Results and discussion
Changes in temperature
Figure 2 shows the temperature levels measured by the sensors in the oil of the eight tanks for 18 months. During the storage period from February to April of the year, the average temperature of oil in aboveground tanks was lower than 20 °C, and the average temperature of oil in the summer from June to September was about 30 °C, and the highest temperature was close to 40 °C. While the average storage temperature of the oil in underground tanks at 12-20 ℃ throughout the year. Storage temperature is an important factor affecting the oxidative rancidity and overall quality of oils. Generally, for every 10 °C increase in oil temperature, the oxidation rate will double [15]. The increase of temperature will also affect the minor components in oil [22].
Changes in the fatty acid composition
The fatty acid composition of peanut oils was determined using GC and listed in Table 2. It can be observed that five major fatty acids were present in peanut oils, which were C18:1 (40 %), C18:2 (39 %), C16:0 (11 %), C18:0 (3 %) and C22:0 (3 %). In a global study about fatty acids composition in peanuts the global average was of 43% of oleic acid and 35% of linoleic acid and the difference with respect to this result is that Chinese peanuts are more unstable with respect to global average [23]. Changes were found in the fatty acid composition during of 18 months stored, particularly in the composition of MUFAs and PUFAs in the oil samples. The fatty acids composition dose not change exceeding 0.6 % of oil sample in tank A. The loss was due to intensive oxidation at natural temperature. However, only slight alteration in the composition of fatty acids (not exceeding 0.3 % with reference to initial oil sample) was found in samples of tanks B, C, D. The above results are in confirmation with the findings of Wroniak et al [1], who reported PUFAs and MUFAs were significant (p <0.05) reduced in stored oil samples, but only slight alteration in the composition of fatty acids (not exceeding 0.3 % with reference to control oil sample) was found in samples exposed to oxygen and kept at room temperature (RT) after 12 months of storage of cold-pressed rapeseed oil. The fatty acid composition of soybean oil under various conditions was presented in Table 2. It can be observed that five major fatty acids were present in soybean oils, which were C18:2 (52 %), C18:1 (24 %), C16:0 (11 %), C18:3 (8 %) and C18:0 (5 %). Changes were found in the composition of MUFAs and PUFAs in the oil samples. The composition of fatty acids (not exceeding 0.3% with reference to initial oil sample) was found in samples of tank E. The loss was due to intensive oxidation at high natural temperatures. However, only slight alteration in the composition of fatty acids dose not exceeding 0.1 % of oil samples in tank F, G, H.
Changes in tocopherol contents
The total content of tocopherols in control peanut oil sample was 467.3 mg/kg (Table 3). Results showed that tocopherols were significantly affected by the storage conditions. Based on the percentage of losses, the highest rate of tocopherols degradation was observed when oil was stored in tank A. Under such conditions, aboveground storage resulted in 20 % loss of α -T, 62 % loss of γ -T and 44 % loss ofδ -T. When oil samples were stored in tank C, α -T,γ -T, δ -T loss of 11 %, 54 % and 40 % respectively. This phenomenon was due to contact of the oil with the oxygen present in the headspace of the tank, which promoted oxidative reactions of the tocopherols and thus their loss in the oil. Interestingly, negligible differences in percentage loss of tocopherols were observed in peanut oils of tank B and D, 2 % and 3 % loss of α -T, 53 % and 51 % loss of γ -T, 41 % and 34 % loss of δ -T respectively. The least losse of tocopherols, and thus the highest tocopherols content was found in peanut oil samples, where the total tocopherols content left 347.8 mg/kg for peanut oil stored in tank D conditioned with air at underground average 18 ℃ after 18 months of storage, the percentage loss of the total tocopherols 25.6 % (tank D). When stored aboveground, the percentage loss of the total tocopherols increased to 38.9 % (tank A), 27 % (tank B), and 31.5 % (tank C).
The total content of tocopherols in control soybean oil sample was 1525.4 mg/kg (Table 3). The least losse of tocopherols, and thus the highest tocopherols content was found in soybean oil samples, where the total tocopherols content left 1218.9 mg/kg for soybean oil stored in tank F conditioned with air and TBHQ added after 18 months of storage, and the percentage loss of the total tocopherols was 20.1 %, TBHQ had the best antioxidant effect on soybean oil. While the percentage loss of the total tocopherols increased to 42.7 % (tank E), the loss of total tocopherols were due to intensive oxidation at high natural temperature. The above results are similar to those of Lee et al [24], who reported the degradation rates for tocopherols of 2.1*103 and 8.9*102 %/day at 25 and 60 ℃ in soybean oil. The percentage loss of the total tocopherols increased to 24.5 % (tank G). At the same time, the percentage loss of the total tocopherols only increased to 21.5 % (tank H).
Changes in phytosterol contents
The peanut oil contained a total of 263.31 mg/100g phytosterols (Table 4). The following phytosterols have been determined in the analysed fresh oil sample: β -sitosterol, ca. 63 %, and campesterol, ca. 15 %, followed by sitosterol, ca. 12 %, while the content of stigmasterol was approximately 10 % of total phytosterols. As can be seen from Table 4, the highest degradation rate of total sterols was found in tank A, when the tanks A was conditioned with air, the percentage loss of total sterols to 23.9 %, while for the oil stored in tank C conditioned with nitrogen-filled, the percentage loss of total sterols to 22.1 %. Respective loss of total sterols to 21.3 % (tank B), When oil were stored in tank D, the percentage loss of total sterols to 16.9 %. In contrary to tocopherols, sterols showed high storage stability over the entire storage period [25]. This may be explained by the fact, that tocopherols were consumed first when protecting PUFAs from oxidation. The soybean oil contained a total of 358.01 mg/100g sterols (Table 4). The following phytosterols have been determined in the analysed fresh oil sample: β -sitosterol, 54 %, and campesterol, 20 %, followed by stigmasterol, 18 %, while the content of sitosterol was approximately 7 % of total phytosterols. As can be seen from Table 4, the highest degradation rate of total sterols was found in tank E, when the tank E were conditioned with air, the percentage loss of total sterols to 26.1 %, while for the oil stored in tank G conditioned with nitrogen, the percentage loss of total sterols to 24.1 %. Respective loss of total sterols in oil conditioned with adding TBHQ to 23.1 % (tank F), When oil was stored in tank H conditioned with air under ground, the percentage loss of total sterols to 19.2 %.
Changes in the volatile compounds
Table 5 listed the identified volatile compounds of the peanut oil stored in the four tanks. A total of 64 volatile compounds were identified, which included 20 pyrazines, 17 aldehydes, 9 ketones, 8 alcohols, 3 acids, 2 furans, 2 pyrroles, 1 lactone, 1 phenol and 1 benzoquinone. During the 18 months experimental period, pyrazines showed a significant decrease with storage time. Pyrazines 36.42%, as characteristic flavors of peanut oil, exhibited flavor characteristics of nutty and roasted aromas [26,27]. The loss of pyrazines in tank A peanut oil was the highest (content decreased to 16.33%), and the content of pyrazines in tank B, C and D peanut oils were reduced to 21.92%, 27.02% and 29.08%, respectively. Aldehydes showed a significant increase with storage time, and the content of aldehydes in tank A, B, C and D peanut oils were increased to 48.43%, 26.08%, 41.34% and 37.20%, respectively. In particular, the content of hexanal as a characteristic of the oxidation degree of oil was significantly increased [28], the content of hexanal in the four peanut oils was 6.40, 2.08, 4.37 and 2.10 times of the initial values, respectively. The peanut oil, in an accelerated oxidation test at 60°C showed that hexanal can reach values of 160µg/Kg after 28 days affecting the sensory properties [29]. This showed that adding TBHQ storage, nitrogen-filled storage and underground storage played a decisive role in the preservation of peanut oil. The initial content of ketones, alcohols, acids and lipids of other small molecules are very small, and the content may be slightly increased or changed little during storage. It is speculated that these compounds have little effect on the flavor change during storage of peanut oil. It is worth noting that the tert-butyl p-benzoquinone, a volatile compounds formed by the oxidative decomposition of TBHQ [5], was detected in the peanut oil added TBHQ, and the content increased with the storage time. After storage for 18 months, its content accounted 22.55% for the total volatile compounds. Tert-butyl p-benzoquinone has certain toxicity [6,7], affecting the safety of peanut oil. Therefore, although adding TBHQ storage can protect the flavor of peanut oil, its harmful compounds counteract its positive effects. In contrast, nitrogen storage and underground storage are greener and safer.
When GC-O method was used in evaluating the odour of the volatile compounds. 2,5-dimethylpyrazine and 2-methylpyrazine respectively showed nutty taste, roasted peanut flavor and nut aroma; benzaldehyde, furfural, phenylacetaldehyde and hexanal respectively showed bitter almond flavor, rotten grass flavor, sweet aroma and sour; 1-inch-3-ol, n-hexanol, n-pentanol and n-pentyl furan respectively give a mushroom flavor, a pungent taste, a grassy taste and a green bean flavor. The flavor wheel produced based on the sensory description of the GC-O sniffing of the main flavor compounds in peanut oil and the sensory evaluation score is shown in Fig. 3.
As the storage time prolonged, the overall flavor intensity of peanut oil gradually decreased. After 6 months of storage, the overall flavor, nutty flavor, baking flavor, sweet aroma and other strengths of the four peanut oils were significantly reduced, and the rancid taste was enhanced. In particular, the tank A peanut oil had the lowest overall flavor score and the highest rancid taste score; peanut oil has the highest overall flavor, nutty flavor, baking flavor and sweet aroma score, and has almost no rancid taste. The overall flavor score of tank B and C peanut oil is the same as that of tank D peanut oil. The nutty, baking and sweet flavor scores are slightly smaller than tank D peanut oil. This corresponds to the tendency of the volatile compounds of pyrazine in Table 5. It shows that nitrogen-filled tank, added TBHQ tank and underground tank can effectively delay the loss of peanut oil flavor compounds and achieve the purpose of quality preservation. After the storage time was extended to 18 months, the overall flavor, nutty flavor, baking flavor, and sweet flavor intensity of the four peanut oils decreased again, and the rancid taste was significantly enhanced. At this time, the characteristic flavor of the peanut oil was seriously lost, and the degree of rancidity was accelerated.
Changes in acid value (AV)
Figure 4a shows the trend in acidity levels for the peanut oil stored in the four tanks. The acid value of the four peanut oils increased slowly from February to April. After April, the acid values of the four peanut oils continued to increase with the increase of temperature. The acid value of peanut oil in tank A increased faster than others, and after storage in April of the following year, the acid values of the four peanut oils increased rapidly again with the temperature rising. The oil stored in tanks B and D showed a consistent trend, and the difference in acid value. This increase in acid value during the 18 months experimental period is lower than the increase for the oil in tank A and C. Statistical analysis indicated no significant differences (p <0.05)in acidity for tanks B and D, However, tank A and C was significantly different from tanks B and D starting from the second month of storage. At the end of 18 months of storage, the acid value of four peanut oils stored in tank A, B, C and D increased from the initial 0.84 to 0.99, 0.89, and 0.97 mg/g, respectively, according to GB/T 1534-2017 [30] of China, limited of peanut oil acid value (KOH) ever, tank A and C wvalue of the three peanut oils did not exceed the national standard limit.
Figure 4b shows the trend in acidity levels for the soybean oil stored in the four tanks. The acid value of the four soybean oils increased with the storage time. The acid value of soybean oil stored in tank E increased more obviously. The oil stored in tanks F, G, and H showed a consistent trend, this increase in acid value during the 18 months experimental period is lower than the increase for the oil in tank E. At the end of 18 months of storage, the acid value of soybean oil stored in tank E, F, G and H increased from the initial 1.01 mg/g to 1.47, 1.36, 1.38, 1.34 mg/g, according to GB 2716-2018 [31] of China, the limit value of soybean oil acid value ≤3 mg/g. The acid value of the four soybean oils did not exceed the national standard limit.
Changes in peroxide value (PV)
Figure 5a showed the tendency of the PV for peanut oil stored in the four tanks. The PV increased as storage time increased, especially after April, the peroxide value increased sharply with the increase of oil temperature, from September to April, the oil temperature is lower, and the peroxide value is also relatively flat. After the next year April, the peroxide value increases sharply after the temperature rises. For the oil stored in tank A increased the most, an elevated increase of peroxide value was observed, with a starting value of 2.5 and a final value of 53.9. The oil in tanks B, C, and D showed a significantly lower peroxide after 18 months of storage, than the oil stored in tank A. When the tanks A were conditioned with air, the PV was higher than 12 meq O2/kg (limit value of the Chinese Standard) only 5 months, while the tank D was underground, the PV stayed within the limit value of the Standard for 12 months. Other study showed that PV can reach until 90 meq O2/kg in accelerated oxidation test during 28 days at 60°C, but this value is for over the value limits for commercialization in the world, this PV affect the sensory performance with reject by consumers [29]. A significant difference (p <0.05) was observed. While the average storage temperature of the tank D is only 18 ℃. This important result shows that it is possible to improve the stability of peanut oil by replacing outdoor with underground storage. When the tank C was conditioned with nitrogen, the PV stayed within the limit value of the standard for 10 months. This confirms that the presence of nitrogen in the tank protects peanut oil from the oxidation process, even at outdoor temperature. When the tank B was conditioned with adding TBHQ, the peroxide value of peanut oil increased the least. The PV was lower than the limit value of the standard for 18 months of outdoor storage. It can be seen that TBHQ has the best antioxidant effect, the antioxidant effect of underground storage on peanut oil was second only to that of TBHQ. A strong correlation has been found between peroxides formation and the temperature of oil, leading to conclusion that the most critical factor affecting the rate of oxidation in the peanut oil is the high temperature.
Figure 5b showed the tendency of the PV for soybean oil stored in the four tanks. The PV increased as storage time increased. The tank E was significantly different from tanks F, G, and H starting from the third month of storage. For the oil stored in tank E, when the tank E were conditioned with air, an elevated increase of PV was observed, with a starting value of 2.9 and a final value of 65.6, the PV was higher than 20 meq O2/kg (limit value of Chinese Standard) only 7 months. While the oil stored in tank G was significantly different from tank E, a low increase of PV was observed, with a starting value of 2.9 and a final value of 21.6, when the tank G was conditioned with nitrogen, the PV stayed within the limit value of the Standard for 18 months. This may be linked to the fact that the use of nitrogen atmosphere and the reduction of the oxygen in the headspace volume can appreciably control oxidation of soybean oil during storage time. Similar findings were stated by, who reported the oil stored under air showed a marked decrease of quality after only two months of storage, while the use of inert gases in the headspace of the container during storage can reduce the presence of oxygen and preserve the compositional, nutritional and organoleptic qualities of the oil. The trends were similar for tank F and H, and the increase in PV was low, with a mean initial value of 2.9 and mean final value of 10.9, the PV was lower than the limit value of the standard for 18 months. No significant differences (p <0.05) were observed among tank F and H for the duration of the experiment. While the average storage temperature of the tank H was only 18 ℃. Other authors reported another indicator to make an easier comparison between different studies. The totox value combines PV and p -Anisidine value because they are complementary states of oxidation and the report is more comparable, but this indicator is not used in commercial scale [32].
Chemometric analysis
Hierarchical cluster ananlysis (HCA) was performed in order to observe similarities or dissimilarities or between the oil samples. The dissimilarities of different clusters was defined by the euclidean distance matrix and calculated by ward’s method. The results of peanut oil were presented in a dendrogram structure (Fig. 6), showing three distinct clusters based on the threshoid value (811.76) with a high similarity between tank B and tank D (Dissimilarity coefficient D=264.97). Whereas tank A and tank C presented two distinct clusters with a high dissimilarity (D=4326.86) between them and the other group of tanks.
The result of soybean oil was presented in a dendrogram structure (Fig. 7), showing three distinct clusters based on the threshoid value (3301.37) with a high similarity between tank F and tank H (Dissimilarity coefficient D=689.20). Whereas tank E and tank G presented two distinct clusters with a high dissimilarity (D=102456.51) between them and the other group of tanks.