Effect of heating time on the physicochemical properties of selected vegetable oils

2.1 Oil samples and reagents

Hydrochloric acid (HCl), sodium hydroxide (NaOH), potassium hydroxide (KOH), ethanol (C₂H₅OH), chloroform (CHCl₃), iodine, sodium thiosulphate (Na₂S₂O₃.5H₂O), starch indicator were of analytical grade and purchased from Merch Darmstadt, Germany.

Vegetable oil (palm olein oil, soybean oil, groundnut oil and palm kernel oil) were purchased from a local supermarket in Benin City, Nigeria.

2.2 Frying process

Oil sample was first heated up to 185 ± 5 °C for different duration of 1 h, 3 h, 5 h, and 7 h, to give a total of four (4) samples for each oil and twenty samples altogether. Immediately after heating, the oil samples were used to fry akara (bean cake) as it is popularly referred to in Nigeria. Frying duration was for five (5) minutes.

2.3 Specific gravity measurement

The densities of oil samples before and after frying were measured using a pycnometer with a 50 mL capacity. The empty bottle was weighed (W₀) and filled with oil sample. Thereafter, the content was emptied and filled with water and reweighed as W_2. The specific gravity was determined as follows using Eq. (1)

2.4 Viscosity measurement

The viscosity of oil samples before and after frying was measured by an Ostwald Viscometer techniconomial constant 0.05Cs/c, ASTMAD 445 England. The flow time (in seconds) of oil samples was recorded with a stop watch. Viscosity was calculated using Eq. (2)

2.5 Saponification value measurement

Saponification value was determined with slight modification according to ASTMD464. Typically, 2.0 g of oil sample was weighed into a conical flask, 25 mL of 0.5 N ethanolic KOH was added into the flask containing the oil and mixed thoroughly. A condenser was fitted to the conical flask and the mixture heated between 60 and 70 °C for an hour. The hot mixture was titrated with 0.5 N HCl using 2 drops of 1% phenolphthalein indicator until a colourless solution was obtained. A blank was determined in the same manner. The SV was obtained with Eq. (3)

2.6 Peroxide value measurement

Peroxide value was determined according to the method described by Cox and Pearson (1962). 1 g of oil sample was placed in tube, followed by 1 g of powdered KI and 20 mL of solvent mixture (2:1 of glacia acetic acid and chloroform). Then the tube was transferred to boiling water and allowed to boil for 30 s. Thereafter, the content of the tube was transferred to a conical flask containing 20 mL of 5% KI solution. 35 mL of distilled water was added to the mixture in the conical flask. The mixture was titrated with 0.002N Na₂S₂O₃ until a pale yellow colour was obtained. 0.5 mL starch was added to the pale mixture and titrated till a colourless mixture was obtained. A blank was determined in the same manner. The peroxide value was calculated using Eq. (4).

2.7 Acid value measurement

The acid value was determined using the ISO 660:1996 method. 2 g of oil was mixed with 10 mL ethanol, followed by two (2) drops of phenolphthalein. The mixture was boiled for about five minutes and then titrated with 0.1N KOH until a pale pink colour appeared. The acid value was calculated using Eq. (5) and FFA was calculated as follows in Eq. (6)

2.8 Iodine value measurement

The iodine value (IV) was determined using the Hubl method. Typically, 0.5 mL of oil was mixed with 5 mL of chloroform and dissolved by gentle swirling. Hubl’s Iodine was added slowly from the burette until the colour of iodine matched with the control. Then the burette reading was noted, and the experiment was repeated. Iodine value was calculated using Eq. (7).

2.9 Statistical analysis

The obtained data were subjected to one way ANOVA using Microsoft Excel 2013. The effect of time, temperature on the physicochemical properties of oil was investigated including analysing the effect of variation of physicochemical properties among the oil samples defined at p < 0.05.

3.1 Specific gravity and viscosity

The specific gravity is an important property affecting the heat transfer by natural convection. The specific gravity of the oils differ due to their different structural composition. Although there were changes on heating, heating at 185 °C at different times (1–7 h) did not have any significant effect on the specific gravity of the oils as can be found in Table 3. In Table 2, the highest values of specific gravity is attributed to soy bean oil, ranging in values from 0.919–0.922 followed by palm kernel oil with value of 0.918–0.921, then palm olein oil with value of 0.910–0.912 while groundnut oil had the least specific gravity range of 0.910–0.911. The values obtained were closely related to the standard range of 0.898–0.907 approved by Standard Organization of Nigeria (SON, 2000). During frying, due to hydrolysis, oxidation and polymerization processes, the composition of the oil degrades, which affects the stability of its compounds (Gloria and Aguilera, 1998). In general, there was reduction of specific gravity as the number of frying times increased except for the specific gravity of palm kernel oil which increased as the frying time increased. The reduction in specific gravity might be due to breakdown of the oil’s structure following heating at high temperature and consequently reducing the density of the oil as frying time progresses. The higher specific gravity observed for soybean oil can be attributed to the higher content of linoleic acid it contains.

Viscosity is another parameter used to measure physical changes. It depends on the specific gravity, molecular weight, temperature and degree of unsaturation. In general, the viscosity of the oils both increased and decreased without a trend when heated at increasing frying times as shown in Table 2. However, changes in viscosity at room temperature (27 °C) follows the trend: palm olein oil > groundnut oil > palm kernel oil > soybean oil as presented in Table 1.

The viscosity trend obtained before heating (seen in Table 1 ) was the same after heating (seen in Table 2). That is, palm olein oil showed the highest viscosity in the range of 810.635–875.412 while soybean oil showed the least viscosity reported in the range of 668.580–748.430. There was no definite trend on viscosity as heating time increased. In addition, there was no significant effect on heating time and temperature on the viscosity of oils as presented in Table 3. The increase in viscosity is likely due to saturation of the triglyceride chain by the process of hydrogenation while the decrease may be due to unsaturation (Santos et al., 2014). Increase in the viscosity of oil may also be due to polymerization of the oil and formation of dimmer (Choe and Min 2007). A decrease on the other hand, as a result of increasing heating time can be attributed to increase in the kinetic energy which enhances the movement of molecules and enhances intermolecular forces.

3.2 Peroxide value

Peroxide is a useful indicator on the extent of oxidation the oil has undergone. In Table1 and Table 2, the range of peroxide values of palm olein and soybean oil indicate that they are below 10 meq/kg oil specified by NIS (1992) and SON (2000). Perhaps, the reason for the wide variation in peroxide values (of palm olein and soybean oil) in comparison with the standards can be attributed to their dissimilar production processes. On account of their very low peroxide values, the two oils will display the tendency to be slow to oxidation or polymerization unlike groundnut oil and palm kernel oil as evident from their peroxide values after heating as presented in Table 2. According to Decker et al. (2010), the addition of antioxidants to oils can lead to low peroxide values. Antioxidants are added so as to prevent rancidity and increase shelf-life. Aside the antioxidants, the low peroxide values also suggest that the oils (palm olein oil and soybean oil) contain low molecular weight fatty acids of low unsaturation as a result of hydrolytic rancidity. Of all the oils, only palm olein oil showed an increase in the peroxide value with increasing heating time while groundnut oil showed a decrease in peroxide value with increasing heating time. Soybean oil and palm kernel oil showed no regular pattern in the peroxide values obtained. For soybean oil, a decrease is observed from 1- 3h and an increase from 5-7 h. As for palm kernel oil, there is an increase from 1 - 3h and a decrease from 5 - 7h. The fluctuation in the peroxide values of soyabean oil and palm kernel oilcan be attributed to decomposition of the peroxides that are formed during primary oxidation to secondary oxidation (Guillén and Cabo, 2002). A decrease in peroxide value with increasing heating time has been observed by several researchers (White, 1991; Vieira and Regitano-D’arce, 1999; Zhang et al., 2007). The decrease can be attributed to the fact that the peroxides formed on heating are unstable compounds towards high temperatures and frying time, transforming them to carbonyl and aldehyde compounds (Serjouie et al., 2010). The results further reveal that groundnut oil and palm kernel oil have gone rancid due to their high peroxide values and there is need for further refining in order to meet specific purpose. The trend in peroxide value before heating and after heating can be reported as palm kernel oil > groundnut oil > palm olein oil > soybean oil.

3.3 Saponification value

Saponification value (S.V) is an index of the molecular weights of triglycerides in the oil. It is inversely proportional to the average molecular weight or chain length of the fatty acids (Muhammad et al., 2011). Therefore, the shorter the chain length, the higher the S.V. The S.V obtained for the oil samples in Table 2 showed 59.69–61.34 mg/KOH/g for palm olein oil, 60.96–62.65 mg/KOH/g for soybean oil, 64.52–70.28 mg/KOH/g for groundnut oil, and 67.04–71.81 mg/KOH/g for palm kernel oil. Thus, S.V obtained for palm kernel oil and groundnut oil were found higher than S.V obtained for palm olein oil and soybean oil. All measured saponification values were below the expected range of 195–205 mg/KOH/g for edible oil as specified by SON (2000) and NIS (1992). The low value of S.V. suggest that the mean molecular weight of triglycerides is low or that the number of ester bonds are low. Therefore, these oils need refining in order to meet the required standard and would be better fitted for industrial purpose. On subjection to statistical analysis, there was no significant effect of temperature and heating time on the saponification values of the oils as can be found in Table 3. However, there was significant effect of the physicochemical properties on the oil types, indicating large variations in their production.

3.4 Iodine value

Iodine value is a measure of the amount of unsaturation in fatty acids. The iodine values obtained for the oil samples are shown in Table 2. For palm olein oil, the iodine value (1.382–1.785), increased from 0 to 3 h and then decreased steadily up to 7 h, due to breakage and saturation of bonds. In soybean oil, the trend is less defined with iodine value of 0.691–1.611. However, in groundnut oil, the iodine value (69.541–76.307) decreased steadily. This may be due to decrease in unsaturation. In contrast, the iodine value (70.302–81.977) of palm kernel oil, increased steadily. This is likely attributed to increase in saturation. Despite the observed fluctuations in the iodine values of the oils, no significant effect was found on the iodine values from heating at different times (ranging 1–7 h) and temperatures (between 27 °C and 185 °C). The low iodine values observed for both palm olein oil and soybean oil at different frying times when heated at 185 °C is likely to have contributed to their greater oxidative stability than palm kernel oil and groundnut oil.

3.5 Free fatty acid

The oxidation and chemical changes in oils during heating are characterized by an increase in free fatty acid content and decrease in the total unsaturation of oils (Perkin, 1992). Oxidative stability of oil is influenced by the fatty acid which is the chief composition of oil. From Table 2, the lowest FFA of 1.122–4.448 was recorded in palm kernel oil while the highest recorded FFA of 7.854–10.261 was found in groundnut oil. Only palm olein oil and palm kernel oil were within the maximum permissible FFA of 3.5mg/KOH/g for edible oils even after 7 h of heating. Before heating, (seen in Table 1), none of the oils were within the maximum permissible FFA as specified by SON (2000). Irregular high values of free fatty acid content as observed, might be due to impurities and variation during production, refining processes and also to the moisture content. Furthermore, high levels of FFA is an indication of hydrolytic rancidity which indicate the need for further refining before it can be used for specific purpose. According to Table 3, there was no significant effect of heating time and temperature on the fatty acid values of the oils, despite seeming that all the oils showed decrease in FFA after 7 h of heating. The decrease in FFA values of oils can be attributed to the esterification reaction between the FFA and the glycerol in the oil structure leading to the formation of triglyceride, and thus a reduction in FFA. There is also the possibility that low molecular weight acids in the oil may be lost through volatilization, hence causing a reduction in FFA (Hau et al., 1986).