COMPARISON OF AROMA COMPONENTS IN THAI MANGO (CV. KAEW) FROM DIFFERENT EXTRACTION METHODS Montatip Yunchalad 1 ; Lozan0 2, Yves; and Dhuique-Mayer 2, Claudie. 1 IFRPD, Kasetsart University, PO Box 1043, Bangkhen, Bangkok '0903, Thailand 2 Cirad-Flhor, Cirad, PO Box 5035, Montpellier -34032, France ABSTRACT Aroma components in Kaew mango were prelimipary identified. Aroma extracts obtained by means of simultaneous distillation-extraction, head space concentration and solvent extraction were analysed by capillary GC and GC-MS. Comparative qualitative chromatogram of different methods showed that almost aroma components were more or less sirnilar. However, the result revealed that the silmutaneous distillation extraction was found to be more suitable than the other two methods. It was also found that monoterpene hydrocarbon was the major aromatic chemical group which contributed about 83% of the total mango aroma components, in tj.ich the most abundant compound was a-terpinolene (70%) and a-car-3-ene (6%). Keywords: Mangifera indica L., Kaew mango, aroma components INTRODUCTION Mango (Mangifera indica L.) is the third importance fruit crop grown in Thailand next to the pineapple and rambutan. Thailand is unique in that mango is consumed in significant quantities as green fruit in addition to the universal eating of ITuit at ripe stage. Kaew mango being the dominant green cu!tivar (Mendoza, 1984) is widely utilised maiilly for cornrnerciaj production of salted pickling. The ripe Kaew mango is also rather cheap in comparison to the other ripe mango cuhivars. Hence, mango juice is made ITom ripe mango that showed the desirable color, and attractive flavour quality. There are some research informations in the field of fruit development, postharvest physiology and preservation of Thai mango (Mendosa, 1984; Phithakpol, 1985). But, data in this particular aroma of Thai mango cultivars are scanty in the literatures. Whereas the reports on the aroma compositions of different mangos cultivars, grown in various part of the world are available in plenty. Therefore preliminary study on the aroma components of Kaew mango should be carried out. There are many extraction techniques and analytical methods used for aroma qualifications. The selection of an appropiate extraction method is very important for both aroma qualification and quantification. Because any losses of important compounds cannot be made up in a later phase of the investigation (Maarse, 1982). Furthermore,each method has its own unique shortcoming, no single method will provide an aromatic profile truely representative of the flavour (Heath & Reineccius, 1986). The aim of this work was to extract the aroma components ITom Kaew mango by comparative extraction methods. The aroma extracts obtained were analysed by GC and GC-MS, then to compare the result obtained by each method of extraction for preliminary qualitative experiment. Key word: Mangifera indica L Kaew mango aroma component
MATERIALS AND METHODS Sample preparation Green mature Kaew mangoes (200 kg) were shipped by air ITeight ITom Bangkok, Thailand to the Cirad Flhor laboratory in Montpellier, France and immediately stored at 11 C for a week. After inducing maturation with C2H. for 24 h at 20 C, the mango fi-uitswere allowed to ripen for 5 days at ambient temperature. Then, the fully ripe mango obtained, were sorted for blemish, washed and pulped. The mango pulp was stored at -20 C for aroma extraction and analysis. Extraction of aroma component 1 : Simultaneous distillation -extraction (SDE) method A simultaneous steam distillation-extraction apparatus (modified Likens & Nickerson apparatus) was employed with the aqueous slurry of 20 g of Kaew mango pulp and 15 ml of pentane used for the flavour extraction simultaneously distillation. In the process for 2h, steam and solvent vapour were condensed together and separated in the V-tube section of the apparatus, after the steam volatile aroma are extracted continuously by pentane vapour: Each liquid phase was flowed back to their respective flasks. Combined extracts were dried over anhydrous Na2S04 then, concentrated under a mild stream of nitrogen gas to 250 ill prior to GC analysis. 2 : Solvent extraction method The simplest method of solvent extraction was carried out on the aqueous slurry of 50g Kaew mango pulp. The mixture obtained was extracted with 3 separate 90 ml of n-pentanedichloromethane mixture (2:I,v/v) in a stoppered erlenmayer flask. The solvent layer was separated ITomaqueous layer by placing the flask in an ultrasonic water bath for 10 min. The combined extracts were dried over anhydrous Na2S04 and then concentrated to 250 ψl under nitrogen gas flow prior to GC analysis. 3 : Headspace concentration method The headspace concentration of Kaew mango pulp (8 g) was conducted at 37 0 C in specially designed flask cormected to a condensing system. A constant flow of nitrogen gas at the rate of 40 ml/min was purged through the mango pulp. The aroma components of headspace carried along with the nitrogen flow were then trapped in a stainless steel tube containing Tenax- GC porous pol)1ticrfixed to the outlet holding over the condenser. The volatile aroma headspace of mango was collected for 30 mm. The Tenax tube was then transferred into the oven of Desorption Concentration Injection (DCI) apparatus built in the GC. The DCI was operated so the aroma components were analysed by capillary gas chromatography.
Gas chromatography (GC) The extracts were analysed by using Hewlett Packard 5890 Series II gas chromatography for direct injection and Girdel Series 3000 gas chromatography for DCI headspace analysis. Both were equipped ith FID. Fused silica capillary columns used were J&W, DB-Wax of 0.25 11m film thickness, 0.32 mm Ld and 30m, 60 m length, respectively. The temperature programme used was 60 C for 5 min, and then increased 2 C/min up to 200 C then held for 20 min. 1m: flow rate of carrier gas helium was I and 0.92 ml / min for 30 and 60 m column, respectively. Sample sizes were 6 III v.ith split ratio 1:50 for 30m column and 3f1l with split ratio 0.72 :17 for 60 m column. Injector and detector temperature were set at 220 C and 240 C, respectively. Aroma components from Tenax trap were directly transferred into another built in concentration trap. This second trap was held at -20 C by means of liquid nitrogen. The nitrogen gas at the flow rate of 30 mumin was used to transfer such components.. After connecting the concentration trap to the capillary column by rotating a gas valve, the trap was rapidly heated to 200 C «5sec). The desorbcd components were then directly flushed into the column and the oven heating programme of the GC was simultaneously started. Concentrated aroma components obtained from solvent extraction and SED methods were analysed by direct injection to Gc. Gas chromatography-mass spectrometry (GC-MS) Identification of all aroma components were achieved by means ofgc-ms using GC 8035 and Fisons Instrument Trio 1000 with an eletronic impact ionisation type(ei+). The operating conditions were as follow: mass scanning range, 40-400 AMU; scanning time 0.85 sec; ionization voltage, 70 ev. The injected volume was lf1l with split ratio, 1:16. The other conditions for GC and desorption steps were the same as described above. The components in aroma extracts were separated by GC and identified by comparison relative retention time between chromatograms of Kaew mango pulp and known standards; then confirmed by mass spectrogram based on the Wiley data bank. Aroma components investigation RESULTS & DISCUSSION Table I lists the aroma components of Kaew mango pulp analysed by GC from the extracts of different extraction methods. Evaluation of chromatograms indicated that the major components included a-terpinolene, δ-car-3-ene, 3-hexen-1-ol, p-myrcene, limonene, β- phellandrene and p-selinene. The most abundant compound found in Kaew mango was α- terpinolene 70% (peak area of total aroma), which was also the major component in Sri Lanka mango, namely 32% in Willard mango and 35% in Parrot mango (MacLeod and Pieris, 1984) and 20%in Kensington mango (Macleod et ai., 1988). o-car-3-ene was the minor contributor (6%) in
Kaew mango. It became the most important aroma component ca 26% (unknown cultivar) in Venezualan mango (Macleod and Troconis, 1982) and also comprised 76.4% in Keitt mango and 60% in Tommy Atkins mango in Florida (MacLeod and Snyder, 1985). The following compounds, a-pinene, trans -caryophellene and (3-selinene were detected only in SDE and solvent extrction methods, but not detected in head space concentration method. The latter two compounds may be impacted in the mango pulp due to high MW and less volatility. However, most of aroma components existed in fully ripe Kaew mango were also found in other mango cultivars as reported in literature (MacLeod and Troconis, 1982; Macleod and Pieris. 1984; Macleod et ai., 1988; Macleod and Snyder, 1985). They varied to their relative amount in the aroma extracts.. In addition, some compound (as ester) was not even detected at all in Kaew mango, but Kensington mango contnduted ca 33% of ester (Macleod et ai., 1988). Effect of extraction method GC analysis of headspace aroma has shown less satisfactory result than in SDE method. Gas chromatograms of both methods were shown in Fig.I and Fig.II. ChromatOgram from headspace analysis showed more components in the lower molecular weight (MW) area; whereas higher MW components were shown in the chromatogram from SDE method. Most of the lower MW aroma components were in traces which could not be identified by GC analysis. In this case, a large amount of mango pulp should be employed to get more detailed aroma profiles in the chromatogram. Since the emulsion problem encountered in the solvent extraction method, only the extracts from the other two methods were further analysed by GC-MS. The accurate identification of GC separated components would be confirmed by this technique as shown in Table II and Table II!. The results were noticed that by headspace method obtained more various alcohol compounds (I-butano~ 2-hexen-I-oI, 2-pentan-1-o1 and 2-ethylhexan-I-oI). This one also provided more aldehyde and ketone compounds (I-phenylethanone, 4,4-hydroxymethylpentan-2- one and benzaldehyde). These compounds were not indicated by GC analysis, due to rather low relative amount of aroma components. thennally chemical changes in steam distillation. For SDE method, those compounds might be induced by Thus such compounds could not be found in mass spectrogram obtained from this method. Furthermore, some chemical class of aroma compounds were concluded as well in Table IV. The aroma components obtained from both extraction methods were indicated that monoterpene hydrocarbon existed as most abundant aroma components ca 83% of the total mango aroma,in which is reponed to be imponant to mango flavour for the characterisation of free volatile aroma of mango (MacLeod and Troconis, 1982; Engel and Tress!, 1983; Sakho et ai., 1985). It was noticed that sequisterpene hydrocarbon were only obtained by SDE method. The reason for this is that headspace technique may be less sensitivity to purge higher MW and more polarity aroma components.
CONCLUSION In summary, a fully ripe Kaew mango pulp were extracted and analysed the aroma components by GC and GC-MS. Investigation of different chromatograms indicated that almost all the aroma profiles were more or less similar in the three extraction methods. The abundant aroma was monoterpene hydrocarbon obtained 83% of mango constituents, in which the major compound wvas a-terpino1ene(70%) and 8-car-3-ene(6%). It is thus concluded that no relatively remarkable changes in the aroma profiles of Kaew mango were detected in those extraction methods described in this experiment. However, regarding the unique character and reproducibility of the extraction method, SED method provided the satisfactory reproducibility of aroma components extraction. [n addition, some reflux apparatus should be used for effective solvent removal with minimal loss of volatile aroma. Furthermore, sniffing system should be available to assess the sensory properties of every separated aroma component. REFERENCE Enge~ K. H. and Tressl, R. 1983. Studies on the volatile components of two mango varieties. Journa[ of Agricultural Food Chemistry, 31 : 796-801. Heath, H.B. and Reineccius, G. 1986. F[avour chemistry and technology. Macmillan Publishers, Newyork. 442 p. Maarse, H. 1982. Recent development in the methodology offla\'our research. [n International symposium on food flavour scientific coordination, J. Adda, and H. Richard, (eds). Paris. Dec. 8-10, 245 p. MacLeod, A. J. and Troconis, N. G. 1982. Volatile flavour component of mango ITuit. Macleod, Macleod, Phytochemistry, 21 (10): 2523-2526. A. J. and Pieris, N. M. 1984. Comparison of the volatile components of some mango cultivars. Phytochemistry, 23 (2): 361-366. A. J. and Snyder, C. H. 1985. Volatile components of two cultivars of mango ITom Florida. Journa[ of Agricultural and Food Chemistry. 33 : 380-38. MacLeod, A. I; MacLeod, G. and Snyder, C. H. 1988. Volatile aroma constituents of mango (cv KENSINGTON). Phytochemistry, 27 (7): 2189-2193. Mendoza JR, D. B. and Wills, R. B. H. 1984. Mango: Fruit development, postharvest physiology and marketing in Asean. Malaysia: ASEAN Food Handling Bureau, III p. Olafsdottir, C. G.; Steineke, J. A. and Lindsay, R. C. 1985. Quantitative performance of simple Tenax-GC adsorption method: for use in the analysis of aroma volatiles. Journal of Food Science, 50: 1431-1436.
Pithakpo~ B., et a!. 1985. Study of canned mango fleshjuice, pickle and conserve. ASEAN WORK SHOP ON FOOD TECHNOLOGY RESEARCH AND DEVELOPMENT. Sakho, M.; Crouzet, J.; Seck, S. 1985. Volatile components of African mango. Journal of Food Science. 50: 548-550. Table I Aroma components analysed by different e>.:traction methods\n Kaew mango Relative rctention Componcnt GC peak arcau(%) time 52.0 unknown 0.25 55.0-0.78 I 2 3 56.0-2.21 61.4-0.30 67.0 α-pinene 0.93 0.81 67.8 Chloroform 0.87 68.5 methylbenzene 1.46 78.3 unknown 0.40 79.2-0.88 92.0-0.73 0.55 0.48 94.1-0.54 97.6 δ-car-ene 6.64 6.62 6.45 100.0 β-myrcene 1.46 2.22 2.32 102.0 unknown 0.73 0.64 0.81 106.0 α-tcrpinene 1.87 0.64 2.51 113.0 Jirnonene 1.87 2.14 1.64 114.0 β-phe1andrene 1.19 1.56 1.04 116.0 2-hexen-I-01 1.38 0.68-119.0 unknown 0.28 1.16-124.0 cis-ocimene 0.21 0.80-130.0 trans-ocimene 0.23 0.39 0.25 138.0 unknown 0.25 0.33 0.24 143.0 α-terpinolene 70.46 68.80 77.46 169.5 1-hexano! 0.94 2.47 0.69 182.0 3-hexene-I-ol 2.90 3.45 1.62
continued Relative retention Component GC peak area**(%) I 2 3 209.0 unknown 0.34 0.21 0.18 277.0 trans-caryophylene 1.75 0.45 304.0 α-humulene 0.19 - - 308.0 unknown 0.24-0.28 320.0 unknown 0.14. - 324.0 β-selinene 2.53 3.27 325.0 unknown 0.42-0.35 ** I: SDE, 2 : Head-spaceconcentration,3: Solventextraction ** Means of3 replications Table II Aroma components identified qualitatively ttom simuhaneous distillation-extraction method in Kaew mango Retention Relative GC-MS time retention time Compound Confinnation 7.30 60.8 α-pinene ***** 10.78 89.8 δ-car-4-ene **** 11.53 96.1 δ-car-3-ene... * 12.00 100.0 β-myrcene... 12.88 107.3 α- terpinene... 13.80 115.0 limonene... * 14.26 118.8 δ-pheuandrene... 14.75 122.9 3-hepten-I-01 ** 17.50 45.8 I, 4-methylethyl benzene... 18.45 153.8 α-terpinolene 27.65 230.4 1,2-methylethylbenzene...... * 29.62 246.8 unknown 37.54 312.8 trans -caryophyuene 41.90 349.2 α-humulene... * 42.60 355.0 β-chamigrene *.. 44.85 373.8 β-selinene...*...* 47.97 399.8 ethan-i-one 51.69 430.8 unknown 52.55 437.9 I,4-methylethenylbenzene -...
Table III Aroma components identified qualitatively from headspace concentration method in Kaew mango Retention Relative GC-MS time retention time Compound Confirmation 2.80 30.0 hexane... 3.28 36.0 unknown - 4.02 44.0 dichloromethane... 4.10 45.0 ethanol.*... 5.27 58.0 chloroform... 5.37 59.0 α-pinene... 5.70 62.0 methylbenzene... 7.85 86.0 unknown 8.53 93.9 δ-car-3-ene *.*.* 8.68 95.6 I-butanol..**** 9.08 100.0 β-myrcene... 9.70 106.0 α-terpinene *... 10.48 115.4 limonene.*** 10.87 119.7 β-phellandrene *... 11.23 123.6 3-methylbutan-I-01... 11.43 125.8 2-hexen-I-ol... 12.27 135.1 cis-ocimene **** 12.65 139.3 γ-terpinene..*** 13.00 143.2 Irans-ocimene.*** 13.85 152.5 1,3-rnethylethylbenzene...*. 14.75 162.4 2-penten-I-01 *.*.* 18.93 208.5 I-hexanol... 19.03 209.6 4,4-hydroxyrnethyl pentan-2-one... 19.57 215.5 unknown - 20.55 226.3 3-hexen-I-01...* 22.00 242 alcoholc6hi206... 23.20 255.5 I,4-methylethenylbenzene... 23.50 258.5 1,2-dichlorobenzene...** 24.42 268.9 acetic acid... 27.14 298.9 2-ethylhexan-I-01... 28.19 310.5 benzaldehyde... 34.23 377 C4H602-35.77 393 I-phenylethanone... 38.55 424.5 Irans-carveol..
Table IV The chemical class of aroma compound identified by different extraction methods. Compound I 2 Monoterpene hydrocarbon 82.55 83.32 Sesquiterpene hydrocarbon 5.62 - Alcohol 4.51 6.41 Total 92.68 89.73 Others 7.32 10.27 * I: simultaneous distillation-extraction 2: headspace concentration