cum-steam method, an yield of about 80% can be presumed if the heads and tailings (i.e.
100 the first and last fractions) are not separated and no further rectification is done.
101 In this context, an interesting change in absinthe production during the 19th century must
102 be noted. In Duplais first edition recipe of absinthe ordinaire (1855), the herbs were
103 macerated in 15 L of alcohol and distilled to the same volume of 15 L.12 In the fourth
104 edition (1882), the recipe was changed in a way that the maceration was done in 16 L
105 that were distilled to 15 L,14 so that a further reduction of thujone would occur by this
5
106 separation of heads and tailings. Other contemporary literature15-17 proves that
107 distillation of absinthe should never be carried on to the end, as the taste of the product
108 would be too strong, and less fine.16 Therefore, only the main fractions (heart) are used
109 for the production of high-quality absinthe. The heads and tailings are collected
110 separately and added to subsequent macerations or used to make absinthe ordinaire after
111 renewed rectification.17
112
113 Thujone concentration of pre-ban absinthe
114 The previous calculations of a thujone concentration of 260 mg L-1 used the assumption
115 of a wormwood oil yield of 1.5% and a thujone concentration in the oil of 67%. We
116 thoroughly reviewed the literature on the composition of wormwood oil and found 29
117 references,18-45 which are summarised in Table 1. A wide variation in the oil content of
118 wormwood and even wider variations of the thujone content in the oil is notable. A
119 56789 4 234 339 95 4 972 9 4 4 3265 4 54 94 6 9 4 4 -thujone, cis-
120 chrysanthenyl acetat, cis-chrysanthenol, cis-epoxyocimene, sabinyl acetate or
121 bornyl acetate as principal component. Some chemotypes did not contain thujone at all,
122 for example chemotypes from France,45 Italy,36,41 Spain,34 Lithuania,26 and Egypt.31 It is
123 therefore possible to produce absinthe without any thujone.
124 The mean essential oil content of A. absinthium is 0.6Ä…0.3% and the mean total thujone
125 content in the essential oil is 17.8Ä…18.4%. Therefore, it can be concluded that the
126 calculation of 260 mg L-1 was done with exceptionally high values for both parameters
127 and may have therefore been an overestimation. If the values from our literature review
128 are used for calculation and considering the determined distillation behaviour, the
6
129 thujone concentrations of absinthe would be as detailed in Table 2. The mean thujone
130 content of absinthe may have been ranged around 20 mg L-1.
131 The even lower concentrations found in the vintage absinthes may be because the
132 recipes prescribe the use of dried wormwood herb. The drying results in loss of volatile
133 oil by evaporation.46,47 Significant differences in terms of essential oils and herb yields
134 occur at the various ontogenetic stages with the highest oil yields during the flowering
135 period.47
136
137 CONCLUSIONS
138 The results confirm our theory of a previous gross overestimation of the thujone content
139 in pre-ban absinthe. During food processing, a loss of thujone occurs during drying of
140 the plants, during distillation, and by the separation of heads and tailings. This study
141 may resolve the controversy as our newly calculated values correspond to the
142 experimental ones.
143 Finally, we want to emphasise that the derived thujone values in pre-ban absinthe are in
144 conformance to today s maximum limits. Therefore, it is possible to produce absinthes
145 after Duplais historical recipes that would be marketable today. A number of such
146 authentic distilled absinthes have recently become available on the market. Now, only
147 the consumers may need to refocus their attention from absinthe s thujone myth and the
148 supposed effects and learn to choose absinthe purely for quality and taste like any other
149 spirit.
150
7
150
151 REFERENCES
152
153 1 Padosch SA, Lachenmeier DW and Kröner LU, Absinthism: a
154 Fictitious 19th Century Syndrome With Present Impact. Subst Abuse
155 Treat Prev Policy 1:14 (2006).
156 2 Lachenmeier DW, Emmert J, Kuballa T and Sartor G, Thujone-Cause
157 of Absinthism? Forensic Sci Int 158:1-8 (2006).
158 3 Hutton I, Myth, Reality and Absinthe. Curr Drug Discov 9:62-64
159 (2002).
160 4 Schaefer I, Bindler F and Lugnier A, Toxicological Rehabilitation of
161 Absinthium Liqueur. Toxicol Lett 74 Suppl. 1:75 (1994).
162 5 Arnold WN, Vincent van Gogh: chemicals, crises, and creativity.
163 Birkhäuser, Boston, MA, USA (1992).
164 6 Strang J, Arnold WN and Peters T, Absinthe: What's Your Poison?
165 BMJ 319:1590-1592 (1999).
166 7 Bonkovsky HL, Cable EE, Cable JW, Donohue SE, White EC,
167 Greene YJ, Lambrecht RW, Srivastava KK and Arnold WN,
168 Porphyrogenic Properties of the Terpenes Camphor, Pinene, and
169 Thujone (With a Note on Historic Implications for Absinthe and the
170 Illness of Vincent Van Gogh). Biochem Pharmacol 43:2359-2368
171 (1992).
172 8 Adams J, Hideous Absinthe. A history of the devil in a bottle. The
173 University of Wisconsin Press, Madison, USA (2004).
174 9 Baker P, The book of absinthe: a cultural history. Grove Press, New
175 York, USA (2001).
176 10 Meschler JP and Howlett AC, Thujone Exhibits Low Affinity for
177 Cannabinoid Receptors but Fails to Evoke Cannabimimetic
178 Responses. Pharmacol Biochem Behav 62:473-480 (1999).
179 11 Hein J, Juckel G, Kienast T and Heinz A, "Die Grüne Fee" Absinth -
180 Historische Und Biochemische Fakten. Sucht 51:19-25 (2005).
181 12 Duplais P, Traité des liqueurs et de la distillation des alcools ou le
182 liquoriste & le distillateur modernes. Lacroix-Comon, Paris, France
183 (1855).
184 13 European Commission, Commission Regulation (EC) No 2870/2000
185 Laying Down Community Reference Methods for the Analysis of
186 Spirits Drinks. Off J Europ Comm L333:20-46 (2000).
8
187 14 Duplais P, Traité de la fabrication des liqueurs et de la distillation des
188 alcools. Gauthier-Villars, Paris, France (1882).
189 15 Fritsch J, Nouveau traité de la fabrication des liqueurs d'aprés les
190 procédés les plus récents. G. Masson, Paris, France (1891).
191 16 Brannt WT, A practical treatise on the raw materials and the
192 distillation and rectification of alcohol, and the preparation of alcoholic
193 liquors, liqueurs, cordials, and bitters. Henry Capey Bairs & Co.,
194 Philadelphia, USA (1885).
195 17 de Brevans J, La fabrication des liqueurs. J.-B. Bailliére et fils, Paris,
196 France (1908).
197 18 Nano GM, Martelli A and Sancin P, Indagini Chimiche Sulle Artemisie
198 Coltivate in Piemonte. Riv Ital Ess Prof Piante Off Arom Sap Cosm
199 Aerosol 98:409-412 (1966).
200 19 Chialva F and Liddle PAP, Sur La Composition De L'Huile Essentielle
201 De Artemisia Pontica Linnaeus Cultivée En Piémont. Riv Ital EPPOS
202 63:350-352 (1981).
203 20 Bos R, Stojanova A, Woerdenbag HJ, Koulman A and Quax WJ,
204 Volatile Components of the Aerial Parts of Artemisia Pontica L.
205 Grown in Bulgaria. Flavour Fragr J 20:145-148 (2005).
206 21 Talzhanov NA, Sadyrbekov DT, Smagulova FM, Mukanov RM,
207 Raldugin VA, Shakirov MM, Tkachev AV, Atazhanova GA, Tuleuov BI
208 and Adekenov SM, Components of Artemisia Pontica. Chem Nat
209 Comp 41:178-181 (2005).
210 22 Hurabielle M, Tillequin F and Paris M, Chemical Study of Artemisia
211 Pontica Essential Oil. Planta Med 31:97-102 (1977).
212 23 Khanina MA, Serykh EA, Korolyuk AY, Bel'chenko LA, Pokrovsky LM
213 and Tkachev AV, Composition of Essential Oil of Siberian
214 Populations of Artemisia Pontica L. - As a Promising Medical Plant.
215 Khim Rast Syr'ya 2000:85-94 (2000).
216 24 Kordali S, Cakir A, Mavi A, Kilic H and Yildirim A, Screening of
217 Chemical Composition and Antifungal and Antioxidant Activities of
218 the Essential Oils From Three Turkish Artemisia Species. J Agric
219 Food Chem 53:1408-1416 (2005).
220 25 Gholami M, Azizi A and Salehi P, Variations in Essential Oil
221 Components in Cultivated and Regenrated Artemisia Absinthium L.
222 Asian J Chem 17:2229-2232 (2005).
223 26 Judzentiene A and Mockute D, Chemical Composition of Essential
224 Oils of Artemisia Absinthium L. (Wormwood) Growing Wild in Vilnius.
225 Chemija 15:64-68 (2004).
9
226 27 Morteza-Semnani K and Akbarzadeh M, Essential Oils Composition
227 of Iranian Artemisia Absinthium L. and Artemisia Scoparia Waldst. Et
228 Kit. J Essent Oil Res 17:321-322 (2005).
229 28 Chiasson H, Belanger A, Bostanian N, Vincent C and Poliquin A,
230 Acaricidal Properties of Artemisia Absinthium and Tanacetum
231 Vulgare (Asteraceae) Essential Oils Obtained by Three Methods of
232 Extraction. J Econ Entomol 94:167-171 (2001).
233 29 Rahimizadeh M, Hassanzadeh MK and Danesh NM, Analysis of
234 Iranian Artemisia Absinthium L. Essential Oil. ACGC Chem Commun
235 13:33-36 (2001).
236 30 Tegtmeier M and Harnischfeger G, Methods for the Reduction of
237 Thujone Content in Pharmaceutical Preparations of Artemisia, Salvia
238 and Thuja. Eur J Pharm Biopharm 40:337-340 (1994).
239 31 Aboutabl EA, El Azzouny AM and El Dahmy SI, Constituents of the
240 Essential Oil of Artemisia Absinthium Grown in Egypt. J Essent Oil
241 Bear Plants 1:82-86 (1998).
242 32 Sefidkon F, Jalili A, Rabie M, Hamzehee B and Asri Y, Chemical
243 Composition of the Essential Oil of Five Artemisia Species From Iran.
244 J Essent Oil Bear Plants 6:41-45 (2003).
245 33 Carnat A-P, Madesclaire M, Chavignon O and Lamaison J-L, Cis-
246 Chrysanthenol, a Main Component in Essential Oil of Artemisia
247 Absinthium L. Growing in Auvergne (Massif Central), France. J
248 Essent Oil Res 4:487-490 (1992).
249 34 Arińo A, Arberas I, Renobales G, Arriaga S and Dominguez JB,
250 Essential Oil of Artemisia Absinthium L. From the Spanish Pyrenees.
251 J Essent Oil Res 11:182-184 (1999).
252 35 Arińo A, Arberas I, Renobales G and Dominguez JB, Influence of
253 Extraction Method and Storage Conditions on the Volatile Oil of
254 Wormwood (Artemisia Absinthium L.). Eur Food Res Technol
255 209:126-129 (1999).
256 36 Nin S, Arfaioli P and Bosetto M, Quantitative Determination of Some
257 Essential Oil Components of Selected Artemisia Absinthium Plants. J
258 Essent Oil Res 7:271-277 (1995).
259 37 Pino JA, Rosado A and Fuentes V, Chemical Composition of the
260 Essential Oil of Artemisia Absinthium L. From Cuba. J Essent Oil Res
261 9:87-89 (1997).
262 38 Mucciarelli M, Caramiello R and Maffei M, Essential Oils From Some
263 Artemisia Species Growing Spontaneously in North-West Italy.
264 Flavour Fragr J 10:25-32 (1995).
10
265 39 Sacco T and Chialva F, Chemical Characteristics of the Oil From
266 Artemisia Absinthium Collected in Patagony (Argentina). Planta Med
267 54:93 (1988).
268 40 Tucker AO, Maciarello MJ and Sturtz G, Essential Oils of Artemisia
269 `Powis Castle' and Its Putative Parents, A. Absinthium and A.
270 Arborescens. J Essent Oil Res 5:239-242 (1993).
271 41 Chialva F, Liddle PAP and Doglia G, Chemotaxonomy of Wormwood
272 (Artemisia Absinthium L.) I. Composition of the Essential Oil of
273 Several Chemotypes. Z Lebensm Unters Forsch 176:363-366 (1983).
274 42 Kaul VK, Nigam SS and Banerjee AK, Thin Layer and Gas
275 Chromatographic Studies of the Essential Oil of Artemisia Absinthium
276 Linn. Indian Perfumer 23:1-7 (1979).
277 43 Blagojevic P, Radulovic N, Palic R and Stojanovic G, Chemical
278 Composition of the Essential Oils of Serbian Wild-Growing Artemisia
279 Absinthium and Artemisia Vulgaris. J Agric Food Chem 54:4780-4789
280 (2006).
281 44 Khalilov LM, Paramonov EA, Khalilova AZ, Odinokov VN, Muldashev
282 AA, Baltaev UA and Dzhemilev UM, Identification and Biological
283 Activity of Volatile Organic Compounds Emitted by Plants and
284 Insects. IV. Composition of Vapor Isolated From Certain Species of
285 Artemisia Plants. Chem Nat Comp 37:339-342 (2001).
286 45 Juteau F, Jerkovic I, Masotti V, Milos M, Mastelic J, re JM and Viano
287 J, Composition and Antimicrobial Activity of the Essential Oil of
288 Artemisia Absinthium From Croatia and France. Planta Med 69:158-
289 161 (2003).
290 46 Guenther E, The essential oils. Vol. V. Van Nostrand, New York
291 (1952).
292 47 Tansi S, Effects of Harvest Times on Wormwood (Artemisia
293 Absinthium L.) of Different Origins in Turkey. Agri Med 130:5-10
294 (2000).
295
296
297
11
297 Legends:
298 Tables:
299 Table 1. Total essential oil content and relative thujone concentrations of the two
300 Artemisia species used for absinthe production (data summarized from the literature).
301 Table 2. Wormwood content in historic absinthe recipes and possible resulting thujone
302 concentration in absinthe.
303
304
305 Figures:
306
307 Figure 1. Behaviour of thujone during ethanolic distillation.
308
309 Figure 2. Behaviour of thujone during water-cum-steam distillation.
310
311 Figure 3. Behaviour of thujone during direct steam distillation.
312
313
12
313 Table 1. Total essential oil content and relative thujone concentrations of the two
314 Artemisia species used for absinthe production (data summarized from the literature).
Species Total Total thujone References
Ä…-thujone ²-thujone
essential oil [% in essential
[% in essential [% in essential
[%] oil]
oil] oil]
18-23
A. pontica Min-Max 0.2-0.9 nd-30 nd-4.2 nd-30
MeanÄ…SD 0.3Ä…0.1 15.4Ä…12.0 1.5Ä…1.8 16.8Ä…12.7
24-45
A. absinthium Min-Max 0.3-1.6 nd-60.0 nd-69.7 nd-70.6
MeanÄ…SD 0.6Ä…0.3 5.9Ä…11.6 12.6Ä…15.5 17.8Ä…18.4
315
316
13
316 Table 2. Wormwood content in historic absinthe recipes and possible resulting thujone
317 concentration in absinthe.
Recipe after Duplais A. pontica A. absinthium Thujone concentration [mg L-1]*
(1855) 12 (kg/100 L) (kg/100 L)
Min. Max. Mean SD
Absinthe ordinaire - 2.5 0 75 18 23
Absinthe demi-fine 1.0 2.0 0 66 18 21
Absinthe fine, 0.5 2.0 0 63 17 20
Absinthe Suisse de
Montpellier ou Lyon
Absinthe Suisse de 1.0 2.5 0 81 22 26
Pontarlier
318
319 * Assumptions: 80% yield during distillation and removal of heads and tailings (5% of
320 the distillate).
321
14
321
Thujone
Alcoholic strength
90
500
85
400
80
Thujone-Recovery:
134 mg/L (52%)
300
75
200
70
100
65
1 2 3 4 5 6 7 8 9 10
Ethanolic distillation [fraction number]
322
323 Figure 1. Behaviour of thujone during ethanolic distillation.
324
Thujone
Alcoholic strength
90
500
85
400
80
Thujone-Recovery:
216 mg/L (83%)
300
75
200
70
100
65
1 2 3 4 5 6 7 8 9 10
Water-cum-steam distillation [fraction number]
325
326 Figure 2. Behaviour of thujone during water-cum-steam distillation.
15
-1
Thujone [mg L ]
Alcoholic strength [% vol]
-1
Thujone [mg L ]
Alcoholic strength [% vol]
 Thujone
Alcoholic strength
90
500
85
400
80
Thujone-Recovery:
209 mg/L (80%)
300
75
200
70
100
65
1 2 3 4 5 6 7 8 9 10
Direct steam distillation [fraction number]
327
328 Figure 3. Behaviour of thujone during direct steam distillation.
16
-1
Thujone [mg L ]
Alcoholic strength [% vol]