Effects on Peroxide Value
PV is an important indicator that provides the initial evidence of primary oxidation in TOs. As shown in Table 2, the evolution of PV was monitored for seven consecutive weeks at 95% CL. The TOs that were not exposed to sunlight showed the lowest increase in PV as a function of time. The PV (meqO2 / kg) in the TOs increased slightly from 2.03\(\pm 0.01\) to 2.14\(\pm 0.02\) (CNO), 6.13\(\pm 0.01\) to 6.23\(\pm 0.04\) (PKO), and 3.09\(\pm 0.01\) to 3.22\(\pm 0.04\) (PO) for the oils. The increase in PV (meqO2 / kg) despite the protection from light can be attributed to the presence of dissolved oxygen in the TOs (Johnson and Decker, 2015). Dissolved oxygen can induce oxidation of unsaturated FFAs in the absence of light (Johnson and Decker, 2015). In addition, the results in Figure 1 show that the increase in PV was more pronounced in the sunlight-exposed TOs. The TOs exposed to light showed a significant increase of the PV from 2.03\(\pm 0.01\) to 4.8\(\pm 0.03\) (CNO), 6.13\(\pm 0.01\) to 9.33\(\pm\)0.03 (PKO), and 3.09\(\pm\)0.01 to 8.43\(\pm\)0.05 (PO). This shows that the effect of sunlight on PV is significantly different (p < 0.05) for the control and exposed TOs. The rapid increase in PV can be ascribed to the chemical changes that occur in the TOs when exposed to sunlight. Therefore, prolonged exposure of TOs to relatively high light intensities caused significant initiation of photooxidation in the oils. These findings are in agreement with the work of Rukmini and Raharjo (2010), who observed a similar pattern. They reported an increase in the PV of VCO upon exposure to fluorescent light at higher storage times. Such an increase in PV has also been studied by Almeida et al. (2019), who stored refined PO (RPO) in an enclosed dark area (20–25 °C) and at ambient temperature with exposure to natural light (26–32 °C). After 12 months of storage with exposure to light, the PV (meq O2/ kg) for the RPO increased drastically from 00.52 to 85.29. The RPO stored away from light increased slightly from 00.59 to 14.64 meq O2/kg.
However, different types of oils can either undergo the same or different changes during the initial stages of oxidation. This is evidenced by the variations in the levels of PV for each TOs shown in figure 2. The maximum PV was recorded for PO on the 7th day (Figure 2), followed by PKO and CNO. At all stages, the increase in the PV of PO was more rapid than in CNO and PKO, which can be attributed to the increase in PV with the increase the degree of unsaturation in the TOs. PO consists of higher levels of unsaturated FFAs (C18:1) compared to CNO (C18:1) and PKO (C18:1), owing to which, PO showed a rapid increase in PV when exposed to sunlight than the PKO and CNO. Therefore, the observed increase in PV agrees well with the degree of unsaturation in the TOs. Suryani et al. (2020) also reported that CNO should theoretically exhibit a low rate of oxidation due to its low unsaturated FFA content. Thus, unsaturated FFAs readily form peroxides (or hydroperoxides) compared to the saturated congeners, which causes the high PV. A higher PV reflects lower chemical stability of the oil and renders the oil less safe for food and non-food uses (Frankel and Huang, 1994). Therefore, the chemical stability of TOs needs to be enhanced by protecting the oils from direct exposure to sunlight. This can be achieved by the appropriate revision and improvement of the packaging and storage conditions of these oils. Attention to such handling of TOs will improve their acceptability.

Effects on Iodine Value

IV is a vital characteristic of oils and represents the degree of unsaturated fatty acids (double bonds) present in the oil (Kumaret al. , 2012). This amount represents the unsaturated fatty acids before and after the oxidation of the oil. Thus, IV indicates the oxidative stability of oil and is a useful parameter for studying the changes in the unsaturated fatty acid contents in TOs exposed to sunlight. Table 3 presents the decrease in IVs of TOs stored under different conditions at 95% CL. The TOs stored away from light exhibited the lowest reduction of IV per week, unlike the oils exposed to light. The IV (g of iodine/100 g of oil) in the unexposed TOs decreased from 6.10\(\pm\)0.01 to 6.03\(\pm\)0.02 for CNO, 18.25\(\pm\)0.01 to 18.12\(\pm\)0.02 for PKO, and 49.35\(\pm\)0.01 to 49.19\(\pm\)0.03 for PO. The exposed TOs showed a significant decrease in IV per week from 6.10\(\pm\)0.01 to 4.94\(\pm\)0.04 for CNO, 18.25\(\pm\)0.01 to 16.6\(\pm\)0.01 for PKO, and 49.35\(\pm 0.01\) to 44.12\(\pm\)0.05 for PO. This shows that the effect of sunlight on IV is significantly different (p < 0.05) for the control and exposed TOs. IV decreases in TOs because it is proportional to the amount of iodine required to saturate the FFAs present in the TOs. The different types of FFA in the TOs account for the varied IV values for each oil. Approximately 51% of unsaturated fatty acids are found in PO (oleic acid), 35% in PKO (myristic acid), and 15% in CNO (myristic acid) (Gunstone, 2011). In addition, Figure 3 provides a comparison between the decreasing rates of IVs of protected and unprotected TOs using a bar chart, which shows that the decrease in IV is more pronounced when the TOs are exposed to sunlight. Thus, when the TOs are exposed to sunlight, a higher number of unsaturated FFA (C=C) in TOs are likely decomposed into radicals, thereby reducing the degree of unsaturation (C=C) in TOs. Dawodu et al. (2015) reported a similar trend when PKO was exposed to different temperatures. The IV in the PKO, initially at 5.00, decreased to 7.00 when the temperature was increased to 300 °C. Fekarurhobo et al. (2009) also reported a decrease in IV from 20.0 to 18.5 and 50.0 to 40.1 in PKO and PO, respectively. In their research, PO and PKO were exposed to sunlight for a shorter time (25 days) than that in the present study (49 days). However, PKO and PO showed a decrease in IV after 25 days of sunlight exposure. They further noted that the reduction of IV could be ascribed to the release of unsaturated FFAs as the π-bonds broke down during photooxidation. In addition, the highest significant decrease of IV was found in PO, followed closely by PKO and CNO. As previously mentioned, the degree of unsaturation of oils accounts for the rate at which the IV decreases. PO with the highest number of unsaturated FFA are likely to have more rapid reduction of IV than PKO and CNO. PKO has a higher number of unsaturated FFA, and therefore, its IV decreases faster than that of CNO. In summary, IV is a useful diagnostic of TOs photooxidation and provides useful information on the oxidative stability of TOs. In addition, it is a straightforward and convenient method for evaluation of TOs in typical laboratory settings and can be used by TOs producers to monitor the changes in TOs exposed to sunlight.

Effects on Colour

Colour is an indirect measure of product quality or product condition. The colours of the TOs were measured using a Lovibond Tintometer Pfx 880 series. The Lovibond colour scale was used to express the colour in the red and yellow AOCS tintometer units. Table 4 depicts the colour changes in the TOs samples during the seven day test period. The initial colours of the TOs were 1.11R\(\pm 3.40\)Y, 2.07R\(\pm\)18.2Y, and 3.63R\(\pm\)9.40Y for CNO, PKO, and, PO, respectively, before exposure to sunlight. After seven days of maintaining the samples in the absence of sunlight, the colours of the TOs reduced to 1.04R\(\pm 3.33\)Y, 2.05R\(\pm\)9.36Y, and 3.57R\(\pm\)9.36Y for CNO, PKO, and PO, respectively. These results indicate an insignificant change in the colour of TOs stored in the absence of sunlight. However, the TOs exposed to light showed a notable decrease in colour to 0.44R\(\pm\)2.51Y, 1.13R\(\pm\)16.8Y, and 1.80R\(\pm\)9.08Y for CNO, PKO, and PO, respectively. This shows that the effect of sunlight on the colour content is significantly different (p < 0.05) for the control and exposed TOs. These results are supported by the findings of Tonfack et al. (2019) (Tonfack Djikeng et al. , 2019) who reported the decrease in colour from 1.00R\(\pm 6.00\) to 0.20R\(\pm\)0.80Y after 90 days of sunlight exposure. Almeida et al. (2019) and Taluri et al. (2019) also reported that high-temperature storage conditions could reduce the colour content of oils. They reported that the degradation of colour pigments in oil during photooxidation affects the colour of the oil. These changes are pronounced in oils with a higher degree of unsaturation during prolonged exposure to light. The colour changes of the exposed TOs can be attributed to the damage of the colour pigments present in the oil. Thus, prolonged exposure to sunlight can deteriorate the colour pigments in oils. Choe et al. (2014) reported that colour changes could also be affected by the type of packaging material used for the oils. Transparent PET bottles used in the packaging of oils allow the transmission of light through oils, thereby destroying the pigments in oils. Hence, the colour content is reduced, changing the appearance of the oil; this change could indicate a problem during storage or the exposure of the oil to adverse conditions. The bar chart (Figure 4) shows that the decrease in colour was faster in PO than in PKO and CNO. As previously discussed, the degree of unsaturation is closely related to the variation of colour in oils. The higher the degree of unsaturation, the more likely are the colour pigments degraded during photooxidation. Therefore, the colour of TOs is a significant indicator of oil stability.

FTIR Studies on Photooxidation

ATR-FTIR analysis was performed to study the photooxidation of the TOs. Computational simulation was used to support the findings of the ATR-FTIR. Figures 5–7 show the experimental and simulated FTIR spectra for the exposed and unexposed (control) TOs. The principal absorption bands relevant for studying photooxidation were located at 3505 cm-3, 3560 cm-3, and 3554 cm-3, corresponding to CNO, PO, and PKO, respectively. The absorption bands were ascribed to the presence of secondary oxidation products, such as hydroperoxides or alcohol-related compounds(Navarra et al. , 2011). The bands indicated the stretching vibration of the O‒H bonds during oxidation. The stretching vibration was observed to be weak in the unexposed TOs due to the low degree of oxidation. A weak vibration was also observed for the exposed CNO due to the high degree of saturation, which renders it more stable against oxidation. These results comply with the those from the simulated spectra demonstrated in Figure 7, producing a strong absorption band at 3640 cm-3 for oxidised fatty acids and a less intense band for the unoxidised fatty acids (Figure 6). A similar result was reported by Poiana et al. (2015) (Poiana et al. , 2015), who used FTIR spectroscopy to evaluate the oxidation of edible oils after heating and frying. The band located at 3008 cm-3 in the spectrum of PO was assigned to the symmetric stretching vibration of the C‒H bonds in the cis-double bonds (Araújo et al. , 2011). This absorption was absent in the spectra of CNO and PKO because of the low amounts of unsaturated double bonds present in the oils. The simulated FTIR spectra from figure 6 &7 also showed located band spectra for the absorption band at 3009‒3006 cm-3. The absorption band at 2850‒2920 cm-3 in the spectra of all the TOs can be assigned to C‒H bonds for the symmetric and asymmetric stretching vibrations of the aliphatic CH2 and CH3 groups (Karabacaket al. , 2012). Another significant absorption band relevant for studying photooxidation was located near 1746 cm-3 in the spectra of the unexposed TOs. This band was assigned to the ester carbonyl functional group, which was more dominant in the less oxidised oils, as seen in the unexposed TOs (Poiana et al. , 2015). However, the vibrational frequencies of the absorption band decreased for the exposed TOs. Absorption bands were detected in the spectra of the oils at 1721 and 3505 cm-3 for the exposed CNO, 1720 and 3560 cm-3 for the exposed PO, and 1721 and 3554 cm-3 for the exposed PKO, and the reduction in these frequencies is attributed to the transformation of secondary oxidation products, such as aldehydes and other carbonyl group-containing compounds of a portion of the esters, during the oxidation process (Poiana et al. , 2015). These transformations caused the lowering of the vibrational frequencies of the ester groups with increasing absorbance. The simulated FTIR spectra also indicated a decrease in the absorption bands of the oxidised fatty acids. The simulated unoxidised lauric, palmitic, and oleic fatty acids exhibited a strong band at 1749 cm-3, which decreased negligibly to 1741 cm-3 after oxidation. Another vital absorption band observed at 888–990 cm-3 was the bending vibration of the trans group of disubstituted olefins (Poiana et al. , 2015). This band provides related oxidation information about the bands at 3009 and 3550 cm-3. The absorption band is extremely weak, with a low intensity compared with the other bands. Additional absorption bands and shifts in wavenumbers and their comparison to the simulated FTIR are summarised in Table 5. The results of the FTIR spectra from this study were comparable to those from the reported FTIR studies on edible oils (Guillén and Goicoechea, 2007). The results also agreed well with the simulated FTIR spectra shown in Figure 6 & 7.