2014 Volume 20 Issue 1 Pages 167-174
Free and glycosidically bound volatile compounds in different parts of Eureka lemon were studied. Free volatile compounds were extracted by solid phase microextraction (SPME). Bound volatile compounds were isolated from juice by adsorption onto an Amberlite XAD-2 column, and then hydrolyzed by almond β-glucosidase. Both the free volatile compounds and the released aglycones were analyzed by using gas chromatography-mass spectromerty (GC-MS). Gas chromatography-olfactometry (GC-O) was also used to determine the aroma-active compounds in juice and peel. Totally 29 and 34 free volatiles were found in juice and peel, respectively. Terpenes and aldehydes were the most abundant compounds in free fractions. Four and six bound volatile compounds were found in juice and peel with a total concentration of 127.8 and 1161.52 µg/L. Benzenic and terpenic compounds were the main bound volatiles found in Eureka lemon fruit. Mannose and glucose were found as sugar moiety in juice, and glucose was found the only sugar linked to bound volatiles in peel.
Eureka lemon originated in America and grew widely under tropical and sub-tropical climatic conditions, including many countries such as Australia, Israel and southern China. It is a common lemon species in the world (Ford, 1942). It is called “the true ‘better’ lemon” and has large quantity of juice and acid content.
Eureka lemon has not been got much attention. The fruit is elliptic in shape, bright yellow and medium thickness of the skin. It is more acidic than Meyer lemon, its peel is thicker, and also the juice is less than Lisbon lemon, a lemon similar to Eureka lemon. Generally speaking, Meyer lemon and Lisbon lemon are more used as juice production, cooking and eating. But the characters of Eureka lemon make it better to be a good material to produce essential oil and other aroma products. The luxuriant aroma compounds of lemon are applied in many fields. d-Limonene plays a potential role in preventing breast cancer (Miller et al., 2011), and the antioxidant activity of lemon essential oils has been confirmed (Di et al., 2010). Lemon products are also important for food industry and health care (González-Molina et al., 2010).
Volatile compounds in Meyer lemon (Moshonas et al., 1972), Italian lemon (Allegrone et al., 2006) and petitgrain Eureka lemon essential oil (Baaliouamer et al., 1985) have been reported, except Eureka lemon. Many studies showed the common volatile compounds in lemon. Several studies showed that the main volatile compounds in lemon essential oil were monoterpenes, such as limonene, γ-terpinene and β-myrcene (Fan et al., 2009a; Verzera et al., 2004). Limonene was the most abundant compound in lemon. Additionally, aldehydes, ketones and esters were the other important volatile compounds in lemon (Lota et al., 2004; Moufida et al., 2003). The free volatile compounds and particular character compounds of Eureka lemon have not been reported yet. The bound volatile fraction has also a potential for production of better processed food.
Bound volatile compounds cannot be detected directly without changing to be free. On the basis of previous research, volatile compounds can be released from glycosides by acid or enzymatic hydrolysis during maturation, storage and industrial treating (Gunata et al., 1985). So far, there are many reports about the bound volatiles in fruit, including grape (Baek and Cadwallader, 1999), mango (Pino et al., 2005), litchis (Chyau et al., 2003), pineapple (Wu et al., 1991) and so on. However, glycosidically bound volatile compounds in citrus have seldom been investigated. Gueguen has found that treatment of β-glucosidase from C. molischiana could significantly increase the levels of linalool, benzyl alcohol and 2-phenylethanol in orange juice (Gueguen et al., 1996). Bound volatile compounds have been also found in sweet oranges. Ethyl 3-hydroxyhexanoate, 3-oxo-α-ionol, p-vinylguaiacol, ethyl 3-hydroxybutanoate, benzyl alcohol have been reported to be the main bound volatile compounds in sweet oranges (Fan et al., 2009b). In general, bound volatile is different from free volatiles in fruit. This character makes bound volatiles a powerful hidden resource of aroma. Alcohols, organic acids, terpenes, C13-norisoprenoid, hydroxy esters and benzenic compounds are the main bound volatile compounds in fruit. The amount and type would vary according to part of plants, grade of maturity, and the species (Bogusz et al., 2012; Escriche et al., 2011; Mukherjee and Litz, 1997; Wirth et al., 2001). There are scarcely studies on the glycosidically bound volatiles from Eureka lemon, especially in the peel and the juice.
The objective of this paper is to study free and bound volatile compounds in peel and juice of Eureka lemon. Volatile compounds were released from glycosides by enzymatic hydrolysis. Then the aroma active compounds in Eureka lemon were determined using SPME, GC-MS and GC-O. Moreover, the sugar moieties of bound fraction were also studied.
Materials Eureka lemons were harvested in October at Dehong, Yunnan Province. After cleaning the fruit, the peel and pulp were separated by hand. The pulp was squeezed to juice by a centrifugal juice extractor then filtered and filled into three bottles. Eureka lemon had a juice yield at 33.93%. The Brix value, total acid, pH and Brix/acid ratio of fresh-squeeze juice was 8.5° Bx, 4.592%, 2.8645 and 1.85, respectively. Both the juice and the peel were preserved in freezer at −18°C.
Reagents Amberlite XAD-2 resin was treated with the method that used by Gunata et al (Gunata et. al. 1986). β-Glucosidase (7.7 u/mg) was obtained from Sigma (USA). The water used in the study was purified by a Millipore-Q system (Millipore Corp., Saint-Quentin, France). An internal standard solution (IS) of cyclohexanone, 99.5% purity (Buchs, Switzerland) in ethyl alcohol was prepared at a concentration of 0.946 mg/mL. All other chemicals were of analytical grade and used as received. Standards of n-paraffins (C6 ∼ C25) were purchased from Sigma Chemical Company. The the flavour molecules standards α-terpineol, β-pinene, carene, citral, β-myrcene, terpinolene, β-phellandrene, bisabolene, carvone, camphene, α-thujene, alloaromadedrene, ocimene, farnesene, undecanal, octanol, nonanol, nerol, eugenol, α-pinene, β-gurjurene, β-elemene, α-cedrene, methyl geranate, δ-cadinene, γ-selinene, copaene, germacrene D, tridecanal, citronellyl acetate, benzoic acid and vanillin were gifts from Shenzhen Boton Flavours & Fragrances Co., Ltd. (Shenzhen, China). α-Pinene, d-limonene, γ-terpinene, caryophyllene, valencene, nonanal, decanal, benzyl alcohol, p-vinylguaiacol and carveol were obtained from Fluka.
Extraction of free volatile compounds in juice and peel Ten milliliters of lemon juice mixed with 3.6 g NaCl and 50 µL cyclohexanone (internal standard solution) were placed into a 20 mL crimp top vial. The SPME manual device assembled with a 50/30 µm DVB/ CAR/PDMS fiber (Supelco, Bellfonte, PA, USA) was conditioned operated in GC injector port at 270°C for 1h. Each sample was equilibrated for 15 min and extracted by the fiber for 40 min at 40°C under stirring (500 rpm). After extraction, the fiber was desorbed in GC injection port for 5 min. Sixty grams of peel mixed with 240 mL distilled water was crushed into slurry. Ten milliliters of slurry was filled in a 20 mL crimp top vial. The samples were treated in the same way described above. Three parallel samples were detected.
Extraction of bound volatile compounds in juice and peel The juice prepared previously was centrifuged (10,000 g, 4°C) for 20 min and the sediment was discarded. A 50 × 1 cm i.d. column was filled with XAD-2 resin and washed by distilled water thoroughly. The prepared juice was poured in the column at 3 mL/min. Then the column was rinsed with water and diethyl ether/pentane (1:1) for removing saccharides, acid and free volatile compounds. The bound fraction was eluted by methanol, and then evaporated to dryness under vacuum at 35°C. The concentrate dissolved in 20 mL citric- phosphate buffer solution (pH 5, 0.06 M). The mixture solution was extracted twice by diethyl ether/pentane (1:1) to remove the rest traces free volatiles. Three parallel samples were detected.
The peel was cut into small cube and 50 g of them were extracted with 300 mL methanol by ultrasonic extraction method at 50°C for 2 h, and the supernatant was filtered. Then the supernatant was evaporated to dryness under vacuum at 35°C. The concentrate was dissolved in 20 mL citric-phosphate buffer solution (pH = 5, 0.06 M). The mixture solution was extracted thrice by diethyl ether/pentane (1:1) to remove the rest traces free volatiles. Three parallel samples were detected.
Enzymatic hydrolysis The bound volatile compounds of juice and peel dissolved in the buffer solution were hydrolyzed by almond β-glucosidase at 40°C for 48 h. The liberated aglycons were extracted three times with 80 mL diethyl ether/pentane (1:1). The organic phase was dried by anhydrous sodium sulfate and concentrated to 0.5 mL under a stream of pure nitrogen. Fifty microliter of cyclohexanone was added to the concentrated as an internal standard. And the aqueous phase was saved. Each three parallel samples were determined.
Derivatization of sugar moiety of volatile glycosides The precedent aqueous phase was concentrated to dryness under vacuum. Fifty milligrams of hydroxyamine hydrochloride and 2.5 mL pyridine were added, and was heated at 90°C water bath for 30 min, stirred during the heating. After cooling to room temperature, 2.5 mL acetic anhydride was added, and was heated at 90°C in a water bath for 30 min. Then it was cooled to room temperature again and added in 2.5 mL chloroform to dissolve the possible existing solid. One microliter of final solution was injected and analyzed by GC-MS. The sugar standards were treated in the same way described above.
GC and GC-MS analysis of volatile compounds Volatile compounds were analyzed by Agilent 6890N GC coupled to an Agilent 5973 mass spectrometer and equipped with a J&W DB-5MS fused silica capillary column (30 m × 0.25 mm i.d., 0.25 µm film thickness). The column temperature was held at 40°C for 12 min, increased to 108°C at 3°C/min, held for 2 min, then increased to 250°C at 5°C/min, and held for 5 min. The carrier gas (helium) flow rate was 1.0 mL/min, the injector temperature was 250°C and the ionization chamber temperature was 230°C. Mass spectra were recorded at 70 eV in electron impact (EI) mode. The injected sample amount was 1 µL. For glycoside derivatives, injector and FID detector temperatures of GC were 250 and 280°C, respectively. The column temperature was raised from 160 to 180°C at 10°C/min, then raised to 210°C at 3°C/ min, and held for 4 min.
Identification of compounds detected by GC-MS analysis was done by comparing mass spectra and retention indices (RI) with the authentic standards and published data, as well as by comparing their mass spectra with the MS library of Wiley7.0 and Nist05. Retention indices were calculated using a mixture of n-paraffin C6 ∼ C25 as standards.
Semi-quantitative determinations were obtained by using cyclohexanone as an internal standard. Volatile compounds content was calculated from the GC-peak areas relating to the GC-peak area of the internal standard.
Sensory analysis and Gas Chromatography—Olfactometry (GC-O) An olfactory detector port 2 (Gerstel) was used for both free and bound volatile compound analysis. GC-MS is equipped also with olfactometry device. Three trained testers analyzed each sample. The average age of these testers is 21, including one man and two women. The odorous volatile compounds were analyzed according to the RT and intensity in GC-O test.
Statistical analysis The statistical significance of the free and bound volatile compounds in juice and peel was determined by one-way ANOVA. Statistical analyses were done with SPSS.
Free and bound volatile compounds in juice Table 1 shows the free and bound volatile compounds released by enzymatic hydrolysis in the Eureka lemons juice. A total of 22 terpenes, 4 aldehydes, 2 alcohols and 1 ester, with a total content of 123 mg/L. Compared to other lemon juice, more abundant volatile compounds were detected in this Eureka lemon (Allegrone et al. 2006). The amount of these compounds in free forms was remarkably higher than those in the bound forms. Terpenes were the main components with 96.7% of these free volatile compounds. In these terpenes, d-Limonene (65%) was the most abundant volatile compound, and the follows were γ-terpinene (10%), α-terpineol (8%), β-pinene (5%) and carene (3%). d-limonene was far higher than the other free volatile compound and was usually a main compound in citrus. It has a citrus and mint aroma with a threshold of 10 ppb (Fazzalari, 1978). In the present study, the content of d-limonene was more than 80 ppm, and it was described as “lemon aroma” in GC-O test. Nevertheless, limonene would not be the most crucial fraction in Eureka lemon, since the most abundant compound did not have major influence of citrus flavour, and the trace concentrations might be the characteristic of flavour (Rouseff et al., 2009).
| RI | Name | Odor | Intensity | Content | ID | |
|---|---|---|---|---|---|---|
| Free (mg/L) | Bound (µg/L) | |||||
| 923 | α-Thujene | Mint | 1 | 0.1 ± 0.02 | n.d. | a |
| 930 | α-Pinene | Lemon | 1 | 1.2 ± 0.18 | n.d. | a |
| 945 | Camphene | Sour | 1 | 0.1 ± 0.01 | n.d. | a |
| 974 | β-Pinene | Wood | 2 | 6.1 ± 0.04 | n.d. | a |
| 990 | β-Myrcene | Lemon | 2 | 1.9 ± 0.00 | n.d. | a |
| 1003 | α-Phellandrene | Lemon, Flower, Milk | 2 | 0.7 ± 0.02 | n.d. | a |
| 1033 | d-Limonene | Lemon | 2 | 80 ± 0.7 | n.d. | a |
| 1043 | Ocimene | None | 0 | 0.02 ± 0.00 | n.d. | a |
| 1049 | Carene | Lemon, Mint | 3 | 3.7 ± 0.02 | n.d. | a |
| 1060 | γ-Terpinene | Lemon, Alcohol, Mint | 1 | 12 ± 0.17 | n.d. | a |
| 1060 | Benzyl alcohol | None | 0 | n.d. | 38 ± 1.27 | a |
| 1073 | Octanol | Alcohol, Lemon, Sour | 1 | 0.79 ± 0.02 | n.d. | a |
| 1087 | Terpinolene | Citrus, Lemon | 2 | 1.2 ± 0.01 | n.d. | a |
| 1105 | Nonanal | Flower, Orange | 1 | 0.63 ± 0.02 | n.d. | a |
| 1174 | Nonanol | None | 0 | 0.14 ± 0.00 | n.d. | a |
| 1188 | α-Fenchene | None | 0 | 0.04 ± 0.02 | n.d. | b |
| 1194 | α-Terpineol | Lemon | 2 | 9.5 ± 0.05 | n.d. | a |
| 1205 | Decanal | Lemon | 2 | 0.24 ± 0.00 | n.d. | a |
| 1239 | Farnesene | None | 0 | 0.03 ± 0.01 | n.d. | a |
| 1246 | Carvone | Meat | 1 | 0.2 ± 0.02 | n.d. | a |
| 1261 | Nerol | Flower | 1 | n.d. | 66 ± 42 | a |
| 1273 | Citral | Mint, Lemon | 3 | 2.2 ± 0.12 | n.d. | a |
| 1307 | Undecanal | None | 0 | 0.04 ± 0.00 | n.d. | a |
| 1324 | p-Vinylguaiacol | None | 0 | n.d. | 2.5 ± 0.41 | a |
| 1337 | Isoterpinolene | None | 0 | 0.02 ± 0.01 | n.d. | b |
| 1354 | Citronellyl formate | None | 0 | 0.04 ± 0.03 | n.d. | b |
| 1420 | Caryophyllene | None | 0 | 0.1 ± 0.00 | n.d. | a |
| 1436 | α-Bergamotene | Lemon, Sour | 1 | 0.2 ± 0.05 | n.d. | b |
| 1457 | Eugenol | Sweet | 2 | n.d. | 20 ± 15.21 | a |
| 1461 | β-Santalene | None | 0 | 0.01 ± 0.00 | n.d. | b |
| 1494 | Valencene | Citrus | 2 | 0.07 ± 0.00 | n.d. | a |
| 1498 | Alloaromadedrene | None | 0 | 0.04 ± 0.03 | n.d. | a |
| 1509 | Bisabolene | None | 0 | 0.3 ± 0.00 | n.d. | a |
RI, Linear retention index on DB-5MS column; n.d., not detected; ± standard deviation;
ID, Identification: a, comparison of mass spectra and retention index with authentic standards;
Intensity, 0, 1, 2 and 3 represent none, weak, medium and strong.
γ-Terpinene was the second major compound in Eureka lemon. And it was also a major compound in citrus such as grapefruit and orange (Steuer et al., 2001). γ-Terpinene smelt like turpentine and gasoline and had lemon, alcohol and mint odor at 12.9 ppm in GC-O, and it was considered to be the major contributor to fresh lemon flavour (Schieberle and Grosch, 1989). α-Terpineol has an oil, anise, mint smell and the threshold was 330-350 ppb (Fazzalari, 1978). Its formation was dependent on pH and temperature. It is reported that limonene was the precursor of α-terpineol (Haleva-Toledo et al., 1999). Moreover, the content of α-terpineol was higher in lemon juice than other citrus juice. It is indicated that limonene influenced the α-terpineol content in lemon juice. In the GC-O test, α-terpineol smelt like lemon.
β-Pinene was also detected in Eureka lemon juice. Its threshold was 140 ppb (Fazzalari, 1978) and had a pine, resin, turpentine odor. β-Pinene widely existed in most fruits as a volatile compound, such as mango, citrus, mushrooms, fermentation and pepper (Breheret et al., 1997; Jirovetz et al., 2002; Lalel et al., 2003; Yoo et al., 2001). In the GC-O test, it was found to have a wood like aroma.
Terpinolene detected in the present study had a citrus and lemon like aroma with a concentration of 1.29 mg/L. It is indicated that it was an important contributor to the total aroma of the Eureka lemon juice. The same result was also obtained by Schieberle & Grosch (Schieberle and Grosch, 1989).
Carene, β-myrcene, terpinolene, α-bergamotene and α-phellandrene were found at relatively low concentrations. These compounds smelt like flowers, lemon and sour. They were considered to be important compounds influencing the entire aroma in lemon juice. In addition, there were also a lot of compounds detected by GC-MS but could not be distinguished by GC-O analysis such as caryophyllene, alloaromadedrene, farnesene.
Citral, nonanal, decanal and undecanal were the four aldehydes detected in the free forms, accounting for 2.5% of the total free volatile compounds in this fruit. Among these aldehydes, citral, nonanal and decanal was detected by GC-O analysis. It is reported that citral is the major contributor to the flavour and aroma of lemon oil and lemon essence (Dugo et al., 2002). Nonanal and decanal smelt like flower, lemon and orange. These compounds had relative low odor thresholds of 1 ppb and 0.2 ∼ 1 ppb, respectively (Fazzalari, 1978). They were not only important in oranges (Fan et al., 2009a) but also in Eureka lemon.
Only two alcohols were found in lemon juice, they were octanol and nonanol with a total concentration of 0.9 mg/L. Octanol had a threshold of 110 ∼ 130 ppb, and nonanol had a threshold of 50 ppb (Fazzalari, 1978). But in the present paper, nonanol was not distinguished by GC-O analysis, while octanol were perceived with an alcohol, sour and lemon odor.
According to the bound volatile compounds in lemon juice, only 4 bound volatiles were detected, including 3 benzenic compounds and 1 terpenic compound with a total concentration of 127.8 µg/L. None of them existed both in free and bound forms.
Nerol was the most abundant compound in the bound fraction, which was the only terpenic compound found as bound form at the level of about 66 ppb with a flower note. It is reported that the threshold of nerol was 300 ppb (Fazzalari, 1978). While the GC-O analysis showed that nerol had a flower aroma. It is indicated that the release of nerol might contribute to the total aroma of the lemon juice.
Benzyl alcohol, eugenol and p-vinylguaiacol were the three benzenic compounds detected in lemon juice. p-Vinylguaiacol was detected at a relative low content of 2.52 µg/L, which is lower than its threshold (10 ppb). This compound was also found in oranges as bound form with a concentration of about 100 ppb (Fan et al., 2009b). Benzyl alcohol was found at the level of about 38 ppb in the bound fraction, but it had a relative high threshold value of 5500 ppb (Fazzalari, 1978). The odors of them were reported as clove, curry and sweet, flower, respectively. But they were not recognized in GC-O test. This indicated that the release of these two bound volatile compounds might not be the contributor to the entire aroma of this juice. Eugenol was the other bound volatile compound detected by GC-O with a sweet and clove aroma. The content of this compound was 20.26 µg/L and was higher than the threshold of 6 ∼ 30 ppb. It is indicated that eugenol might be an important contributor to the lemon flavor after release.
To sum up, these bound volatile compounds were considered as ubiquitous in many other plants. While the contents of these compounds were found not be abundant and none of them existed as free forms in the lemon juice. Eugenol and nerol were considered to be the probable contributors to the total aroma of the lemon juice due to their relative high concentrations. After released from the glycosides, they provided a mild sweet odor.
Free and bound volatile compounds in peel As shown in Table 2, a total of 34 free volatile compounds were detected in the peel, including 27 terpenes, 5 aldehydes, and 2 esters. Similar to the juice, terpenes (98%) was also the most abundant group in free fractions of the peel. Among these terpenes, limonene was also the predominant compound and it took up 66% of the free volatile compounds, following by β-pinene (13%), γ-terpinene (12%) and β-myrcene (2%). This result was also obtained in many other studies, and these compounds were easily to degrade by oxidation and UV (Nguyen et al., 2009). Among these free volatiles, totally 20 compounds were also found in juice. It could be found that the amounts of these common compounds in peel were obviously higher than that in juice. And there were 14 free volatiles detected only in peel. GC-O test showed that many of these different compounds were not perceived by the testers. It is indicated that many of these compounds might not be active to human scent and they were not important contributors to the total aroma.
| RI | Name | Odor | Intensity | Content | ID | |
|---|---|---|---|---|---|---|
| Free (mg/L) | Bound (µg/L) | |||||
| 923 | α-Thujene | Sour, Flower | 1 | 65 ± 6 | n.d. | a |
| 929 | α-Pinene | Sweet, Fruit, Grass | 1 | 562 ± 46 | n.d. | a |
| 995 | β-Myrcene | Orange | 3 | 598 ± 43 | n.d. | a |
| 979 | β-Pinene | Wet wood | 1 | 3790 ± 185 | n.d. | a |
| 1057 | d-Limonene | Lemon | 3 | 18572 ± 135 | n.d. | a |
| 1064 | Ocimene | Fruit | 2 | 68 ± 8 | n.d. | a |
| 1071 | Linalool oxide | Sweet, Fruit, Alcohol | 1 | n.d. | 31 ± 3 | b |
| 1078 | γ-Terpinene | Lemon, Flower | 3 | 3470 ± 60 | n.d. | a |
| 1087 | α-Terpineol | Lemon | 1 | n.d. | 18 ± 9 | a |
| 1096 | Carene | Lemon, Flower | 2 | 270 ± 6 | n.d. | a |
| 1101 | Terpinolene | Lemon, Sweet | 2 | 1.2 ± 0.03 | n.d. | a |
| 1109 | Nonanal | Lemon, Sweet | 2 | 99 ± 6.40 | n.d. | a |
| 1167 | Carveol | Roast | 2 | 16 ± 1.2 | n.d. | a |
| 1185 | Limonene oxide | None | 1 | 27 ± 2 | n.d. | b |
| 1195 | Sylvestrene | Lemon | 1 | 77 ± 0.1 | n.d. | b |
| 1204 | Benzoic acid | None | 0 | n.d. | 703 ± 406 | a |
| 1207 | Decanal | Lemon, Roast | 1 | 46 ± 1.5 | n.d. | a |
| 1225 | cis-Verbenol | None | 0 | n.d. | 12 ± 3 | b |
| 1278 | Citral | Lemon | 2 | 366 ± 9 | n.d. | a |
| 1308 | Undecanal | Pungent, Sweet, flower | 1 | 18 ± 0.6 | n.d. | a |
| 1325 | Methyl geranate | None | 0 | 2.2 ± 0.1 | n.d. | a |
| 1350 | Copaene | None | 0 | 0.7 ± 0.1 | n.d. | a |
| 1354 | Citronellyl acetate | Coconut, flower | 1 | 8 ± 0.2 | n.d. | a |
| 1381 | 8-Hydroxy linalool | Lemon, Alcohol | 1 | n.d. | 194 ± 48 | b |
| 1393 | β-Elemene | None | 0 | 2.7 ± 0.06 | n.d. | a |
| 1403 | α-Fenchene | Pomelo | 1 | 0.1 ± 0.03 | n.d. | b |
| 1409 | Tridecanal | None | 1 | 4 ± 0.03 | n.d. | a |
| 1416 | α-Bergamotene | Lemon | 1 | 7.7 ± 0.6 | n.d. | b |
| 1421 | Vanillin | Mint, Lemon, Flower, Sweet | 1 | n.d. | 152 ± 85 | a |
| 1422 | Caryophyllene | None | 1 | 93 ± 5 | n.d. | a |
| 1429 | Alloaromadendrene | Lemon | 2 | 1.2 ± 0.2 | n.d. | a |
| 1457 | Farnesene | Lemon, Flower | 1 | 7.5 ± 0.3 | n.d. | a |
| 1480 | α-Cedrene | None | 0 | 2.7 ± 0.1 | n.d. | a |
| 1495 | Valencene | None | 0 | 17 ± 0.6 | n.d. | a |
| 1503 | Bisabolene | Citrus | 1 | 14 ± 1.7 | n.d. | b |
| 1511 | β-Gurjurene | Mint | 1 | 132 ± 2 | n.d. | a |
| 1519 | α-Panasinsen | None | 0 | 1.5 ± 0.09 | n.d. | b |
| 1525 | δ-Cadinene | None | 0 | 2.1 ± 0.2 | n.d. | a |
| 1648 | Germacrene D | None | 0 | 0.1 ± 0.02 | n.d. | a |
| 1671 | γ-Selinene | Lemon | 1 | 1.2 ± 0.1 | n.d. | a |
RI, Linear retention index on DB-5MS column; n.d., not detected; ± standard deviation;
ID, Identification: a, comparison of mass spectra and retention index with authentic standards;
Intensity, 0, 1, 2 and 3 represent none, weak, medium and strong.
Although most of the free volatile compounds in the juice and peel were the same, such as limonene, γ-terpinene, β-pinene and β-myrcene, the different percentage of each compound might cause the characteristic aroma. It made the entire aroma largely identical but with minor differences between the juice and peel. For instance, the percentage of limonene in total free volatile compounds of juice and peel were 65% and 66%, respectively. The percent of citral in juice and peel were 1.7% and 1.3%, and the percent of β-myrcene in juice and peel were 1.5% and 2.1%. These compounds made the similarities of aroma between peel and juice. However, it could be found that many compounds had very different percents in juice and peel, such as γ-terpinene (10% and 12% in juice and peel), 3-carene (2.9% and 1.4%), α-pinene (0.96% and 1.97%). And this might be the main reason for the difference on the total aroma between the juice and peel.
According to the GC-O test, it elucidated that the compounds in free forms of the peel, which was mainly described as fruit and mint, such as ocimene and β-gurjurene, distinguished from those of the juice. Other compounds were described as pungent, such as undecanal. The amount of each kind of free volatiles in peel could also indicate that the aroma of peel was more intense.
Aldehydes were the second dominant compounds in free forms of the peel, which were in total 176 mg/kg. They are secondary metabolites formed during ripening and will increase with fruit maturity (Perez-Cacho and Rouseff, 2008). Nonanal, decanal, and undecanal could be found in the free fractions of juice. Furthermore, the GC-O results showed that they were all active compounds to human scent in peel with low threshold (nonanal at 1 ppb, decanal at 0.1 ∼ 2 ppb, and undecanal at 5 ppb) (Fazzalari, 1978), and they provided a strong sweetness and fragrance of flowers.
Methyl geranate and citronellyl acetate were the two esters detected in the free fractions of the peel. Citronellyl acetate smelt like coconut or sweet flower by GC-O analysis, while methyl geranate was not perceived by the panelists.
Totally 6 bound volatile compounds were identified in peel, including 4 terpenic compounds and 2 benzenic compounds. The aroma of these glycosidically bound volatiles extracted from peel presented a more various and mellow odor than that of the juice. After the release of the bound volatiles, the total concentration would be in total about 1,110 µg/kg. It was greatly higher than that of the juice. None of these compounds was found as free forms in peel.
Benzoic acid and vanillin were the two benzenic compounds in the bound fractions. Although benzoic acid was the most abundant bound compound in peel, it was not detected in GC-O test due to its high threshold. It is a general compound in many plants or fermentation products. Vanillin detected in the present study is also an aroma compound in common use. It is reported that it was an important compound influencing the aromas of many citrus fruits such as orange, lemon, lime (Goodner et al., 2000). It had a sweet, flower odor, and their threshold was 20 ∼ 200 ppb. It was the third highest amount compounds in bound fraction, so it would have a potential contribution to enhance the total aroma.
Linalool detected in this study was reported as a positive compound in citrus juice (Selli and Kelebek, 2011). 8-Hydroxy linalool and cis-linalool oxide detected in this study had a lemon and sweet fruit aroma in GC-O test, respectively. α-Terpineol was detected both in the free fraction of juice and the bound fraction of peel, and the amount of the former was greatly larger than that of the latter. It was a catabolite of limonene and linalool, smelt like lemon peel, existing in many fruit and plants. Thus, if in the other grade of maturity, it might be greater than that in present.
To sum up, compared to the free volatile compounds in peel, the levels of bound volatile compounds were much lower. Furthermore, the amount of many bound volatile compounds in peel or juice was lower than its threshold. And some of the glycosidically bound volatile compounds were detected at relative high levels. The release of these potential compounds would be positive to the total aroma of Eureka lemon. Benzene derivatives were abundant both in juice and peel. The same results were also obtained in many other studies. Contrast to other fruits, lemon had relative low pH value and low concentrations of glycosidically bound volatile compounds. It is reported that bound volatile compounds could be hydrolyzed at its acidic conditions
Sugar Moieties of the Bound Volatile Fraction Monosaccharide was the only glycoside found in this study. Mannose and glucose were proved to be sugar moiety of glycosidically bound volatile compounds in juice of Eureka lemon, while glucose was the only sugar moiety in peel.
Free and bound volatile compounds in juice and peel of Eureka lemon were studied. The amount of volatile concentration in peel and juice had great differences. Either free or bound volatile compounds in peel were much more than in juice. Terpenes constitute the main framework of the entirely aroma. Glycosidically bound volatile compounds were totally different from free fractions both in juice and peel. These results suggest the potential ability to enhance the aroma during lemon processing.
Acknowledgement We thank Yunnan Academy of Agricultural Sciences for providing the lemon samples. This work was supported by National Natural Science Foundation of China (Program No. 31101239), National Science and Technology Support Program (2012BAD31B10-6), Fundamental Research Funds for the Central Universities (Program No. 2013PY097) and the Chenguang Program of Wuhan Science and Technology Project (Program No. 201271031404).
