LWT - Food Science and Technology 44 (2011) 1946e1951

Contents lists available at ScienceDirect

LWT - Food Science and Technology

Response surface optimization and characteristics of rambutan (Nephelium lappaceum L.) kernel fat by hexane extraction
Wanrada Sirisompong, Wannee Jirapakkul, Utai Klinkesorn*
Department of Food Science and Technology, Faculty of Agro-Industry, Kasetsart University, 50 Ngam Wong Wan Road, Chatuchak, Bangkok 10900, Thailand

a r t i c l e i n f o
Article history: Received 12 October 2010 Received in revised form 7 April 2011 Accepted 26 April 2011 Keywords: Rambutan kernel Fat extraction Response surface methodology Fat characteristics

a b s t r a c t
Response surface methodology (RSM) was used to study the effect of moisture content (1.59e18.41 g/ 100 g), extraction time (2.3e10.7 h) and particle size (0.09e2.11 mm) on the fat yield from rambutan kernels using hexane extraction. The physical and chemical characteristics of rambutan fat were also determined. The optimum conditions obtained from response surface analysis was 4.99 g/100 g moisture, 1.05 mm particle size and 9.2 h extraction time. Under these optimum conditions, the maximum fat yield was 37.35 g/100 g. The extracted fat was a white solid at room temperature. The physical and chemical characteristics of the extracted fat compared well with those of conventional fats. The high level of arachidic acid (w 34.3 g/100 g fat) and low iodine value in rambutan kernel fat permits the use of the fat, especially where oxidation may be a concern, without its being subjected to hydrogenation. 2011 Elsevier Ltd. All rights reserved.

1. Introduction Rambutan (Nephelium lappaceum Linn.) is a seasonal fruit native to west Malaysia and Sumatra. It is cultivated widely in Southeast Asian countries and Thailand has become the leading producer (0.6e0.7 million tons a year) and exporter (w 10e15 million dollars US) of the fruit (Salakpetch, 2000). This fruit is generally consumed fresh, although in the main producing countries like Malaysia and Thailand it is industrially processed to obtain juice, jams, jellies and marmalades (Morton, 1987). In addition, rambutan fruits are also processed as rambutan stuffed with a chunk of pineapple and canned in syrup. The rambutan fruits are deseeded during processing and these seeds (w 4e9 g/100 g) are a waste by-product of the canning industry (Tindall, 1994). Some studies have reported that rambutan seed possesses a relatively high amount of fat with values between 14 g/100 g and 41 g/100 g (Augustin & Chua, 1988; Kalayasiri, Jeyashoke, & Krisnangkura, 1996; Morton, 1987; Sols-Fuentes, Camey-Ortz, Hernndez-Medel, Prez-Mendoza, & Durn-de-Baza, 2010; Winayanuwattikun et al., 2008), and the increasing demand for oils and fats, whether for human consumption or for industrial purposes, necessitates the search for new sources of novel oils and fats. Therefore, the extracted fat from rambutan seed not only could be used for manufacturing candles, soaps, and fuels, it also has a potential to be a source of natural edible fat with possible industrial use.
* Corresponding author. Tel.: 662 562 5031; fax: 662 562 5021. E-mail address: [email protected] (U. Klinkesorn). 0023-6438/$ e see front matter 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.lwt.2011.04.011

The main process for the separation and recovery of oils and fats from seeds is solvent extraction. The fat yield from the seeds depends on the nature of the solvent, the extraction temperature and time, seed particle size and pretreatment conditions (Becker, 1978). Recently, a little information has become available on rambutan fat extraction (Azam, Waris, & Nahar, 2005; SolsFuentes, Camey-Ortz, Hernndez-Medel Mdel, Prez-Mendoza, & Durn-de-Baza, 2010). However, no work has been reported on the effect of moisture content, extraction time and particle size on the extraction efciency of fat from rambutan seed kernels. In order to study the effects of independent variables on fat extraction, central composite design (CCD) with response surface methodology (RSM) is often used (Mani, Jaya, & Vadivambal, 2007; Shao, Sun, & Ying, 2008; Tan, Jinap, Edikusnadi, & Hamid, 2008; Wei, Liao, Zhang, Liu, & Jiang, 2009). RSM has increasingly been used for optimizing purposes due to its efciency and lower data requirements. This experimental methodology consists of a group of mathematical and statistical procedures that can be used to study the relationships between one or more responses and a number of independent variables. It also denes the effect of the independent variables, alone or in combination, in the process. Beside analysis the effects of independent variables, RSM creates a mathematical model that accurately describes the overall process (Hu, 1999; Montgomery; 2001). The main objective of the present study was to determine and explain the effects of moisture content, extraction time and particle size on the yield of rambutan kernel fat using hexane extraction. A model equation that would predict and determine the optimum conditions for total fat yield was

W. Sirisompong et al. / LWT - Food Science and Technology 44 (2011) 1946e1951

1947

developed, and as well, the physical and chemical characteristics of the extracted fat were also determined. 2. Materials and methods 2.1. Materials Rambutan seeds (N. lappaceum L.) were obtained from Universal Food Public Company Limited (Nakornpathom, Thailand). The seeds were washed and the kernels were removed manually from the seeds. The kernels were then ground and dried using a tray dryer (Kan Seng Lee Machinery (1960) Ltd. Part, Thailand) at 55  C for 5 h (w 5 g/100 g moisture). The prepared kernels were stored at 5  C until used for experiments. The selected extraction solvent was hexane which was purchased from Mallinckrodt Baker, Inc. (Phillipsburg, NJ). All other chemicals were reagent grade or higher. 2.2. Sample preparation and extraction process Based on preliminary experimental results, the chosen independent variables were moisture content, extraction time and particle size. The moisture content of ground kernels was adjusted by slowly adding sprayed water while the ground kernels were continuously mixed in an Imarex mixer (IF 309, Thailand). These samples were left sealed overnight at 4  C to reach equilibrium (Sandoval & Barreiro, 2007). The moisture content of the rambutan kernel samples was determined in triplicate according to the air oven method AOAC 931.04 (AOAC, 2000). The ground particles were separately by size using an AS 200 Control sieve shaker (Retsch, Germany) and different particle sizes ranging from 0.09 to 2.11 mm were obtained. The particles sizes based on sieve numbers used for analysis were: 10 (2.0e2.36 mm), 12 (1.7e2.0 mm), 18 (1.0e1.8 mm), 35 (0.5e0.6 mm) and 170 (0.09e1.1 mm). A lab scale Soxhlet apparatus was used to extract fat from the rambutan kernels. About 20 g of ground and dried kernels was used for each combination of process parameters. The amount of solvent used for fat extraction was 150 ml for each sample. The extraction temperature was approximately 65  C. The extracted fat was expressed as a percentage, which is dened as the weight of the extracted fat over the dry weight of the sample. 2.3. Characterization of rambutan kernel fat 2.3.1. Color, refractive index and crystal polymorphism The color of the extracted fat was measured using a Miniscan XE (Hunter Association Laboratory Inc., USA) according to the modied method of Cheikh-Rouhou et al. (2007). The refractive index of kernel fat was determined using an instrument from Atago Co. Ltd. Series No. 11506, Japan. The crystal polymorphism of the fat sample was analyzed using an X-ray diffractometer (JEOL JDX-3530, Japan) and the modied method of Reshma, Saritha, Balachandran, and Arumughan (2008). 2.3.2. Iodine value, sponication number and unsponication matter The iodine value of fat samples was determined according to AOCS Cd 1d-92 (AOCS, 1997). The sponication number and unsponication matter of rambutan fat samples were measured according to AOCS Cd 3e25 and AOCS Ca 6a-40, respectively. 2.4. Analysis of fatty acid composition The fatty acids were hydrolyzed from fat and then methylated to fatty acid methyl esters (FAMEs). The fatty acid composition of fat was investigated using gas chromatography (Model 6890N

G1530N, Agilent Technologies, Inc., USA). The gas chromatography apparatus was equipped with a Supelco SP-2560 capillary column 100 m 0.25 mm id with the lm thickness of 0.2 mm (Supelco Inc., USA) and FID detector, and operated in a split mode with a split ratio of 100:1. The injector and detector temperatures were 250  C. The column temperature was held at 140  C for 5 min, and then programmed to rise to 250  C at 3  C/min and then held for 17 min. The carrier gas used was helium set at a ow rate of 1.1 ml/min. The area percents were used to determine the relative amounts of each fatty acid. The response factors for each fatty acid according to AOAC 996.06 method were used to correct area percents. The fatty acid content was reported as grams per 100 g of fat. 2.5. Analysis of phytosterol content The phytosterol content in rambutan fat was determined according to the method of Schwartz, Ollilainen, Piironen, and Lampi (2008). Gas chromatography analysis was conducted using an HP 6890 (Hewlett Packard Co., USA) instrument equipped with an HP-5 capillary column 30 m 0.32 mm id with the lm thickness of 0.25 mm and FID detector, and operated in a splitless mode. The injector and detector temperatures were 300  C. The column temperature was held at 245  C for 1 min, and then programmed to rise to 275  C at 3  C/min and then held for 20 min. The carrier gas used was helium set at a ow rate of 3 ml/min. The relative amount of each phytosterol was reported as milligrams per gram of fat. 2.6. Analysis of a-tocopherol content The a-tocopherol in rambutan fat was measured using an Agilent 1100 Series reverse-phase high performance liquid chromatography apparatus with a diode array detector (Agilent Technologies, Inc., USA). A Hypersil ODS column (250 4.0 mm, Alltech Associates, Inc., USA) was used as the stationary phase. The mobile phase with this column was 98 percents methanol and 2 percents water and a ow rate of 1 ml/min was used. Column temperature was maintained at 25  C and the a-tocopherol was detected at a wavelength of 292 nm. 2.7. Thermal behavior of rambutan fat The thermal behavior experiment was conducted using a differential scanning calorimeter (Model DSC1, Mettler-Teledo International Inc., USA) and a method modied from Saloua, Eddine, and Hedi (2009) and Besbe, Blecker, Deroanne, Drira, and Attia (2004). Approximately 4e5 mg of the unrened fat sample was placed in an aluminum sample pan and an empty aluminum pan was placed on the reference platform. A linear heating and cooling rate of 5  C/min over a temperature range of 50 to 90  C was used. The thermogram peak was used to provide an estimate of enthalpy (DH) and the thermogram peak points were used to determine the melting point. 2.8. Experimental design and statistical analysis The effect of three independent variables: moisture content (X1; 1.59e18.41 g/100 g), extraction time (X2; 2.3e10.7 h) and particle size (X3; 0.09e2.11 mm) on the response variable (Y, percent fat yield) was evaluated using a three factor central composite design (CCD). The ve coded levels of moisture content, extraction time and particle size were incorporated into the design and were analyzed in 20 combinations (Table 1). The central point of the design was repeated six times to calculate the reproducibility of the method (Montgomery; 2001). For each combination of the independent variables in the experimental design, the dependent parameter percent for extracted fat was determined. The effect of

1948

W. Sirisompong et al. / LWT - Food Science and Technology 44 (2011) 1946e1951

Table 1 Uncoded and coded levels of independent variables in the experimental design and percent extracted fat. Treatment Moisture runs content (X1, g/100 g) Extraction time (X2, hours) Particle size (X3, mm) Extracted fat (Y, g/100 g)a

3.2. Response surface analysis The estimated regression coefcient of the response surface models for the extracted rambutan kernel fat along with the corresponding coefcient of determination values (R2) and lack of t test are shown in Table 2. The linear coefcient indicates that the fat yield is positively correlated with all independent variables. On the other hand, the quadratic coefcient of all independent variables shows a negative correlation with extracted fat yield. The R2 and adjusted R2 were 0.974 and 0.967, respectively. These results imply that the response surface model can explain more than 96 percents of the variation in the response variables studied. Therefore, the R2 values of the response models are sufciently high, to indicate that the response surface model is workable and can be used for estimation of the mean response and the subsequent optimization stages. The lack of t, which measures the tness of the model, was found to be non signicant (p > 0.05), indicating that the number of experiments were sufcient for determining the effect of independent variables on percentage fat yield (Montgomery; 2001). As shown in Table 2, the extracted fat yield was signicantly (p < 0.05) inuenced by the main effects of moisture content, extraction time and particle size and the quadratic effects of all independent variables, as well as the interaction effect between moisture content and particle size and extraction time and particle size. The following response surface model (Eq. (2)) was t to the three independent variables (X1, X2 and X3) of the response variables (Y):
2 2 Y 24:97 1:39X1 1:31X2 5:56X3 0:08X1 0:12X2 2 4:76X3 0:53X1 X3 0:77X2 X3

Uncoded Coded Uncoded Coded Uncoded Coded 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 10 5 10 10 15 10 10 5 10 15 10 5 18.41 10 15 10 1.59 10 5 15 0 10.705 1 4 0 6.5 0 6.5 1 9 0 6.5 0 6.5 1 9 0 6.5 1 4 0 6.5 1 9 1.682 6.5 0 2.295 1 9 0 6.5 1.682 6.5 0 6.5 1 4 1 4 1.682 1 0 0 1 0 0 1 0 1 0 1 0 1.682 1 0 0 0 1 1 1.1 0.5 1.1 1.1 1.7 2.109 0.091 1.7 1.1 0.5 1.1 0.5 1.1 1.1 0.5 1.1 1.1 1.1 1.7 1.7 0 1 0 0 1 1.682 1.682 1 0 1 0 1 0 0 1 0 0 0 1 1 35.53 35.74 34.89 33.57 23.13 23.58 34.52 35.40 33.63 29.83 34.75 35.39 20.63 28.17 32.46 33.88 35.34 34.48 30.07 18.86 0.47a 0.38a 0.40a 0.24ab 1.48d 0.63d 0.55ab 0.84a 0.66ab 0.73c 0.09a 0.62a 1.11e 1.49c 1.34b 0.68ab 1.10a 1.09ab 0.61c 1.63e

a Means standard deviation having the same letters are not signicantly different at the 5% level.

these independent variables X1, X2 and X3 on the response Y was investigated using the second-order polynomial regression equation with stepwise elimination. This equation, derived using RSM for the evaluation of the response variables, is as follows:
2 2 2 Y b0 b1 X1 b2 X2 b3 X3 b11 X1 b22 X2 b33 X3

(2)

3.3. Effect of independent variables on percent extracted fat

b12 X1 X2 b13 X1 X3 b23 X2 X3

(1)

where Y is the response variable (fat yield, g/100 g); bo, b1, b2, b3, b11, b22.. are regression coefcients and X1, X2 and X3 are uncoded values for moisture content, extraction time and particle size, respectively. An analysis of variance (ANOVA) was performed to determine the lack of t and the effect of linear, quadratic and interaction terms on fat extraction. The analysis of data and the optimizing process were generated using Minitab statistical software v.15 (Minitab Inc., USA). 3. Results and discussion 3.1. Percent extracted fat Fat from rambutan kernels (N. lappaceum L.) was extracted with hexane using a Soxhelt apparatus. The effect of different levels of the independent variables namely moisture content, extraction time and particle size on percent fat yield was determined and the results are summarized in Table 1. The yields of extracted fat ranged from 18.86 to 35.74 g/100 g with a mean value of 31.19 g/100 g. Augustin and Chua (1988) analyzed the seed composition of rambutan grown in Malaysia and reported that the seed contained 37.1e38.9 g/100 g crude fat (petroleum ether extracted), which is higher than the value (33.4 g/100 g) reported by other researchers in Mexico (Sols-Fuentes et al., 2010). In Thailand, it has been shown that rambutan seed contains a wide range of fat, between 14 and 41 g/100 g (Kalayasiri et al., 1996; Winayanuwattikun et al., 2008). Those differences in the fat content of the seed may be attributed to the variability of the studied cultivars, a diversity in the maturity of the seeds used and the agricultural conditions in the area cultivated (Augustin & Chua, 1988; Cheikh-Rouhou et al., 2007).

The response surface plot of fat yield using hexane for various combinations of moisture content, extraction time and particle size is shown in Fig. 1. The percent fat yield was higher at lower moisture content and vice versa (Fig. 1a). Similarly, the fat yield was higher for smaller particle size and vice versa (Fig. 1b). The fat yield decreased when the moisture content and particle size increased. Higher moisture content samples have more resistance to penetration by the solvent into the samples, which resulted in low fat yield. As the particle size decreased, the surface area increased and this enhanced fat extraction, resulting in a higher fat yield (Shahidi & Wanasundara, 2002). The fat yield increased in particles up to 1.1 mm and then gradually decreased as particle size increased. As shown in Table 2, the fat yield was signicantly (p < 0.05) inuenced by the interaction effect of moisture content and particle
Table 2 Regression coefcients and p-value of the response surface models and statistical analysis. Regression Term Extracted Fat (Y, g/100 g) Coefcient Intercept (b0) Moisture (b1) Time (b2) Particle size (b3) Moisture2 (b11) Time2 (b22) Particle size2 (b33) Moisture/Particle size (b13) Time/Particle size (b23) R2 R2 (adj) Regression Lack-of-t
a

Probability (p-value)a 0.000 0.000 0.004 0.004 0.000 0.000 0.000 0.000 0.000 e e 0.000 0.122

24.9732 1.3885 1.3091 5.5557 0.0836 0.1157 4.7579 0.5284 0.7722 0.9743 0.9673 e e

The p-value more than 0.05 is not signicantly different at the 5% level.

W. Sirisompong et al. / LWT - Food Science and Technology 44 (2011) 1946e1951

1949

content and with medium particle size but with a long extraction time (Fig. 1). To determine the exact optimum points for all the independent variables necessary to achieve the optimized condition, a numerical optimization was utilized. The results showed that the extraction using 1.05 mm particles at 4.99 g/100 g moisture for 9.2 h of extraction time provided the maximal fat yield. Under these optimum conditions, the predicted fat yield was 37.35 g/100 g. The adequacy of the response surface equations was demonstrated by a comparison between the experimental values and predicted values based on a response regression (Mirhosseinia, Tan, Hamid, & Yusof, 2008). Under the recommended optimum conditions, the experimental value for fat yield was 37.20 0.69 g/100 g which was not signicantly different (p > 0.05) from the predicted value (37.35 g/100 g). Good agreement must exist between the values calculated using the model equations and the experimental values at the points of interest, and no signicant (p > 0.05) difference was reported between the actual and the predicted values (Predicted value 1.0016 Experimental value). The high correlation coefcients (z0.978) also conrmed that a close agreement between experimental data and predicted values calculated using the models had been obtained. The closer the experimental and predicted results, the better they explain the adequacy of the regression equation (Rossa, de S, Burin, & Bordignon-Luiz, 2011). 3.5. Physical and chemical characteristics of rambutan kernel fat 3.5.1. Color The rambutan kernel fat is consistently a white solid at room temperature (25 2  C). The L*, a* and b* values of solid fat were 86.87, 2.06 and 3.55, respectively (Table 3). When the solid fat was heated (w 60  C), the melted fat had a golden yellow color with L*, a* and b* values of 66.34, 2.31 and 6.62, respectively. The L*, a* and b* values of other vegetable oils, such as palm, soybean, sunower, olive, and corn range from 63.4 to 69.5, 3.8 to 4.4 and 9.2 to 10.4, respectively (Hsu & Yu, 2002). This shows that the melted fat has b* values lower than those of other vegetable oils. It may suggest the presence of less yellow pigments, e.g. carotenoids, in rambutan kernel fat (Cheikh-Rouhou et al., 2007).
Fig. 1. Response surface plots showing the interaction effects of extraction variables on the extracted fat yield; a extraction time-moisture effect, b particle size-moisture effect, and c particle size-extraction time effect.

3.5.2. Crystal polymorphism In regard to polymorphic forms of rambutan kernel fat as determined by XRD, it can be seen that the extracted fat contains

size. From the results (Fig. 1b) it is apparent that with a smaller particle size (less than 1.1 mm) the fat yield slightly increased with an increase in moisture content up to w 10 g/100 g and the fat yield decreased when the moisture content was further increased. On the other hand, for the bigger particle sizes (more than 1.1 mm) the fat yield slightly decreased with increasing moisture content and when the moisture content was increased from 5 to 15 g/100 g a marked decrease in fat yield was observed. In Fig. 1c, the fat yield was increase with an increase in extraction time and did not increase after 6.5 h for small particle size (w 0.5 mm). At intermediated particle size (w 1.1 mm), the fat yield increased as the extraction time was increased and reached a maximum at 9 h. Any further increase in extraction time did not increase the fat yield. 3.4. Optimization and validation of regression model Response optimizations were performed to measure the optimum levels of independent variables required to achieve the desired fat yield. For hexane extraction, the process providing the maximal fat yield would involve samples with low moisture

Table 3 Physical and chemical characteristics of rambutan kernel fat. Properties Color (CIE Lab) Solid fat Melted fat Polymorphic forms (g/100 g) Beta (b) Beta-prime (b0 ) Refractive index Iodine value (g/100 g oil) Saponication value (mg KOH/g oil) Unsaponiable matter (g/100 g) Phytosterol (mg/g) Stigmasterol b-Sitosterol Campesterol a-Tocopherol (mg/100 g)
a b

Valuea L* 86.87 0.97; a* 2.06 0.48; b* 3.55 0.35 L* 66.34 0.18; a* 2.31 0.46; b* 6.62 0.56 84.7 1.2 15.3 1.2 1.469 0.001 41.6 1.2 166 3 0.19 0.04 0.32 0.03 0.61 0.06 NDb 0.103 0.001

Mean standard deviation of duplicate analysis. ND Not detected.

1950

W. Sirisompong et al. / LWT - Food Science and Technology 44 (2011) 1946e1951 Table 4 Fatty acid composition of rambutan kernel fat. Fatty acid composition Saturated fatty acid Myristic acid (C14:0) Palmitic acid (C16:0) Stearic acid (C18:0) Arachidic acid (C20:0) Heneicosanoic acid (C21:0) Behenic acid (C22:0) Tricosanoic acid (C23:0) Lignoceric acid (C24:0) Monounsaturated fatty acid Palmitoleic (C16:1u7) Trans-9-Elaidic acid (C18:1u9t) Cis-9-Oleic acid (C18:1u9c) Erucic acid (C22:1u9) Polyunsaturated fatty acid Cis-9,12-Linoleic acid (C18:2u6) a-Linolenic acid (C18:3u3) Cis-11,14-Eicosadienoic acid (C20:2)
a

mixtures of b (84.7 g/100 g) and b0 (15.3 g/100 g) polymorphic forms (Table 3). The structure, composition and polymorphic forms of fatty crystals are most important criteria for determining the functional properties of fat (Reddy & Jeyarani, 2001). From the results, it appears that rambutan kernel fat tends to crystallize in b form which is also seen in cocoa butter. For cocoa butter, a representative confectionery fat, the most functional polymorph (form V) is a b type, since this form results in optimal density, melting behavior, and surface appearance (Sato, 1999). Therefore, rambutan kernel fat has the potential to be used for confectionery products. 3.5.3. Refractive index, iodine value and saponication number The refractive index of rambutan kernel fat is 1.469 0.001 (Table 3). This value is consistent with the values of other vegetable oils (Padley, Gunstone, & Harwood, 1986) and the results of Sols-Fuentes et al. (2010), who reported that rambutan fat had a refractive index of 1.468. The iodine value of the fat, 41.6 1.2 g per 100 g fat (Table 3), places it in the non-drying group of oils which also included palm oil, palm kernel oil and coconut oil (Tan & Che Man, 2000). This value is similar to the previous studies (Azam et al., 2005; Sols-Fuentes et al., 2010), which reported that rambutan fat had an iodine value of about 44e47 g per 100 g fat. The saponication number, which is an indication of the average molecular weight of the fat, was 166 3 mg KOH per gram fat for rambutan kernels (Table 3). The low saponication value suggests that rambutan fat contains a long chain fatty acid and a relatively high average molecular weight (Onyieke & Acheru, 2002). 3.5.4. Unsaponication matter The estimated unsaponication matter in rambutan kernel fat found to be 0.19 0.04 g/100 g, as compared to 0.2e0.6 g/100 g in cocoa butter (Padley et al., 1986). In the present study, the phytosterol along with a-tocopherol content were also determined (Table 3). The results show that the rambutan kernel fat had 0.61 0.06 and 0.32 0.03 mg per gram fat of b-sitosterol and stigmasterol, respectively. The fat from rambutan kernel has lower level of total phytosterol than do commercial oils and fats (Padley et al., 1986). The extracted fat also contained 0.103 mg per 100 g fat of a-tocopherol. The low content of a-tocopherol in rambutan kernel fat was similar to that of cocoa butter, cod liver oil and beef fat (Swern, 1964). These results suggest that rambutan kernel fat may not be a good source of the natural antioxidants, phytosterol and a-tocopherol. 3.6. Fatty acid composition The fatty acid composition of rambutan kernel fat is described in Table 4. The total saturated and unsaturated fatty acids found in rambutan fat were 49.6 and 45.9 g/100 g, respectively. The most abundant fatty acids in rambutan seed fat was arachidic acid (C20:0) for the saturated fatty acids (34.3 g/100 g), and oleic acid (C18:1) was the main unsaturated fatty acid (36.8 g/100 g). These two fatty acids comprised more than 70 g/100 g of the total fatty acids, which is in agreement with previously published data (Augustin & Chua, 1988; Azam et al., 2005; Sols-Fuentes et al., 2010). Likewise, measurable amounts of palmitic (C16:0), stearic (C18:0) and behenic (C22:0), for saturated fatty acids as well as palmitoleic (C16:1), linoleic (C18:2n-6) and linolenic (C18:3n-3) for unsatutated fatty acids were detected (Table 4). The high level of arachidic acid is the main reason for the low iodine value in rambutan seed fat (Table 3) would allow the fat to be used without being subjected to hydrogenation, especially where autoxidation may be a concern.

Content (g/100 g)a 49.57 0.02 4.69 7.03 34.32 0.05 3.10 0.03 0.33 37.97 0.49 0.03 36.79 0.66 7.89 1.37 6.48 0.04 0.14 0.00 0.15 0.08 0.01 0.00 0.04 0.01 0.06 0.22 0.04 0.00 0.16 0.03 0.01 0.02 0.03 0.00

Mean standard deviation of duplicate analysis.

3.7. Thermal properties of rambutan kernel fat The thermal behavior, melting and crystallization of rambutan kernel fat is shown in Fig. 2a. These results show three well-dened peaks with shoulders that correspond to groups of triglycerides with high, middle, and low temperatures of melting and crystallization. These results illustrate the complex nature of triglycerides in fat samples (Tan & Che Man, 2000). For the melting behavior, the rst peak had the lowest melting temperatures, with the temperature peak at 4.2  C. The results suggest that this region includes a group of triglycerides with a greater abundance of unsaturated fatty acids. The second and third peaks represent intermediate and high melting temperatures, with a peak temperature at 37.4 and 49.8  C, respectively. These regions might be represented by groups

Fig. 2. Melting and crystallization curves of rambutan kernel fat (a) and commercial cocoa butter (b).

W. Sirisompong et al. / LWT - Food Science and Technology 44 (2011) 1946e1951

1951

of triglycerides with more saturated fatty acids. Similar results were observed for crystallization behavior (Fig. 2a) where the peak temperatures were 40.9, 19.1 and 16.1  C for the rst, second and third peaks, respectively. The beginning and completed phase changes were determined based on the onset and offset temperature values of the peaks. The crystallization process for rambutan fat began with the rst crystal formation at 37.8  C and ended at 14.8  C. The total crystallization enthalpy was 92.1 J/g. For the melting process, the beginning temperature was 4.5  C and the phase change was completed at about 58.9  C. The total value of enthalpy for the melting process was 85.4 J/g. These values indicate a large range of melting points for the rambutan fat studied, which is slight different from previously published data (Sols-Fuentes et al., 2010). It is often difcult to compare the thermal prole of oils and fats from various sources because of the lack of uniformity of analytical techniques used in qualitative or quantitative analysis (Tan & Che Man, 2000). Based on the results of crystal polymorphism analysis in the present study (Table 3) which reported that rambutan kernel fat, like cocoa butter, tends to crystallize in b form, the thermal behavior of commercial cocoa butter (Siam Cocoa Products Co Ltd., Thailand) was investigated under the same analytical conditions used for extracted rambutan fat. The commercial cocoa butter showed only one peak and a limited melting range which indicates a homogeneity of triglycerides (Fig. 2b). Its melting and crystallization occurred at the peak temperatures of 36.25  C and 13.24  C, respectively. Similarities in the thermal behaviour were observed for the second region of rambutan fat and cocoa butter. These results suggest that this fraction of rambutan kernel fat has the potential to be used as a cocoa butter replacer for confectionery products. 4. Conclusions The seed kernel of rambutan fruit is a typical waste material obtained annually in large amounts as a by-product of the canning industry and hexane can be used successfully as a solvent for extraction of fat from rambutan seed kernels. A moisture content of 4.99 g/100 g, particle size of 1.05 mm and extraction time of 9.2 h provided the maximum fat yield of 37.35 g/100 g. The rambutan seed kernels give a considerable yield of fat and the high arachidic acid content makes the fat highly stable to oxidation. Because of these physical and chemical characteristics, the merit of using rambutan kernel fat in the cosmetic and food industries may be justied. However, through the seeds are considered edible after roasting in some Asian countries, the safety of rambutan kernel fat must still be evaluated before it is used as an ingredient in the food industry. Acknowledgment Funds for this research were provided by the Kasetsart University Research and Development Institute (KURDI) with the project No. V-T(CH) 2.6.51. References
AOAC. (2000). Ofcial methods of analysis of AOAC International (17th ed.). Maryland: AOAC. AOCS.. (1997). Ofcial methods and recommended practices of the AOCS (5th ed.). Champaign, IL: AOCS. Augustin, M. A., & Chua, B. C. (1988). Composition of rambutan seeds. Pertanika, 11, 211e215. Azam, M. M., Waris, A., & Nahar, N. M. (2005). Prospects and potential of fatty acid methyl esters of some non-traditional seed oils for use as biodiesel in India. Biomass & Bioenergy, 29, 293e302.

Becker, W. (1978). Solvent extraction of soybeans. Journal of the American Oil Chemists Society, 55, 754e761. Besbe, S., Blecker, C., Deroanne, C., Drira, N. E., & Attia, H. (2004). Date seeds: chemical composition and characteristic proles of the lipid fraction. Food Chemistry, 84, 577e584. Cheikh-Rouhou, S., Besbes, S., Hentati, B., Blecker, C., Deroanne, C., & Attai, H. (2007). Nigella sativa L.: chemical composition and physiochemical characteristics of lipid fraction. Food Chemistry, 101, 673e681. Hsu, S. Y., & Yu, S. H. (2002). Comparisons on 11 p.ant oil fat substitutes for low-fat Kung-wans. Journal of Food Engineering, 51, 215e220. Hu, R. (1999). Food product design: A computer-aided statistical approach. Boca Raton, Florida: CRC Press. Kalayasiri, P., Jeyashoke, N., & Krisnangkura, K. (1996). Survey of seed oils for use as diesel fuels. Journal of the American Oil Chemists Society, 73, 471e474. Mani, S., Jaya, S., & Vadivambal, R. (2007). Optimization of solvent extraction of moringa (Moringa Oleifera) seed kernel oil using response surface methodology. Trans IChemE, 85(C4), 328e335. Mirhosseinia, H., Tan, C. P., Hamid, N. S. A., & Yusof, S. (2008). Optimization of the contents of Arabic gum, xanthan gum and orange oil affecting turbidity, average particle size, polydispersity index and density in orange beverage emulsion. Food Hydrocolloids, 22, 1212e1223. Montgomery, D. C. (2001). Design and analysis of experiments (5th ed.). New York: John Wiley and Sons. Morton, J. F. (1987). Rambutan. In J. F. Morton (Ed.), Fruits of warm climates (pp. 262e265). Miami, Florida: Julia F. Morton. Onyieke, E. N., & Acheru, G. N. (2002). Chemical composition of selected Nigerian oil seeds and physiochemical properties of the oil extracts. Food Chemistry, 77, 431e437. Padley, F. B., Gunstone, F. D., & Harwood, J. L. (1986). Occurrence and characteristics of oils and fats. In F. D. Gunstone, J. L. Harwood, & F. B. Padley (Eds.), The lipid handbook (pp. 49e112). New York: Chapman and Hall. Reddy, S. Y., & Jeyarani, T. (2001). Trans-free bakery shortenings from mango kernel and mahua fats by fractionation and blending. Journal of the American Oil Chemists Society, 78, 635e640. Reshma, M. V., Saritha, S. S., Balachandran, C., & Arumughan, C. (2008). Lipase catalyzed interesterication of palm stearin and rice bran oil blends for preparation of zero trans shortening with bioactive phytochemicals. Bioresource Technology, 99, 5011e5019. Rossa, P. N., de S, E. M. F., Burin, V. M., & Bordignon-Luiz, M. T. (2011). Optimization of microbial transglutaminase activity in ice cream using response surface methodology. LWT - Food Science and Technology, 44, 29e34. Salakpetch, S. (2000). Rambutan production in Thailand. In Proceedings of the international tropical fruit conference (pp.13e17). Florida: Hawaii Tropical Fruit Growers. Saloua, F., Eddine, N. I., & Hedi, Z. (2009). Chemical composition and prole characteristics of Osage orange Maclura pomifera (Ran.) Schneider seed and seed oil. Industrial Crops and Products, 29, 1e8. Sandoval, A. J., & Barreiro, J. A. (2007). Off-line capillary rheometry of corn starch: effects of temperature, moisture content and shear rate. LWT-Food Science and Technology, 40, 43e48. Sato, K. (1999). Solidication and phase transformation behaviour of food fats a review. Fett/Lipid, 101, S.467eS474. Schwartz, H., Ollilainen, V., Piironen, V., & Lampi, A. M. (2008). Tocopherol, tocotrienol and plant sterol contents of vegetable oils and industrial fats. Journal of Food Composition and Analysis, 21, 152e161. Shahidi, F., & Wanasundara, J. P. D. (2002). Extraction and analysis of lipids. In C. C. Akoh, & D. B. Min (Eds.), Food lipids, chemistry, nutrition, and biotechnology (pp. 115e136). New York: Marcel Dekker. Shao, P., Sun, P., & Ying, Y. (2008). Response surface optimization of wheat germ oil yield by supercritical carbon dioxide extraction. Food and Bioproducts Processing, 86, 227e231. Sols-Fuentes, J. A., Camey-Ortz, G., Hernndez-Medel, R., Prez-Mendoza, F., & Durn-de-Baza, C. (2010). Composition, phase behavior and thermal stability of natural edible fat from rambutan (Nephelium lappaceum L.) seed. Bioresource Technology, 101, 799e803. Swern, D. (1964). Structure and composition of fats and oils. In D. Swern (Ed.), Baileys industrial oil and fat products (4th ed.). (pp. 3e53) New York: John Wiley and Sons. Tan, C. P., & Che Man, Y. B. (2000). Differential scanning calorimetric analysis of edible oils: comparison of thermal properties and chemical composition. Journal of the American Oil Chemists Society, 77, 143e155. Tan, T.-J., Jinap, S., Edikusnadi, A., & Hamid, N. S. A. (2008). Extraction of cocoa butter by supercritical carbon dioxide: optimization of operating conditions and effect of particle size. Journal of Food Lipids, 15, 263e276. Tindall, H. D. (1994). Rambutan cultivation. Rome, Italy: Food and Agricultural Organization of the United Nation. Wei, Z.-J., Liao, A.-M., Zhang, H.-X., Liu, J., & Jiang, S.-T. (2009). Optimization of supercritical carbon dioxide extraction of silkworm pupal oil applying the response surface methodology. Bioresource Technology, 100, 4214e4219. Winayanuwattikun, P., Kaewpiboon, C., Piriyakananon, K., Tantong, S., Thakernkarnkit, W., Chulalaksananukul, W., et al. (2008). Potential plant oil feedstock for lipase-catalyzed biodiesel production in Thailand. Biomass & Bioenergy, 32, 1279e1286.