Supercritical carbon dioxide extraction and studies of mango seed kernel for cocoa butter analogy fats

Abstract

Supercritical carbon dioxide (SC-CO₂) was introduced for obtaining premium grade cocoa butter quality fat from the waste of mango seed kernel (MSK), where the Soxhlet method was also used for the comparison. Six different varieties of MSK were selected to be extracted using SC-CO₂ at pressures of 35 and 42.2 MPa, temperatures of 60°C and 72°C, and constant CO₂ flow rate at 3.4 ml/min. The total fat contents of MSK varieties ranged from 64 to 135 g/kg at SC-CO₂ extraction and from 76 to 137 g/kg at Soxhlet extraction methods. The fatty acid constituents of fat yield of all varieties extracted using SC-CO₂ ranged from 6.9% to10.9% palmitic acid, 32.8% to 47.6% stearic acid, 37% to 47.3% oleic acid, and 3.7% to 6.9% linoleic acid. However, the physicochemical properties and fatty acid constituents of SC-CO₂ extracted MSK fats were found to be comparable to that of commercial cocoa butter.

Con el fin de obtener una grasa equivalente a la mantequilla de cacao de primera calidad a partir del desperdicio del núcleo de la semilla de mango (NSM), se introdujo dióxido de carbono supercrítico (CO₂-SC) en dicho núcleo. Asimismo, con fines de comparación, se usó el método de extracción Soxhlet. Se seleccionaron seis variedades diferentes de nsm para realizar su extracción mediante el proceso CO₂-SC a presiones de 35 y 42,2 MPa, a temperaturas de 60°C y 72°C, con un flujo constante de CO₂ de 3,4 ml/min. Con la extracción mediante CO₂-SC, los contenidos de grasa totales de las variedades de nsm fluctuaron entre 64 y 135 g/kg, mientras que usando los métodos de extracción Soxhlet, mostraron una variación de entre 76 y 137 g/kg. Los componentes de ácidos grasos de la grasa de todas las variedades que fueron extraídas usando CO₂-SC, mostraron las siguientes variaciones: el ácido palmítico aumentó de 6,9 a 10,9%, el ácido esteárico lo hizo de 32,8 a 47,6%, el ácido oleico de 37 a 47,3% y el ácido linoleico de 3,7 a 6,9%. Sin embargo, las propiedades físico-químicas y los componentes de ácidos grasos de grasas de nsm extraídas a través del método CO₂-SC fueron comparables a los de la mantequilla de cacao comercial.

Keywords:

• mango seed kernel fat
• cocoa butter quality fat
• supercritical carbon dioxide extraction
• fatty acid constituents
• physicochemical properties

Palabras claves:

• grasa del núcleo de la semilla de mango
• grasa de calidad de mantequilla de cacao
• extracción mediante dióxido de carbono supercrítico
• componentes de ácidos grasos
• propiedades físico-químicas

Introduction

Mango (Mangifera indica L.) is one of the most commercially important and popular tropical fruits around the world. In some countries, it is known as the ‘king of fruits’. It is well acknowledged that mango fruits and its processed products have gained popularity in both the national and international market due to their nutritional quality, delicious, succulence, sweet taste, and delicate flavor. According to Food and Agriculture Organization (FAO, 2004), mango fruit occupied fifth rank in total world production among all major fruit crops. Recently, Solís-Fuentes and Durán-de-Bazúa (2011) reported that about 28.5 million metric tons of mango fruits are produced and traded in the world annually. Commercially, mangoes are produced in over 103 countries; among them India, China, Thailand, Mexico, Indonesia, Pakistan, Brazil, Philippines, Nigeria, Egypt, and Australia are the major mango growing countries contributing about 85% of the world’s mango production (FAO, 2007; Vasanthaiah, Ravishankar, & Mukunda, 2007). From 2003 to 2005, India was the largest producer of mango fruit, accounting for 38.6% of the world production followed by China (12.9%), Thailand (6.2%), Mexico (5.5%), Indonesia (5.3%), Pakistan (4.5%), Brazil (4.3%), Philippines (3.5%), Nigeria (2.6%), and Egypt (1.3%) (FAO, 2007). During that period, India dominated the world mango export trade followed by Mexico, Brazil, and Pakistan. Malaysia imported 1,22,330 metric tons of mangoes from 2001 to 2004 (Department of Agriculture Malaysia, 2009).

The mango fruit is not only widely used for eating purposes but also used for the manufacture of juice, jam, and jelly. Loelillet (1994) reported that about 20% of the mango fruits are used to produce puree, nectar, leather, pickles, and canned slices which have gained worldwide popularity. Thus, huge amount of mango seeds are discarded as agricultural waste which are produced either by industrial processing or after consumption of the fruit, amounting to 35–60% of the total fruit weight (Puravankara, Boghra, & Sharma, 2000). In another study, Abdalla, Darwish, Ayad, and El-Hamahmy (2007) reported that more than one million of mango seeds are discarded as agricultural waste annually. Depending on the mango varieties, the seed represents 10–25% of the weight of the total fruit and the kernel represents 45–85% of the seed and about 20% of the whole fruit (Arogba, 1997; Hemavathy, Prabhakar, & Sen, 1988; Solís-Fuentes & Durán-de-Bazúa, 2011). On a dry basis, mango seed kernel (MSK) contains about 71–150 g/kg crude fat (Abdalla et al., 2007; Ali, Gafur, Rahman, & Ahmed, 1985; Gunstone, 2011). MSK fat is the rich sources of palmitic (C_16:0), stearic (C_18:0), and oleic acids (C_18:1) containing about 3–18%, 24–57%, and 34–56%. The other major fatty acids are linoleic, C_18:2 (up to 13%), linolenic, C_18:3 (0.6–1.0%), arachidic, C_20:0 (1.7–2.6%), and myristic, C₁₄ (0.5–8%) acids.

Currently, MSK fats are extracted and fractionated using solvent extraction methods (mostly Soxhlet extraction), and their qualities are evaluated by many researchers in the literature (Abdalla et al., 2007; Ali et al., 1985; Gaydou & Bouchet, 1984). Although Soxhlet extraction has successfully been used by many researchers, it produced large amounts of hazardous solvent wastes and it is non-selective, costly and requires long extraction time. However, the greater concern over the disposal of such toxic organic solvents and their effect on the human health as well as on the environment has led to a move toward cleaner extraction methods such as SC-CO₂ extraction (Hultin, 1994; Staby & Mollerup, 1993). The SC-CO₂ extraction is a method of choice for the extraction and fractionation of edible natural fats and oils from various plant sources. Over the last 20 years, SC-CO₂ has been well acknowledged as a promising alternative to organic solvent extraction method in the field of natural fats and oils. The major merits of SC-CO₂ method is a lack of solvent residue left in the final products and better retention of valuable components (Asep et al., 2008; Herrero, Cifuentes, & Ibanez, 2006; Jahurul et al., 2012a, 2012b; Zaidul, Norulaini, Omar, & Smith Jr., 2007). Carbon dioxide is used as a solvent due to its nontoxic, non-flammable, inexpensive, and clean solvent, which offers great opportunities for complex separation problem. Until now, no studies have been conducted for the extraction of MSK fat using SC-CO₂ method.

In recent years, MSK fat extracted from MSK has attracted a considerable interest of scientists. This interest is due to their unique physical and chemical characteristics, in addition to their fatty acid constituents which are similar to those of cocoa butter, illipe, shea, kokum, and sal butter (Hemavathy et al., 1988; Muchiri, Mahungu, & Gituanja, 2012). Another most important advantage of MSK fat is that it contains no trans fatty acids which are responsible for the development of various diseases and have several adverse effects on human health (Solís-Fuentes & Durán-de-Bazúa, 2011). However, it is a promising and safe natural source of edible fats. The aim of the present study is to extract the total fat from several MSK varieties using SC-CO₂ extraction methods (continuous and soaking methods) to achieve the highest yield that might be use in formulations for cocoa butter substitute. The triglycerides in terms of fatty acid constituents and physicochemical properties such as iodine value, saponification value, acid value, and slip melting point of MSK fats extracted using SC-CO₂ methods were also studied. Thus, the SC-CO₂ extracted MSK fat could be used as blending components as cocoa butter substitute.

Materials and methods

Materials and chemicals

Six different mango varieties, namely, Apple, Arumanis, Chokonan, Elephant, Sala, and Waterlily were obtained from Balic Pulau Penang and local weight markets, Malaysia. Commercial liquefied carbon dioxide (purity, 99.9%), was purchased from Malaysian Oxygen Ltd., Penang, Malaysia. The n-hexane, AR Grade (Merck, Germany), cyclohexane, glacial acetic acid, iso-propanol, ethanol, potassium hydroxide, potassium iodide free from iodine or iodate AR Grade, Wij’s solution for iodine number analysis (QREC, Asia Sdn Bhd, Malaysia), sodium thiosulphate, potassium dichromate (Bendosen Laboratory chemicals, Malaysia), hydrochloric acid (HACH, Germany), starch (Bendosen Laboratory chemicals), and phenolphthalein indicator (R & M Essen, UK) were purchased from Penang, Malaysia.

Preparation of MSK for experiment

The MSKs were manually separated from the pulp and washed with water. The MSKs were stored at –18°C, and then freeze-dried (FreeZone⁶, Labconco Corporation, Kansas City, Missouri 64132, USA). The dried samples were ground into powder (particle size <200 μm) and then kept in desiccator until further use for experimental analysis.

Determination of moisture content

The determinations of moisture content of the MSK powders were carried out using the PORIM test method No. p^5.2 (1995). The range of moisture content of MSKs was found to be 61–75 g/kg.

Extraction of MSK fats using Soxhlet method

To obtain total yield, the Soxhlet extraction was carried out for comparison with SC-CO₂ methods. The Soxhlet extraction was performed in triplicates for about 15 g of MSK powder with 250 ml n-hexane for 8 h. The solvent was evaporated from the extracted fat using rotary vacuum evaporator (Buchi RE 121, Germany) and then dried in an oven at 45°C for 2 h, weighted and kept at –20°C ready for GC analysis.

SC-CO₂ extraction methods

The supercritical fluid extraction (SFE) equipment consisted of a SC-CO₂ extractor (ISCO, Inc., Lincoln, NE; Model SFX 220; two extraction vessels), a high pressure syringe pump (ISCO, Inc.; Model 100 DX) with a maximum capacity of 69 MPa, a chiller (B/L-730, YIH DER, Taipei) for CO₂ liquefaction, and a carbon dioxide cylinder. The volume of each extraction cell was 2.5 ml. The SC-CO₂ extractor is equipped with a heated capillary restrictor (ISCO, Inc.) with an outer diameter of 50 µm and a maximum operating temperature of 150°C. The software (ISCO, Inc.; Model SFX 200) which is integrated into the SC-CO₂ extractor system was used to control the pressures and temperatures. Four grams of ground MSK samples were loaded into two extraction vessel and were then placed in the heating extractor unit to equilibrate the operating temperature. The continuous and soaking methods of SC-CO₂ extraction were carried out at pressures of 35 and 42.2 MPa, at temperatures of 60°C and 72°C, and at constant CO₂ flow rate of 3.4 ml/min. When pressurization was initiated, the CO₂ from the cylinder was passed through the chillier at 0°C and pumped into the heated extraction vessel by a high-pressure pump. The fat was collected from fat-rich CO₂ using an expansion valve and a heated capillary restrictor. As per the experimental design, the fat yield was weighed using analytical balance in certain time interval. For ‘continuous method’, the system was programmed with desired pressures, temperatures, and CO₂ flow rates for 5 h continuously. At that condition, the system was pressurized and CO₂ was flowed through the chillier at 0°C and pumped into the heated extraction vessel by a high-pressure pump for 5 h continuously. For ‘soaking method’, at given temperatures (60°C and 72°C) and pressures (35 and 42.2 MPa), the sample was soaked with CO₂ in the vessel for 3 h and extracted continuously over 2 h. The yields were collected at the end of the extraction. The total fat yields were defined in g/kg of MSK sample on dry basis as described in the following equation:

Fatty acid methyl ester

The total MSK fat yields extracted by soaking method of SC-CO₂ and Soxhlet extractions were analyzed to determine the fatty acid constituents using a gas chromatography with flame ionization detector (GC-2010 Plus + AOC-5000, Shimadzu, Osaka, Japan). The test samples were melted at 60–70°C and homogenized before taking the samples. According to the PORIM Test Methods No. p^3.4 (1995), the fatty acid methyl esters (FAMEs) were prepared prior to GC injection. Approximately, 50 mg of the test sample was measured into a 2 ml vial, then 0.95 ml of n-hexane was added and the mixture was shaken to dissolve the fat. About 0.05 ml sodium methoxide was added to the vial and shaken vigorously for 5–8 s with vortex mixture (Janke & Kunkel, VF2, Germany). The mixture first gone clear and turbid as sodium glyceroxide was precipitated. After a few minutes, 1 µl clear upper layer of the fatty acid methyl ester was taken and injected into GC for fatty acid analysis using PORIM Test method No. p^3.5 (1995). The BPX70 (70% Cyanopropyl polysilphenylene-siloxane, 30 m in length, with a 0.25 µm film coating, 0.25 mm ID, SGE France) capillary column was used to determine the fatty acid constituents. The initial oven temperature was set at 155°C, held for 3 min, and increased with a heating rate of 2°C/min until 180°C and 4°C/min up to the final temperature of 220°C. The injector and detector temperatures were retained at 250°C, respectively. Helium was used as carrier gas at a flow rate of 1.29 ml/min. The resultant FAME chromatograms were identified based on the elution order of the reference standard Supelco 37 component FAME mixtures (Sigma-Aldrich, Supelco, Bellefonte, PA., USA). The mass of the FAMEs were determined using PORIM Test Methods No. p^3.5 (1995). It was based on the percentage represented by the area of the corresponding peak relative to the sum of the areas of all the peaks. To get better accuracy, a correction factor K_i, Equation (2), was used to convert the percentage of peak area into weight percentage of the FAME components (Equation (3)). A correction factor was determined with the help of a chromatogram derived from the analysis of the FAME reference standard under operating conditions identical with those used for the sample.

Where K_i is the correction factor of component i, m_i is the weight percentage of component i in the FAME standard solution, and A_i is the area under the peak corresponding to component i.

Determination of physicochemical properties of MSK fats

The determination of iodine value, saponification value, acid value, and slip melting point (SMP) were all carried out according to PORIM test method No. p^3.1-2 and p^4.2 (1995).

Results and discussion

Total fat in different MSKs

The total fat content of various mango varieties obtained by Soxhlet extraction are shown in Table 1. The values are given as mean standard deviation of three replications. On dry basis, the fat contents in MSK varied from 76 to 137 g/kg. The highest yields were found in Apple mango (137 g/kg) and Sala mango (121 g/kg) followed by Waterlily (117 g/kg), Elephant (105 g/kg), and Arumanis mango varieties (87 g/kg). On the other hand, the lowest fat was found in Choconan (76 g/kg), which is one and half fold lower than that of Apple mango variety. Our findings are good in line with the findings of Lakshminarayana, Rao, and Ramalingaswamy (1983), Gaydou and Bouchet (1984), and Ali et al. (1985), who reported the fat contents in different MSK varieties. The fat obtained from Chokonan mango variety (76 g/kg) was lower compared with the value reported by Abdalla et al. (2007) for the mixed varieties of Zebda, Balady, and Succary (123 g/kg). The possible explanation of these variations of the total fat contents in MSK may be due to the different mango varieties and geological condition of the regions. However, the MSK fats found in Apple (137 g/kg) and Sala mango kernel (121 g/kg) were comparable with the figures reported by Lakshminarayana et al. (1983) for Fernandis (126 g/kg), Muchiri et al. (2012) for Kent (104 g/kg) and Boribo (100 g/kg), and Ali et al. (1985) for Lakhanvoge mango varieties (100 g/kg). Moreover, the fat contents of Waterlily (117 g/kg), Elephant (105 g/kg), and Arumanis mango varieties (87 g/kg) were also similar as compared to those results reported by Lakshminarayana et al. (1983) for Lohera (119 g/kg), Solís-Fuentes and Durán-de-Bazúa (2004) for Manila (94 g/kg), and Muchiri et al. (2012) for Dodo (97 g/kg) and Kagege mango varieties (85 g/kg).

Table 1. Total fats extracted from MSK using continuous and soaking techniques of SC-CO₂ extraction and Soxhlet method.

Table 1. Grasas totales extraídas de NSM usando técnicas de extracción CO₂-SC continuas y en remojo, además del método Soxhlet.

		Total MSK fat^b (g/kg)
		Continuous		Soaking		Soxhlet
Varieties^a	Pressures	60°C	72°C	60°C	72°C
Waterlily	35 MPa	101.2 ± 0.12	101.8 ± 0.11	108.9 ± 0.08	110.8 ± 0.11	117.2 ± 0.18
	42.2 MPa	106.1 ± 0.08	106.9 ± 0.08	114.4 ± 0.1	115.1 ± 0.12
Sala	35 MPa	102.1 ± 0.11	102.9 ± 0.12	113.0 ± 0.11	114.3 ± 0.14	121.0 ± 0.11
	42.2 MPa	112.3 ± 0.12	113.1 ± 0.14	117.8 ± 0.09	119.1 ± 0.17
Apple	35M Pa	123.1 ± 0.1	124.4 ± 0.09	132.0 ± 0.08	132.8 ± 0.11	137.0 ± 0.15
	42.2 MPa	130.1 ± 0.09	130.8 ± 0.1	134.3 ± 0.1	135.1 ± 0.1
Elephant	35 MPa	91.5 ± 0.1	92.1 ± 0.09	98.1 ± 0.09	99.8 ± 0.18	105.5 ± 0.09
	42.2 MPa	97.7 ± 0.12	99.3 ± 0.08	102.5 ± 0.12	103.1 ± 0.11
Arumanis	35 MPa	74.3 ± 0.09	75.2 ± 0.11	79.5 ± 0.1	80.1 ± 0.09	87.1 ± 0.08
	42.2 MPa	80.5 ± 0.08	81.0 ± 0.13	83.9 ± 0.08	84.6 ± 0.1
Chokonan	35 MPa	64.4 ± 0.1	65.0 ± 0.09	68.2 ± 0.1	68.8 ± 0.14	76.2 ± 0.1
	42.2 MPa	70.5 ± 0.13	71.5 ± 0.1	73.2 ± 0.11	73.9 ± 0.12

Note: ^aDry weight basis; and ^bmean value ± standard deviation of three replications.

Effects of SC-CO₂ methods on the yield of MSK fats

The total MSK fats of six varieties were extracted using continuous and soaking methods of SC-CO₂ extraction at pressures of 35 and 42.2 MPa, at temperatures of 60°C and 72°C, and at constant CO₂ flow rate of 3.4 ml/min. The SC-CO₂ extracted MSK fats were light white to yellowish in color, physically cleaner than those of Soxhlet extracted fats, and semi-solid at ambient temperature. The total fats were also extracted by Soxhlet method for comparison. There were no significant differences found in the yields obtained by both SC-CO₂ and Soxhlet extraction methods, although hexane is a most powerful solvent traditionally used for extracting fats from seed kernels (Table 1). Only minor variations were observed between the yields obtained by Soxhlet and SC-CO₂ extraction methods for all varieties. The reason for similar yields obtained using both SC-CO₂ and Soxhlet extractions in this study is that at temperature of 72°C and at pressure of 42.2 MPa, the density of SC-CO₂ could exhibit similar as hexane. Kuk and Hron (1994) reported that at extraction pressure of 48.2 MPa and temperature of 80°C, the density of SC-CO₂ was heavier (30%) than hexane at its normal boiling point. At this condition, authors extracted neutral cottonseed lipids as efficiently as hexane. The extraction yields of MSK fat increased with increasing pressures and temperatures for all conditions applied. The highest yield was obtained at pressure of 42.2 MPa and at temperature of 72°C for both continuous and soaking methods of SC-CO₂ extraction. The range of SC-CO₂ extracted yields varied from 64 to 135 g/kg for all conditions applied. The highest yield was 135 g/kg for soaking method at pressure of 42.2 MPa, at temperature of 72°C, and at CO₂ flow rate of 3.4, while it was 131 g/kg for continuous method with the same conditions applied. The yields obtained in this study for all methods (continuous and soaking) and conditions were comparable to each other. Sahena et al. (2010) extracted fish oil using various methods of SC-CO₂ extraction and showed that the soaking method gave a higher yield than the continuous method. However, in this study, the comparable results obtained by both continuous and soaking methods were presumably due to small particle size of MSK powder (particle size < 200 μm). Reverchon and Marrone (2001) reported that the particle size is one of the controlling parameters of SC-CO₂ extraction rate of oil yield. The authors also reported that due to the grinding breaks the intake oil cells of seeds or nuts and releases the oils present in these cells and as particle size decreases the amount of the released oil increases. The total fat contents of MSK varieties in the present study were comparable to that of Sal butter (140–160 g/kg) but significantly different from those of Kokum butter (440 g/kg), Shea butter (400–550 g/kg), Illipe butter (400–600 g/kg), and cocoa butter (600 g/kg) (Gunstone, 2011).

Fatty acid compositions

Tables 2 and 3 shows the fatty acid composition of MSK fats extracted by the soaking method of SC-CO₂ and Soxhlet extraction methods. Thirteen fatty acids, namely, myristic (C₁₄), pentadecanoic (C₁₅), palmitic (C₁₆), palmitoleic (C_16:1), heptadecanoic (C₁₇), stearic (C₁₈), oleic (C_18:1), linoleic (C_18:2), linolenic (C_18:3), arachidic (C₂₀), 11-eicosenoic (C_20:1), Behenic (C₂₂), and lignoceric (C₂₄) acids were identified in MSK fats. The fatty acids, such as C_16:1, C_20:1, and C₂₄, identified in the present study were the first report for MSK fats. The fatty acid compositions of studied MSK fats of all varieties varied considerably with the mango varieties. The C₁₄ content ranged from 0.02 to 0.09%, where Apple mango had the lowest and Sala mango had the highest (Table 2). For C₁₅ and C₁₇, Arumanis had the lowest and Chokonan had the highest, while the content ranged from 0.02% to 0.07% and from 0.1% to 0.2% in other studied varieties. The long chain fatty acids with odd number of carbon atoms such as C₁₅ and C₁₇ are highly remarkable for plant fats. However, Gaydou and Bouchet (1984), and Muchiri et al. (2012) have also reported these fatty acids in different mango varieties and their results were very close to the present study. For both C₁₆ and C_16:1 contents, Chokonan had the highest and Apple had the lowest, while their contents ranged from 6.9% to 10.9% (C₁₆) and from 0.06% to 0.1% (C_16:1). The C₁₈ content was found to be highest in Apple and lowest in Sala varieties, while their contents ranged from 32.8% to 47.6%. On the other hand, Sala had the highest C_18:1 content and Arumanis had the lowest and its contents ranged from 37.0% to 47.3%. The essential fatty acids such as C_18:2 and C_18:3 were found to be highest in Sala, while they were lowest in Apple mango variety. These two fatty acids are essential for human being and can not be synthesized by human body, and that they can only be supplied from the diet. For C₂₀, Arumanis had the highest and Sala had the lowest, while its contents ranged from 1.8% to 2.4%. Significant differences in the fatty acid constituents were not found in the fat yields extracted by both SC-CO₂ and Soxhlet methods (Tables 2 and 3).

Table 2. Fatty acid composition in fats extracted from various MSK using soaking technique of SC-CO₂ extraction at 44.2 MPa, 70.2°C, and 3.4 ml/min.

Table 2. Composición de ácidos grasos en grasas extraídas de varios NSM usando la técnica de extracción de CO₂-SC en remojo a 44,2 mpa, 70,2°C y 3,4 ml/min.

Fatty acid composition (%)^a	Arumanis mango kernel	Apple mango kernel	Chokonan mango kernel	Elephant mango kernel	Sala mango kernel	Waterlily mango kernel	Commercial CB^b
C14	0.04 ± 0.004	0.02 ± 0.002	0.07 ± 0.006	0.05 ± 0.004	0.09 ± 0.005	0.05 ± 0.001	0.7
C15	0.02 ± 0.001	0.04 ± 0.001	0.07 ± 0.001	0.03 ± 0.003	0.05 ± 0.002	0.03 ± 0.001	0.0
C16:0	7.98 ± 0.16	6.95 ± 0.08	10.93 ± 0.13	7.09 ± 0.1	8.74 ± 0.05	7.71 ± 0.09	24.1–25.8
C16:1	0.07 ± 0.004	0.06 ± 0.001	0.1 ± 0.006	0.06 ± 0.005	0.08 ± 0.004	0.06 ± 0.001	0.0
C17	0.21 ± 0.001	0.14 ± 0.001	0.17 ± 0.02	0.16 ± 0.004	0.14 ± 0.001	0.2 ± 0.002	0.0
C18:0	44.78 ± 0.09	47.62 ± 0.1	34.30 ± 0.08	46.91 ± 0.12	32.8 ± 0.03	42.27 ± 0.09	33.3–37.6
C18:1	37.01 ± 0.05	38.23 ± 0.01	43.87 ± 0.05	38.38 ± 0.01	47.28 ± 0.06	41.41 ± 0.02	32.7–36.5
C18:2	5.75 ± 0.02	3.66 ± 0.02	6.29 ± 0.004	4.06 ± 0.009	6.87 ± 0.04	5.09 ± 0.001	2.6–3.5
C18:3	0.65 ± 0.007	0.40 ± 0.001	0.96 ± 0.02	0.57 ± 0.002	1.15 ± 0.005	0.62 ± 0.005	0.1–0.2
C20	2.43 ± 0.03	1.89 ± 0.01	1.96 ± 0.02	1.89 ± 0.02	1.77 ± 0.06	1.78 ± 0.02	1–1.2
C20:1	0.15 ± 0.001	0.08 ± 0.001	0.19 ± 0.002	0.11 ± 0.001	0.16 ± 0.008	0.14 ± 0.002	0.0
C22	0.52 ± 0.01	0.35 ± 0.01	0.52 ± 0.04	0.39 ± 0.006	0.42 ± 0.01	0.38 ± 0.001	0.1–0.2
C24	0.39 ± 0.001	0.34 ± 0.02	0.47 ± 0.04	0.3 ± 0.001	0.45 ± 0.03	0.26 ± 0.004	0.0

Note: ^aMean value ± standard deviation of three replications, and ^bLipp and Anklam (1998).

Table 3. Fatty acid composition in fats extracted from varieties MSK using Soxhlet method.

Table 3. Composición de ácidos grasos en grasas extraídas de variedades de NSM usando el método Soxhlet.

Fatty acid composition (%)^a	Arumanis mango kernel	Apple mango kernel	Chokonan mango kernel	Elephant mango kernel	Sala mango kernel	Waterlily mango kernel
C14	0.04 ± 0.003	0.03 ± 0.001	0.05 ± 0.001	0.04 ± 0.002	0.07 ± 0.002	0.05 ± 0.001
C15	0.03 ± 0.001	0.04 ± 0.001	0.07 ± 0.001	0.05 ± 0.001	0.04 ± 0.002	0.03 ± 0.001
C16:0	7.78 ± 0.003	7.05 ± 0.11	10.58 ± 0.31	6.96 ± 0.04	8.52 ± 0.03	7.80 ± 0.12
C16:1	0.05 ± 0.002	0.05 ± 0.001	0.09 ± 0.003	0.06 ± 0.003	0.09 ± 0.002	0.06 ± 0.001
C17	0.21 ± 0.005	0.13 ± 0.002	0.21 ± 0.003	0.16 ± 0.001	0.15 ± 0.002	0.19 ± 0.005
C18:0	44.35 ± 0.02	47.11 ± 0.2	33.95 ± 0.26	46.62 ± 0.31	33.39 ± 0.02	42.44 ± 0.09
C18:1	37.21 ± 0.01	38.53 ± 0.02	44.16 ± 0.05	38.50 ± 0.15	47.01 ± 0.03	41.46 ± 0.03
C18:2	5.91 ± 0.002	3.89 ± 0.01	6.73 ± 0.002	4.25 ± 0.13	6.75 ± 0.02	4.93 ± 0.001
C18:3	0.74 ± 0.001	0.50 ± 0.002	0.97 ± 0.003	0.54 ± 0.02	1.21 ± 0.003	0.58 ± 0.007
C20	2.47 ± 0.005	1.81 ± 0.02	1.94 ± 0.001	1.97 ± 0.03	1.76 ± 0.04	1.70 ± 0.004
C20:1	0.17 ± 0.001	0.09 ± 0.002	0.19 ± 0.01	0.11 ± 0.003	0.17 ± 0.007	0.13 ± 0.01
C22	0.53 ± 0.001	0.44 ± 0.01	0.49 ± 0.01	0.41 ± 0.004	0.40 ± 0.02	0.37 ± 0.001
C24	0.41 ± 0.003	0.33 ± 0.03	0.57 ± 0.03	0.33 ± 0.008	0.44 ± 0.02	0.26 ± 0.004

Note: ^aMean value ± standard deviation of three replications.

In the present study, the C₁₆, C_18:0, C_18:1, and C_18:2 were the major fatty acids found in all mango varieties. Among these major fatty acids of the MSK fats, the C_18:0 and C_18:1 were the dominant fatty acids. The amounts of major fatty acids of MSK fat determined in this study were in close agreement with that reported by Ali et al. (1985) and Abdalla et al. (2007), but the amounts varied to a small extent from those reported by Muchiri et al. (2012). Moreover, the concentrations of C_18:3 and C₂₀ in MSK fats were also good in line with that reported by Ali et al. (1985) and Lakshminarayana et al. (1983) for different mango varieties. However, the investigated fatty acid profiles of the promising MSK fats were also good in line with that reported by many researchers in the literature (Abdalla et al., 2007; Ali et al., 1985; Lakshminarayana et al., 1983; Solís-Fuentes & Durán-de-Bazúa, 2004). The fatty acid compositions of the studied MSK fats were comparable in the contents of C₁₆, C_18:0, and C_18:1 with those of sal, shea, kokum, and illipe butter, but varied with respect to other component fatty acids. Thus, the total amount of these three fatty acids in MSK fats constituted more than 96%, which is close to these three fatty acids of cocoa butter (about 95%).

Physicochemical properties of MSK fats

Table 4 shows the physicochemical properties such as iodine value, saponification value, acid value, and SMP of the MSK fats of six varieties studied.

Table 4. Physicochemical properties^a of MSK fats extracted using soaking technique of SC-CO₂ extraction.

Table 4. Propiedades^a físico-químicas de grasas NSM extraídas usando la técnica de extracción de CO₂-SC en remojo.

Mango varieties	Saponification value (mg KOH/g fat)	Iodine value (g iodine/100 g fat)	Acid value (%)	Slip melting point (°C)
Arumanis	191.3 ± 1.27	47.21 ± 1.21	4.58 ± 0.57	38.4 ± 0.09
Apple	189.9 ± 0.96	48.1 ± 0.8	5.12 ± 0.46	39.1 ± 0.1
Chokonan	194.6 ± 0.57	51.28 ± 2.01	3.83 ± 0.71	36.0 ± 0.08
Elephant	191.7 ± 0.82	48.91 ± 0.09	4.41 ± 0.29	37.8 ± 0.05
Sala	190.7 ± 1.03	52.69 ± 0.46	3.57 ± 0.41	35.8 ± 0.03
Waterlily	195.5 ± 1.16	42.9 ± 1.12	3.23 ± 0.62	36.2 ± 0.1
CB	193.58–196.71^b	34.74–37.94^b	0.42–3.11^b	35^c

Note: ^aMean values of three replications; ^bChaiseri and Dimick (1989); and ^cKheiri (1982).

Iodine value

Determination of iodine value is important in the fats and oils chemistry. It is used to determine the degree of unsaturation and hardness of fats or oils. The studied iodine values of MSK fat varied from 42.9 to 52.69. Sala mango kernel fat showed the highest iodine value (52.69), while Waterlily mango kernel showed the lowest (42.9). The iodine values of Sala (52.69), Chokonan (51.28), and Elephant (48.91) varieties were comparable with the iodine values of Brindaboni (51.2), Kent (51.08), and Kuipahari (50.3) varieties (Ali et al., 1985; Muchiri et al., 2012). Moreover, the Apple (48.1), Arumanis (47.21), and Waterlily (42.9) had the similar iodine values as reported by Solís-Fuentes and Durán-de-Bazúa (2004) for Manila (47.7), and Ali et al. (1985) for Lakhanvoge variety (46.8). The iodine values determined for six mango varieties were in good agreement with that of palm oil, sal butter, and shea butter. On the other hand, the determined iodine values of MSK fat are a little bit higher than that of cocoa butter, kokum butter, and illipe butter. Thus, the fractionation or hydrogenation methods could be used to reduce iodine value of MSK fat to that desired for cocoa butter.

Saponification value

Saponification value is the number that is represented by potassium hydroxide in milligrams required to saponify 1 g of fat. Saponification value is a measure of the average molecular weight of all the fatty acids present in fats or oils. It increases with decreasing molecular weight. From Table 4, it can be seen that the saponification values varied from 189.9 to 195.5. The fat extracted from Waterlily MSK showed the highest saponification value, while Apple MSK fat showed the lowest. The saponification value of Apple MSK fat was also lower to a small extent as compared to the results obtained from different mango varieties in the literature (Ali et al., 1985; Muchiri et al., 2012). The lower saponification value of Apple mango variety could be due to the higher content of stearic acid (C_18:0) than that contained in the other studied varieties. However, the saponification values investigated for Waterlily (195.5), Chokonan (194.6), and Sala (190.7) mango varieties were in close agreement with those of Kent (195.9), Aroenanis (195), Kanchamitha (194), Misrakanta (191), and Kagege (190.2) mango varieties (Ali et al., 1985; Muchiri et al., 2012; Van Pee, Boni, Foma, Holyaerts, & Hendrikx, 1980). Moreover, the saponification values of Elephant (191.7), Arumanis (191.3), and Apple (189.9) mango varieties were also comparable to Dodo (193), Kalabau (192), Manila (189), and Boribo (188.8) mango varieties (Ali et al., 1985; Hussain, Haque, Gafur, A1i, & A1i, 1983; Muchiri et al., 2012; Solís-Fuentes & Durán-de-Bazúa, 2004). Thus, the saponification values of studied mango varieties (189.9–195.5) were in good agreement with those of kokum butter, hemp seed oil, corn oil, flax seed oil, and apricot kernel oil and were not so significantly different from those of sesame seed oil, shea butter, and illipe butter. However, the range of saponification values obtained in this study was comparable to those of illipe butter, shea butter, kokum butter, and cocoa butter.

Acid value

The acid values of fat extracted from six MSK varieties varied from 3.23 to 5.12. The acid value was found to be higher in Apple MSK fat, while it was lower in Waterlily kernel fat among all varieties studied. The acid values of Apple (5.12), Arumanis (4.58), Elephant (4.41), and Chokonan (3.83) mango varieties were similar as compared to Kagege (5.46) and Boribo (4.94) mango varieties (Muchiri et al., 2012). Moreover, the acid values found for Sala (3.57) and Waterlily (3.23) were comparable to that of Baramasi (3.8) and Hamlett (3.5) varieties (Lakshminarayana et al., 1983). However, the studied acid values of six MSK fats were a little bit higher than that of cocoa butter, tallow butter, illipe butter, and palm oil while it were lower than that of shea butter. The acid value of palm kernel oil was reduced by SC-CO₂ fractionation without any physical or chemical refining (Zaidul, Norulaini, Omar, & Smith Jr., 2006). The authors also used these fractionating palm kernel oils as cocoa butter replacer’s blending components. Thus, the acidity of MSK fats could be reduced by SC-CO₂ fractionation or by physical and chemical refining to that desired for cocoa butter.

Slip melting point

The slip melting point of the MSK fats determined for all varieties ranged from 35.8°C to 39.1°C with Apple having the highest and Sala the lowest figures. The variations of SMP found in varieties MSK fats could be due to the different amount of unsaturated fatty acids present in studied MSK fats. The Chokonan, Sala, and Waterlily mango varieties showed the similar SMP to each other; whereas the SMP of Apple, Arumanis, and Elephant mango varieties showed a little higher than the other varieties (Table 4). The SMPs of the all MSK fats investigated in the present study were very close to the figures reported by Sonwai, Kaphueakngam, and Flood (2012), and Baliga and Shitole (1981). Thus, the SMPs of the MSK fats investigated were higher to a small extent to that of cocoa butter, sal, shea, illipe, and kokum butter, but still comparable to each other.

Conclusions

Total MSK fats were extracted using continuous and soaking methods of SC-CO₂ extraction and the yields were compared with the solvent extraction (Soxhlet). The triglyceridges in terms of fatty acid constitutes were analyzed and compared for both soaking method of SC-CO₂ extraction and Soxhlet extractions. Moreover, the physicochemical properties of SC-CO₂ extracted fats were also compared with different vegetable fats. Results of total fats (76-137 g/kg) obtained by Soxhlet extraction were closer to the total fats extracted using soaking method (69–135 g/kg) at 42.2 MPa pressure, 72°C temperature, and 3.4 ml/min CO₂ flow rate. The Apple MSK variety contained a high amount of fats which placed it to the best among all other studied varieties. The Waterlily mango variety, on the other hand, was found to be the best in terms of fatty acids and physicochemical properties, similar to that of cocoa butter. Furthermore, palmitic, stearic, oleic, and linoleic acids were the predominant fatty acids which accounted more than 96.6% of the total fatty acids present in the studied MSK fats. Extracts with high content of such fatty acids, in addition to their physicochemical properties, are ideal for use in formulations for cocoa butter replacer’s blends in food industry; in particular, chocolate industry, and also extracts should be improved specifically chemical residues free to apply in the food and nutraceuticals industry. Moreover, the physicochemical properties of the studied MSK fats suggest that these are comparable to that of shea, sal, illipe, kokum, and cocoa butter. Thus, the MSK fat obtained by soaking method of SC-CO₂ extraction could be regarded as premium grade cocoa butter analogy fats which need to be blended with other vegetable fats to that of desired cocoa butter.

Acknowledgement

The authors are glad to acknowledge the USM-Fellowship of Universiti Sains Malaysia for providing financial support to conduct the study.