62

E. T. Contis et al. (Editors) Food Flavors: Formation, Analysis and Packaging Influences © 1998 Elsevier Science B.V. All rights reserved

679

Effect of rosemary and 1,4-dihydropyridines on oxidative and flavour changes of bergamot oil F. Pudil, J. Volfovd, V. Janda, H. Valentovd and J. Pokomy Department of Food Chemistry and Analysis, Prague Institute of Chemical Technology, Technicka5,CZ-166 28 Prague 6, Czechia

Abstract Bergamot oil was oxidized in the temperature range of 40-60°C at restricted oxygen levels. The course of oxidation was investigated by gas chromatography (GLC) and gas chromatogaphy with mass spectrometric detection (GC/MS). Flavour acceptabilities and sensory profiles were determined at the same time. The sensory acceptability decreases on storage, and sensory profiles change, too. Limonene oxides and carvone, which were identified as main oxidation products, have particular sensory profiles. An addition of rosemary extracts inhibits the oxidation, but it influences the ratios of various components of stored oil. Therefore, it changes moderately the sensory profile. The sensory profile changes are more pronounced, when 1,4-dihydropyridine antioxidants are used. Both oxidative and sensory changes depend on the structure of dihydropyridines, too.

1. INTRODUCTION Citrus oils are widely used as flavorings for food and cosmetic products. Bergamot oil, producedfi'ombitter orange {Citrus aurantium subsp. Bergamie Risso et Poiteau Engler) were applied for various foods and beverages, such as fiuit juices [1], soft drinks [2], liqueurs [3], firuit jams [4],flavouredtea [5] and even flying oil [6]. The disadvantage of bergamot oil is its high content of monoterpenes [7], particularly limonene, linalyl acetate and linalool, which decrease resistance against oxidation under storage. Black tea, flavoured with bergamot oil, was stable only for two months at room temperature [8]. At 50 or 60°C, the peroxide value of 300 mval/kg was soon reached which caused deterioration of flavour quality [9]. Stored samples of bergamot oil have high peroxide values [10], unless they have been stabilized with antioxidants; they are oxylabile even in cold orfi^ozenstorage [11]. The peroxide value was found a good indicator of sensory quality of orange oil, but tiie gas-chromatographic profile was preferable [12]. More than 160 components were identified in aged bergamot oil, using GC/MS [13]. Oxidation mechanisms of terpenic hydrocarbons and their oxygenated compounds were discussed [14]. Oxidation products modify the flavour of essential oils. Therefore, we studied the relationship between the chemical composition and the sensory profile of stored bergamot oil, and the effect of antioxidants on the product.

680 2. MATERIALS AND CHEMICALS Bergamot oil (free of antioxidants, without deterpenation) was supplied by Sigma (UK), (S)(-)-limonene (96 % purity) and (±)-linalool (97 % purity) by Aldrich (Gemany). Rosemary extract was produced by extracting 100 g of fresh dried rosemary leaves (grown in Poland, in 1995) four times with 1 L acetone, macerating overnight at room temperature of 22°C, combining the filtrates, removing the solvent by distillation, and drying on a boiling water bath. The main active components were identified as camosic acid and camosol. Diludine (2,6-dimethyl-3,5-diethoxycarbonyl-l,4-dihydropyridine) of 99 % purity and OSI 7284 (2,6dimethyl-3,5-dibutoxycarbonyH,4-dihydropyridine) of 98 % purity were prepared in the Institute of Organic Syntheis (Director: Prof Dr. G. Duburs) in Riga, Latvia.

3. EXPEMMENTAL PROCEDURES 3.1. Oxidation of bergamot oil A 100 fiL portion of bergamot oil was placed into a 10 mL vial containing an internal standard (5 mg of «-decane). The vial was sealed and conditioned at 40°C (in some experiments, at 60'^C) in a thermostat. A volume of 0.1 mL of vapour phase was injected by gas-tight Hamilton syringe mto the gas- chromatograph (using a headspace autosampler). Alternatively, the solid phase microextraction (SPME) sampling technique was used. To study the antioxidant efficiency, 1 mg of the antioxidant dissolved in 100 mL methanol, was added, and the solvent evaporated. Essential oil was added afterwards, the vial was sealed, and thoroughly shaken. 3.2. Sotid phase microextraction (SPME) A 65 pm Carbowax-divinylbenzene fiber for a manual holder (Supelco, USA) was used for extracting volatiles for GC/MS. The fiber was inserted into the vial containing the stored sample (see Section 3.1). Extraction time was 10 min at 40°C and desorption time 2 min at 220°C. The fiber was cleaned for 30 min At the same temperature. Details of the apparatus are given in another chapter in these proceedings[16]. 3.3. Gas chromatography (GLC) and GLC with mass spectrometric detection (GC/MS) The gas chromatograph GC8000 (Fisons Instruments) was equipped with an autosampler HS800 (injector temperature 220°C) and a 60 m x 0.32 mm column, coated with Supelcowax 10 (fibn thickness 0.25 mm) was used. The column temperature was programmed from 50*^0 (2 min isothermally), heating rate 2°C/min to 220°C (isothermal for 30 min) Tnc fl^me ionization detector (FID) was used (temperature 250*^0). The c^^T?er gas was helium at an initial pressure 100 kPa. The input/split ratio 1:25. Retention indices were calculated using a mixture of «-alkanes as reference substances, as well as for calculation of peak areas. A MSD8000 mass spectrometer was used for GC/MS; the ionizing energy was 70 eV. Pure standards (Aldrich, purified by GLC) were used for identification of mass spectra.

3.4. Sensory analysis Conditions for the sensory analysis were in agreement with the international standard ISO [17]; the test room was equipped with six standardized test booths [18]. The panel of assessors consisted of selected, trained and monitored persons [19] with experience in sensory profiling of at least six months. The amount of 100 mg of sample was placed into a 250-mL wide-neck

681 ground-glass bottle, and left at least 2-3 hours to equilibriate. The odour intensity was evaluated by sniffing. For the analysis of stabilized samples, 100 jiL of 1% methanolic solution were added, the solvent evaporated, the sample was added, the bottle thoroughly shaken, and left for 2-3 h. For the hedonic rating, the odour acceptability was determined using an unstructured graphical scale - straight lines 100 mm long [19] (0% = rather bad; 100% = excellent). The sensory profile [20] consisted of 24-36 descriptors (they are shown in the respective figures in spider-web diagrams); unstructured graphical scales 100 mm long were used (0% = imperceptible; 100% = very strong). Two samples were served at a session in random order. The total of 24 responses were used for the calculation of means. The standard deviation of the means varied between 2-6% of the scale.

4. RESULTS AND DISCUSSION 4.1. Composition of bergamot oil Bergamot oil was analyzed by GC/MS, and a chromatographic profile is shown in Figure 1. The sample was moderately oxidized (20 h at 40°C) in order to better determine the peak position of some oxidation products. The chemical structure of identified compounds was determined by comparison of mass spectra and retention indices with those of authentic substances; they are listed in Table 1. Linalool, linalyl acetate and limonene were present as major terpenic components, in agreement with the literature [7, 22, 24], and myrcene, cis- and /ran^-ocimene, and p-cymene were found in smaller amounts. Caryophyllene was the only sesquiterpene detected.

3.5] 21

3.0J

S

141 1<

4

If

r

2.5

|| ii

f

• 1.5-1

3

1

iMuJLuLjLL^U 0,5-]

1 11

1

14 1

11 1 1

t

i.oj 1

1

1

17 1

h

k

JLJ 30

0 1

^

40

Time [min] Figure 1. Chromatographic profile of bergamot oil after storage at 40°C for 20 h. Numbers above each peak refer to the compound shown in Table 1.

682 Table 1 List of identified substances in the sample of very slightly oxidized bergamot oil (20 h at 40°C, corresponding to Figure 1) Retention Retention Identified Peak substance index time [min] No. 1 6.53 840 acetone 7.54 2 methanol 910 3 8.23 946 ethanol 4 16.79 1174 p-myrcene 5 18.65 1209 limonene 6 20.83 1246 c»5-ocimene 7 21.85 1263 trans-ocimeiae 8 23.01 1281 /j-cymene 9 34.04 1453 c/f-linalool oxide 10 34.95 1467 acetic acid 11 35.91 1481 /rans-linalool oxide 12 39.38 1536 3,7-dimethyl-1 -octen-3-ol 13 39.62 1540 y-terpineol 14 40.97 1561 linalool 15 41.68 1571 iinalyl acetate 16 43.15 1594 sesquiterpene (probably »o-caryophyllene) 17 49.36 1648 citral 18 50.13 1654 a-terpineol 19 51.98 1668 neryl acetate 20 52.67 1675 carvone 21 53.78 1681 geranyl acetate 22 66.47 1975 caryophyllene oxide

4.2. Oxidation of bergamot oil Bergamot oil was oxidized under simulated storage conditions of flavored food products, i. e. under limited access of oxygen at 40°C and in thin film of about 1 mm. Examples of chromatographic profiles are shown in Figure 2 where a sample taken near the beginning of oxidation is compared with a sample stored for about a month, when oxygen has been consumed and various secondary (non-oxidative) reactions proceeded. Various oxidation reaction proceeded in the beginning of storage, but when oxygen in the headspace has been exhausted, various secondary reactions followed. Changes at 60°C were very sunilar, therefore, they are not presented here. Limonene was gradually destroyed firom the start of storage and the course followed the zero order kinetics until complete exhaustion of oxygen. Data for several other compounds are given in Table 2. Myrcene was oxidized from the beginning, too. On contrary, the degradation of/?-cymene followed after a distinct lag period. The contents of a- and y-terpineols and 3,7dimethyH-octen-3-ol mcreased during llie storage. The two ocimenes reached their maximum in the beginning of storage. Acetone, a typical oxidation product formed by decomposition of a hydroperoxide, was produced very rapidly, reaching a maximum after 40 hours. Both cisand /ran^-linalool oxides were formed slowly, and attained their maxima when oxygen was already consumed.

683 Table 2 Changes of some compounds during the storage of bergamot oil at 40'^C 428 h 39 h Peak No. 739 h 354 h 185 h 612 h 4.38 4.16 5.22 6.88 14.60 1 6.48 31.33 4 16.75 46.89 107.45 148.63 20.52 5.34 9.21 7.69 6.65 6 6.53 8.47 16.62 18.15 27.59 7 9.50 11.46 27.08 40.23 179.57 216.48 288.96 8 6.18 6.36 6.05 5.46 4.26 9 6.07 5.87 5.65 3.01 5.73 4.60 11 1.74 0.55 12 1.82 1.61 1.31 0.91 1.27 1.69 1.12 0.50 1.35 0.69 13 7.54 6.25 4.42 3.50 2.40 18 6.03 Note: Peak numbers are the same as in Table 1. Numbers refer to FID peak area. 4.3. Effect of rosemary extract on the degradation of bergamot oU Rosemary extract, an efficient natural antioxidant used for the stabilization of fats and oils [25, 26], was found active in stabilizing limonene and linalool under storage conditions at 40°C. No changes or very small changes were observed in most components for the first 400 hours of storage. The activity was similar to that observed in Citrus hystrix oil [16]. Changes of the composition of bergamot oil stabilized with rosemary extract during the storage at oxygen are partially shown in Table 3. The contents of acetone, 3,7-dimethyl-locten-3-ol, /7-cymene, P-myrcene, a- and y-terpineols changed similarly during the storage as in non-stabilized oil. The degradation oicis- and /ra^w-ocimenes was moderately inhibited by rosemary extract. Both cis- and /ra/w-linalool oxides were formed in inhibited samples at the same rate at the beginning as in the control containing no antioxidants, however, they were decomposed at a faster rate at later stages of storage (after 400 h) if rosemary extract was present. Table 3 Changes of some components of bergamot oils stored with rosemary extract at 40°C Peak No. 41 h 187ji 356 h 543 h 713 h 1.34 2.39 20.91 1 1.60 5.10 57.77 120.74 58.78 24.26 31.89 4 8.04 8.24 7.80 7.98 8.82 6 22.04 30.96 13.83 17.02 29.36 7 90.01 96.29 2.16 21.67 175.51 8 4.71 2.28 3.87 4.95 5.06 9 4.01 3.77 2.33 3.65 4.07 11 1.28 0.79 1.95 0.92 1.56 12 0.64 1.81 1.11 0.69 1.23 13 4.91 7.59 3.54 3.07 5.55 18 The structure and numbers of peaks are the same as in Table 1.

Since bergamot oil is a complicated mixture of terpenes, the sensory changes that occur with storage are difficult to evaluate and compare. Therefore, the effect of rosemary extract

684 was studied using purified limonene and linaiooi. An example of limonene is shown in Figure 3 (changes occurring in stored linaiooi were very sunilar). Degradation of citrus notes was very pronounced in the non-stabilized sample, but it was slightly smaller in the stabilized sample. In presence of rosemary extract, the formation of woody, acidic and heavy odour notes were suppressed in stored bergamot oil; on contrary, the formation of spice and fresh odour notes was stimulated. The hedonic rating decreased by 28% of the scale in case of unstabilized sample and by 14% only in case of stabilized sample, respectively. 70-

60-

SO

a e

<0

0S 30

., 1

1

20-

10-

-1 t

• 1

0-

LJLL-

IJJi

JI II

^

^ •.

.-

t

Time [min] Figure 2. Chromatographic profiles of oxidized bergamot oil, A - 40 hours at 40*^0, B - 740 hours at 40°C.

lemon juice heavy ^ pungent ^

V

terpenic^

y 1

*7

> orange juice

/\^^^} r 7 i \ y ^ \ ^ ^ , . ^ - ^ orange peel

^^^^"^^^^

menthol ^—^-^ '

I _ , _ _ _ ^ lemon peel

'Z^^^^^i^^^^^^^^^^'*^~^~~~' ^ ' ^

/

i

^

\

^floral -o-O hours

woody

f anise

\

acidic spicy

- o — rosemary ~-^r- dliudlne - • ^ OSI7284

Figure 3. Sensory profiles of limonene stored with and without rosemary extract

685 4.4. Effect of 1,4-dihydropyridine antioxidants on the degradation of bergamot oil Dihydropyridine derivatives are used for the stabilization of feeds and phannaceutical preparations such as carotene supplements [27, 28]. In terpenes they were found less efficient as inhibitors of oxidation, similarly as in pure terpenes [29], but they influenced the formation of various oxidation and other degradation products. Changes of several important components are summarized in Table 4 for oil stabilized with Diludine and in Table 5 for oil stabilized with OSI 7284, respectively.

Table 4 Changes of some components during the storage of bergamot oil at 40°C in presence of Diludine 717 h 547 h 358 h 189 h 43 h Peak No. 15.71 15.40 15.55 21.76 27.25 1 29.34 142.14 14.17 53.72 125.20 4 6.77 4.32 8.60 6.52 7.70 6 13.41 7 7.56 14.90 23.25 25.43 21.30 361.06 525.62 266.62 S 9.71 12.67 10.36 9 6.03 4.00 2.86 9.25 11.33 9.72 5.45 11 1.11 0.60 1.00 1.57 0.85 12 0.57 0.66 1.67 1.11 0.87 13 5.24 5.89 0.35 2.69 4.12 18 Compounds and numbers of peaks are the same as in Table 1.

Table 5 Changes of some components during the storage of bergamot oil stabilized with OSI 7284 Peak No. 45 h 207 h 375 h 548 h 721 h 4.05 6.07 11.16 18.01 40.28 1 154.47 217.14 4 18.55 29.77 70.95 12.87 6.04 8.32 9.43 10.76 6 10.47 45.91 7 15.56 24.18 35.87 282.24 53.46 207.30 8 1.89 334.63 6.64 4.22 5.80 6.95 6.05 9 4.17 5.51 6.14 5.46 4.73 11 1.47 1.70 1.25 1.11 0.98 12 0.93 1.10 1.00 1.25 1.03 13 4.44 3.82 6.12 6.06 5.20 18 Compounds and numbers of peaks are the same as in Table 1. The destruction of myrcene was efficiently inhibited in presence of OSI 7284, but no effect was observed with Diludine. The maximum contents of cis- and trans—linalool oxides were higher than in absence of antioxidants or in presence of rosemary extract. The maximum was particularly high, when Diludine has been added. The content of p-cymene changed similarly. The maxima of cis- and /ran^-ocimenes were much higher in oil stabilized with OSI 7284 than m oil stabilized with Diludine. During 300-400 h of storage the concentrations of a- and

686 y-terpineols were almost the same in all samples studied. During further storage their content decreased in presence of dihydropyridines, especially Diludine, while it continued to increase in non-stabilized oil or in oil stabilized with rosemary extract. Similar differences were observed for 3,7-dimethyl-l-octen-3-ol. The formation of acetone was also stimulated by dihydropyridine derivatives. These differences could be explained by presence of a nitrogen group in dihydropyridines. Their weakly basic character may influence not only the course of hydroperoxide decomposition, but the course of non-oxidative degradative changes such as retroaldolization. From thesedata, it is evident that changes in sensory profiles of stored bergamot oil are not due only to the inhibition of primary oxidation reactions, but also depend on the structure of antioxidant added.

5. ACKNOWLEDGEMENTS Supported by the research grants Copernicus CIPA-CT94/0111 and OK 175 ( M S M T CR). The rosemary extract was prepared and characterized by Dr. J, Korczak, Agricultural Univeristy, PoznapPoland. The dihydropyridine antioxidants were preprared and characterized by prof. Dr. G. Duburs and co-workers. Institute of Organic Synthesis, Riga, Latvia.

6. REFERENCES 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

16 17

E. Benk, Mineralbrunnen, 34 (1984) 410. C. Arena, Ind. Bevande, 10 (1980) 209. E. Benk, Alkohol-Ind., 98 (1985) 278. E. Benk, Ind. Obst- GemOseverwertung, 70 (1985) 358. H. P.Neukom, D. J. Neier and D. Blum, Mitt. Geb. Lebensm.-Unters. Hyg., 84 (1993)337. W. J. Begemann and P. D. Harkes, EP No. 0191 519 Al (1986). P. E. Shaw, J. Agr. Food Chem., 27 (1979) 246. N. Dhanaraj, C. P. Natarajan and R. Seshadri, J. Sci. Food Agr., 37 (1986) 185. J. Pokomy, D. Wiewi6ricova, E. Nfiiec, H. S^ankova and D. Curda, Sb. V§CHTPrazeE,34(1972)5. A. DiGiacomo, F. Bovalo and M. Caliciuri, Essenze, Deriv. Agrumeri, 45 (1975)110. P. Rovesti, Essenze, Deriv. Agrumari, 45 (1975) 100. J. Pokomy, D. Curda, L. Jirdskova, D. Wiewiorkovd and G. Janidek, Sb. V§CHTPrazeE,34(1972)21. G. Mazza, J. Chromatogr., 362 (1986) 87. G. Mazza, Essenze, Deriv. Agrumari, 57 (1987) 5. J. Pokomy, F. Pudil, K. Ulmannovd and E. Ficov^ in: Food Flavors: Generation, Analysis and Process Influence (G. Charalambous, ed.). Barking, Elsevier Science, 1995, p 815. F. Pudil, H. Vijaya, V. Janda, J. Volfova, H. Valentovd and J. Pokomy, These Proceedings, p. ISO 6658: Sensory analysis - Methodology - General Guidance. ISO, Geneva, 1985.

687 18 19 20 21 22 23 24 25 26 27 28

ISO 8589: Sensory analysis - General guidance for the design of test rooms. ISO, Geneva, 1988. ISO 8586: Sensory analysis - General guidance for the selection, training and monitoring of assessors. ISO, Geneva, 1989. ISO 6564: Sensory analysis - Flavour profile. ISO, Geneva, 1985. ISO 4121: Sensory analysis - Grading of food products by methods using scale categories. ISO, Geneva, 1988. L. Mondello, P. Dugo, K. D. Bartle, G. Dugo and A. Cotroneo, Flavour Fragrance J., 10 (1995) 33. R. Huet, Fruits, 46 (1991) 501, 551, 671. N. Nakatani, Nippon Nogeikagaku Kaishi, 62 (1987) 170. J. Pokorny, H. T. T. Nguyen and J. Korczak, Nahrung, 41 (1997) 176 L. Koui^imska, J. Pokorny and G. Tirzitis, Nahrung, 37 (1993) 91. L. Farnikova, J. Pokorny, G.Tirzitis, Potrav. Vedy, 13(1995) 225. J. Pokorny, F. Pudil, J. Volfova and H. Valentova, These Proceedings, p.