2.2. Phenolic Compounds
In this study, a comprehensive quantitative analysis of twenty phenolic compounds was carried out in the investigated cold-pressed oils. The identified compounds covered a range of chemical classes, including flavonoids (apigenin, quercetin, luteolin, kaempferol), flavanols (catechin), glycosides (rutoside, vitexin), hydroxybenzoic and hydroxycinnamic acids (such as ferulic, caffeic, sinapic, and chlorogenic acids), and aromatic aldehydes (vanillic acid, vanillin). Individual compound concentrations, including standard deviations, are presented in
Table 3.
Among all quantified phenolics, sinapic acid was the most dominant, reaching exceptionally high concentrations in BPSO (1838.5 mg/kg), BSO (1678.8 mg/kg), and MTSO (1630.0 mg/kg). The lowest content was recorded in SO (143.9 mg/kg), representing a more than twelvefold difference between the most and least concentrated samples. These results confirm sinapic acid’s central role in the phenolic composition of cold-pressed oils and support previous findings on its antioxidant potential and function in lipid peroxidation protection [
31,
46,
47]. Chlorogenic acid levels were also highly variable, ranging from 58.3 mg/kg in BSO to 402.6 mg/kg in FSO. Notably high values were observed in EPSO (332.4 mg/kg), SBO (297.4 mg/kg), and BCSO (279.2 mg/kg). Oils rich in chlorogenic acid generally also contained high amounts of caffeic acid, with PSO (197.9 mg/kg), EPSO (194.8 mg/kg), and BCSO (189.4 mg/kg) showing the highest levels. The co-occurrence of these biosynthetically related phenolics suggests favorable extraction conditions and reflects their combined antioxidant activity [
48]. Ferulic acid reached its peak in CSSO (292.8 mg/kg), followed by PSO, BCSO, and EPSO, where levels also exceeded 100 mg/kg. This compound is of particular technological interest due to its potential synergistic role with tocopherols in improving oxidative stability [
49].
Among the hydroxybenzoic acids, 4-hydroxybenzoic acid was most abundant in CSSO (196.2 mg/kg), MTSO (168.7 mg/kg), and BPSO (136.0 mg/kg). Other acids in this group, including gallic and protocatechuic acids, exhibited more consistent levels across oils, typically ranging between 16 and 68 mg/kg, supporting their chemical persistence in lipid matrices [
47]. Within the flavonoid class, quercetin was the most prevalent, particularly in MTSO (209.9 mg/kg) and BSO (167.4 mg/kg), while SO and EPSO exhibited the lowest concentrations (24.5 and 65.6 mg/kg, respectively). Apigenin exceeded 100 mg/kg in EPSO, SFO, and BSO, but dropped below 40 mg/kg in MSO and SO. Luteolin showed a more even distribution, with the highest levels in BSO (66.1 mg/kg), EPSO (62.9 mg/kg), and SO (58.6 mg/kg). Kaempferol was most abundant in MSO (49.8 mg/kg) and showed notably lower levels in MTSO and BSO. Naringenin was primarily detected in BCSO (51.2 mg/kg) and PSO (48.9 mg/kg), while MSO and SO again showed the lowest concentrations. The only flavanol detected, catechin, showed the highest content in BCSO (75.0 mg/kg), FSO (65.8 mg/kg), and BPSO (46.7 mg/kg), while remaining below 16 mg/kg in EPSO, PSO, and SO. Glycosides, such as rutoside and vitexin, occurred in moderate concentrations across all oils. Vitexin levels were particularly elevated in BCSO (51.1 mg/kg) and remained consistently above 30 mg/kg in most samples, while rutin levels generally ranged from 12 to 18 mg/kg. Though minor in quantity, these glycosides may still influence oil stability due to their potential conversion to active aglycones [
50]. Phenolic aldehydes, including vanillin and vanillic acid, were found in the lowest concentrations among all quantified phenolics. Vanillin appeared in trace amounts only, peaking in CSSO (4.3 mg/kg) and averaging below 1 mg/kg in most oils. In contrast, vanillic acid reached 33.2 mg/kg in FSO and was relatively stable across the samples, suggesting better chemical resilience. Despite their low abundance, these compounds may contribute to the sensory and oxidative profiles of the oils [
48].
The total phenolic content (TPC), calculated as the sum of all quantified compounds, was highest in BPSO (2884.9 mg/kg), MTSO (2860.2 mg/kg), and BSO (2761.0 mg/kg), reflecting their richness in bioactive and antioxidant agents. On the other hand, the lowest TPC values were recorded in SO (937.5 mg/kg), MSO (941.6 mg/kg), and SFO (1002.5 mg/kg). This wide variation supports previous findings that phenolic content is strongly influenced by seed variety, oil matrix composition, and extraction efficiency [
27].
2.3. Tocopherols
Among the studied oils, SBO and FSO contained the highest total tocopherols contents, at 109.88 and 84.16 mg/100 g, respectively (
Table 4), with γ-tocopherol (γ-T) clearly dominating in these oils, at 75.02 and 79.93 mg/100 g in SBO and FSO, respectively. These γ-T levels were similar to, or slightly higher than, those reported in previous studies [
31,
38,
51]. However, significantly lower values of γ-T in FSO, with a mean of 12.4 mg/100 g, have also been reported [
52]. At the same time, FSO characterized only a few percentages of α-tocopherol (α-T) (4.23 mg/100 g), which is consistent with data reported by other authors for this oil [
31]. SFO and SO showed comparable total tocopherol contents, 65.93 and 64.85 mg/100 g, respectively, with very similar compositions of α-T and γ-T (27.48 and 26.07 mg/100 g for α-T and 36.74 and 38.36 mg/100 g for γ-T, respectively).
There were no significant differences between BPSO, CSSO, and MTSO in total tocopherols contents; however, notable variations were observed in the amounts of individual tocopherols. BPSO was characterized by nearly equal levels of α-T and γ-T, whereas MTSO contained 1.7 times more γ-T than α-T, which is opposite to some previous reports [
31]. In CSSO, the highest level of α-T and simultaneously the lowest level of γ-T were noted, 44.87 and 10.37 mg/100 g, respectively, which is opposite to [
31], where α-T was not detected in CSSO. PSO exhibited a comparable amount of γ-T and a slightly lower level of α-T (11.60 and 35.03 mg/100 g, respectively). At the same time, it was the only oil that contained a significant amount of δ-tocopherol (δ-T), at 13.63 mg/100 g. However, some of the literature sources show a significantly higher γ-T content in PSO along with similar levels of α-T and δ-T [
31].
BSO, MSO, and BCSO had similar α-T contents, ranging from 16.79 to 19.23 mg/100 g. Among these oils, in terms of γ-T, BSO (36.61 mg/100 g) showed the highest content, followed by BCSO (25.63 mg/100 g) and MSO (17.85 mg/100 g). This means that MSO was also characterized by the lowest total tocopherol content (37.88 mg/100 g). BCSO was the only oil in which a few percentages of β-tocopherol (β-T) were determined (3.34 mg/100 g), although a higher β-T level in BSCO was found in [
53] and a lower one in [
24]. EPSO, after FSO, contained the lowest α-T amount (10.63 mg/100 g) and had a γ-T content (32.77 mg/100 g) comparable to MTSO, BSO, SO, and SFO.
The differences observed in tocopherol composition and concentration between the present results and those in the literature can be explained by genetic and agronomic factors, quality characteristics of seeds, and technological factors [
54].
The tested oils were characterized by high levels of total tocopherols, ranging from 37.87 in MSO to 109.88 mg/100 g in SBO, where α-T and/or γ-T were the dominant forms, concerning previous reports [
22]. β-T and δ-tocopherols (δ-T) were either not detected or found at low levels (below 3.34 mg/100 g) in the analyzed oils with the exception of PSO, which contained 13.63 mg/100 g of δ-T, the secondary tocopherol after α-T in this oil.
Although α-T is appreciated by consumers as a beneficial component of oil, γ-T, when present at elevated levels, contributes significantly to the oil’s antioxidant capacity [
55]. Among the analyzed oils, the highest levels of γ-T were detected in FSO and SBO, which may influence their oxidative stability. It is also known that antioxidant properties depend, among other factors, on the concentration of bioactive compounds. Studies have demonstrated that α-T exhibits its highest antioxidant activity in vegetable oils at a concentration of 100 mg/kg, whereas concentrations above 250 mg/kg may lead to a slight pro-oxidant effect. On the contrary, γ-T shows the most potent antioxidant properties at concentrations between 250 and 500 mg/kg [
31]. Furthermore, during prolonged storage of oils, a significantly greater reduction in α-T concentration has been observed compared with γ-T, which strongly correlates with the formation of lipid hydroperoxides [
31]. In contrast, δ-T has shown stronger antiradical properties than the γ-homologs in some studies [
56]. Although tocopherols are known as the main lipid-phase antioxidants, in some cases, other bioactive compounds can also play a significant role in antioxidant activity. For instance, in pumpkin seed oil, 59% of the antioxidant capacity was attributed to polar phenolics, while only 41% was due to tocopherols [
55].
2.4. Pigments
Pigments such as chlorophylls and carotenoids in cold-pressed oils originate from the plant seeds and are not removed during the cold pressing. Their content is influenced by many factors, including plant cultivation, seed maturity, and the type of pressing method used [
27,
51]. As reported, the chlorophyll level in edible oils should not exceed 50 mg/kg due to its pro-oxidative effects, which negatively impact the oxidative stability of the oil and accelerate rancidity, thereby shortening its shelf life [
45]. However, in the absence of light, chlorophylls can act as antioxidants [
9,
31]. In contrast, carotenoid pigments function as antioxidants by protecting triacylglycerols, unsaturated lipids, membranes, and phenolic quinones from photo-oxidative processes. They are also known for their cancer-preventive properties [
51].
Chlorophyll and carotenoid pigments were determined spectrophotometrically and expressed as chlorophyll
a and β-carotene in mg per 100 g of oil (
Table 4). Chlorophyll content ranged from 0.02 mg/100 g (in SO and SFO) to 0.89 mg/100 g (in BCSO), which is below the acceptable level in edible oils. The exception was PSO (21.14 mg/100 g), mainly due to the presence of a high content of pheophytins, the main dye in this oil [
31]. Similarly low chlorophyll levels have previously been published for camelina oils, varying from 1.02 to 2.18 mg/kg [
30]. In contrary, higher chlorophyll levels were noted in [
52]. Although most of the analyzed oils have chlorophyll levels below the tolerable limit (< 0.9 mg/100 g), this does not exclude the possibility of photochemical oxidation being induced at such concentrations [
57].
Concerning carotenoid pigments, among all the analyzed oils, PSO (8.96 mg/100 g) contained the second-highest level of carotenoids after SBO (229.22 mg/100 g), which showed the highest one. The carotenoid levels in the remaining oils ranged from 0.10 to 1.35 mg/100 g, which were lower in CSSO, EPSO, MTSO, and BCSO than the values reported in the literature [
45]. However, some studies have shown similar or even lower carotenoids levels in these oils [
9,
22,
27,
51]. As previous results on the effect of β-carotene on the light stability of soybean oil have shown, the presence of this pigment at concentrations between 5 and 20 mg/kg of oil provides a protective effect against oxidative damage induced by light [
58]. Regarding these results, only the carotenoid content of FSO, BCSO, and MSO could be considered sufficient to provide some protection against photo-oxidative degradation.
A source of discrepancy in the spectrophotometric measurement of pigments in unrefined oils may be the calibration of the method for only a specific type of pigment, most often chlorophyll a or pheophytin a, and β-carotene or lutein, for chlorophylls and carotenoid pigments, respectively. Although their absorption maxima occur at the same wavelength, the molar absorption coefficients differ significantly, additionally depending on the solvent used. This may explain the different results obtained by diverse researchers employing different methods.
Previous research has demonstrated that in unrefined oils, the ratio of carotenoids to chlorophylls is variable [
59]. In our study, most of the tested oils contained higher carotenoid levels than chlorophyll (ranging from 0.10 to 229.22 mg/100 g), except for BCSO, PSO, and BSO, where chlorophylls predominated. Earlier studies also confirmed that chlorophylls are the most abundant pigments in these oils [
45]. As shown in
Table 4, the richest source of carotenoids, excluding SBO, was PSO (8.96 mg/100 g), and the poorest one was SO (0.10 mg/100 g). A similar trend among unrefined oils was also observed previously [
59]. SBO was characterized by the highest carotenoid content, reaching a level similar to that noted in the literature (approx. 200 mg/100 g) [
22]. Such a high carotenoid content provides regenerative and antiwrinkle properties of this oil, with β-carotene identified as the main carotenoid responsible for sea buckthorn oil’s characteristic orange-red color [
60].
2.5. Specific Extinction Coefficients (K232 and K268)
Lipid oxidation represents the primary mechanism responsible for quality degradation in fat-containing foods. The oxidative stability of oils is primarily influenced by their fatty acid composition and the presence of antioxidants. The primary changes occurring in stored oils result from the presence of double bonds in fatty acids, with susceptibility to oxidation increasing rapidly as the number of double bonds per molecule rises. Consequently, PUFAs are considerably more susceptible to oxidative degradation than MUFAs or SFAs [
61]. The amounts of conjugated products formed from PUFAs, described as conjugated dienes, are proportional to the content of hydroxyl peroxide compounds [
30]. Secondary oxidation products, such as conjugated trienes, were also determined at the beginning of the heating process and within a time of storage of 21 days at 60 °C. The results were expressed as K
232 and K
268 extinction coefficients, respectively. The obtained results for SO, MSO, SBO, BPSO, BSO, and SFO oils are given in
Table S1, while data for the remaining oils are presented in our earlier study [
20].
Significant differences in conjugated products between the oils were noted at the start point of the heating test. The analyzed oils showed K
232 values ranging from 1.60 to 12.30, while K
268 values were much lower, ranging from 0.03 to 2.70. The highest concentrations of conjugated dienes were noted in SFO and SBO, at 10.55 and 12.30, respectively, indicating a significant accumulation of oxidation products. The highest K
268 value was detected for PSO (2.70), simultaneously with a significant value of conjugated trienes (7.50). Among the analyzed oils, CSSO exhibited the lowest K
232 (1.60) and K
268 (0.03) extinction values. Similar results were achieved for CSSO by Ratusz et al. [
30]. Although there are no standardized limit values for K
232 and K
268 for cold-pressed oils, some researchers have discussed their results referring to the EU Commission Regulation 2568/91 for extra virgin olive oil [
15,
30]. According to this regulation, K
232 and K
268 values should not exceed 2.50 and 0.22, respectively. Following these guidelines, only FSO and CSSO met the given criteria, while BSO was at the threshold.
The reason for the higher K values observed in other studied oils, indicating a significant content of oxidation products, may resulted from technological processes or oil storage conditions. However, it also may be attributed to the co-occurrence of some compounds absorbing in the same wavelength range, such as carotenoids present in pumpkin seeds [
62]. Similarly, values exceeding the aforementioned norms for several cold-pressed oils, including BSCO, PO, and EPSO, have been noted by other researchers [
15,
31].
During storage, the level of both primary and secondary oxidation products increased at diverse rates depending on the oil, highlighting variations in oxidative stability and susceptibility to degradation. In all oils, a significant increase in these parameters was observed after about 10–14 days of the test. After 18 days, in some oils (MSO, SBO, SFO, EPSO) the K
232 value began to decrease, which may be connected with the peroxide concentration reaching its maximum and next decomposing into secondary oxidation products [
5]. Hence, based on the obtained K values, to better visualize the oils’ susceptibility to oxidation, the growth rates of K
232 and K
268 were calculated after 10 days, 14 days, and at the endpoint—either after 18 days (for MSO, SBO, SFO, and EPSO) or after 21 days (for the remaining oils). These results are summarized in
Table 5.
At the initial stage of storage, the growth rates of the K parameters ranged from 1.04 to 6.80 for K
232 and from 0.82 to 24.40 for K
268, with BPSO and CSSO showing the fastest growth rates of K
232 and K
268, respectively. A significant growth rate of secondary oxidation products was also observed for EPSO (18.58). At the same time, CSSO, together with MSO, exhibited a significant higher rate of K
232 growth compared with the other oils, while the lowest increases in both primary and secondary oxidation products were noted for BCSO and PSO. The high oxidative stability of BCSO and PSO, in comparison with some other cold-pressed oils, has also been reported by other studies [
5,
15].
After 14 days, the growth rate of K
232 increased comparably for most oils, enlarged 1.3 (for CSSO and SO) to 2.14 (for BSO) times. Exception were BCSO and PSO, where no significant differences were observed, and MSO, where the K
232 growth rate slightly decreased. By the end of the test, the fastest growth of oxidation product occurred in FSO (growth rate 33.17), achieving an 8- and 4.25-fold increase compared with 10 and 14 days, respectively. Similarly low oxidative stability during storage for FSO has also been observed by other researchers [
31]. In the case of FSO, the oxidation rate was particularly high despite its significant levels of γ-T (79.93 mg/100 g) and polyphenols (2293 mg/kg). This suggests a possible threshold beyond which antioxidant compounds are insufficient to prevent lipid peroxidation, or which counteracts the effects of the presence of trace pro-oxidant agents such as metal ions. However, further studies are warranted to confirm this hypothesis.
More significant variances occurred in the K
268 growth rate, which increased most notably after 14 days in BPSO (by 3.8 times). The largest percentage change in conjugated products during storage of poppy seed oil was also reported by [
23]. By the end of the storage period, the K
268 growth rate increased most significantly for EPSO and CSSO, rising rapidly by about 4–8 times compared with the initial stage of storage, indicating a sharp increase in the oils’ degradation.
Significant differences were observed between the oils at both the first and subsequent stages of oxidation. The lowest degradation rates were observed for PSO and BCSO throughout the entire storage period. A slow increase in K values was also noted for SO, but only in the initial storage period. The fastest oxidation process was recorded in FSO, confirming previous reports [
15,
21,
22,
38].
2.6. Correlations
To investigate the influence of the analyzed FAME parameters and phytochemicals (fatty acid composition and the content of chlorophylls, carotenoids, phenolics, and tocopherols) on the oxidative stability of the oils, a statistical analysis was performed. The results of the correlation analysis between selected quality variables and oxidative stability, determined based on the presence of conjugated dienes and trienes (K
232, K
268, and their growth rates) as well as theoretical Cox values, are presented in
Table 6.
In our previous study, it was found that the K
232 and K
268 values for MTSO, EPSO, FSO, CSSO, BTSO, and PSO mostly reflected the susceptibility of these oils to oxidation related to their PUFA contents [
20]. Similarly, when considering all oils analyzed in the present study, a comparable relationship was found between PUFA content and K values at the end of the thermostatic tests, with a significant (
p ≤ 0.05) positive correlation (r = 0.66 and r = 0.71 for K
323 and K
268, respectively). Moreover, a strong positive correlation (r = 0.91) between K
232 and K
268 confirmed a close connection between the decomposition of primary into secondary oxidation products [
5]. Additionally, the presented relationship results revealed a significant positive correlation between K
268 and C18:2 content (r = 0.59). However, when considering the growth rate of K values, no statistically significant correlations were obtained between them and either the fatty acid groups (PUFA, MUFA, or SFA) or individual fatty acid contents. The exception was C18:3, as one of the most susceptible fatty acids to oxidation, which had the greatest impact on the growth rate of primary oxidation products at the end of the test (r = 0.67). Symoniuk et al. [
15,
45] also noted no significant effect of the overall content of individual fatty acid groups on the oxidative stability of some cold-pressed oils determined by the Rancimat method (oil oxidation at 100 °C), simultaneously with a significant impact of C18:3 on the oxidative stability of the oils (r = −0.64 and r = −0.54). Similarly, Ratusz et al. [
30] found no significant correlations for camelina oils using both the Rancimat and Pressure Differential Scanning Calorimetry (PDSC) methods. On the other hand, Maszewska et al. [
18] indicated that oxidative changes in some cold-pressed oils were of a similar nature and comparable in both the Rancimat and thermostatic test, as confirmed by the high correlation between results obtained from each method. The correlation results in our study, compared with the above-mentioned outcomes, seem to confirm this relationship.
As indicated earlier, the relationship between an oil’s susceptibility to oxidation and its potential oxidizability is rarely noted, since oil is a complex mixture of compounds. The oxidative stability of oils is also influenced by the presence of antioxidants. Relationships between oxidative stability and antioxidant content have been found significantly more often [
15,
30,
45]. However, in our study, merely significant negative correlations (
p ≤ 0.05) were identified between
α-T and K
268 growth rates (r = (−0.72)–(−0.63)). Based on the data given in
Table 6, it can be concluded that the oxidative stability of the tested oils does not depend directly on any individual component. Instead, it has a complex nature influenced by many parameters. At first, vegetable oils are a rich source of various substances with antioxidant activity. Moreover, their composition and concentration can also impact their synergistic or antagonistic activity as well as lipophilic or hydrophilic properties [
5,
22,
31]. Hence, high correlations can be found between individual bioactive components of the oils [
22,
27,
31], although in our results, only carotenoids, which are considered to be the main lipid antioxidants, were highly positively correlated with total tocopherols content (r = 0.79). Synergistic interactions, including those between tocopherols and β-carotene, have been demonstrated in palm oil and in sea buckthorn oil blends [
22].
The theoretical Cox values and their relationship to the growth rates of K
232 after 10, 14, and 18 (or 21) days of the thermostatic test for the investigated oils, presented in
Figure 1, confirm the complexity of the observed processes. The lowest Cox index, obtained for SO (
Section 2.1.3), followed by PSO, indicated higher oxidative stability for these oils. However, BCSO exhibited the slowest K
232 increase. On the other hand, the highest Cox value was calculated for CSSO, although FSO, rich in bioactive compounds, exhibited the lowest oxidative stability, similar to BPSO, which had a moderate Cox value. Therefore, the Cox index was positively correlated with the growth rates of K
232 after 14 days (r = 0.59) and more significantly with the K
232 growth rate at the end of the test (r = 0.79). This finding is in contrast with other results where no correspondence among Cox and the oxidation induction time, as measured by the Rancimat method, was observed [
30,
45]. However, the literature also provides data showing a high correlation between the Cox index and the oxidative stability established with the PDSC method [
45]. Moreover, some studies showed that FAME could be a predictor for oil oxidation stability at an early stage of oil oxidation but not in later steps [
21]. This is opposite to our findings, where correlations improved over time. In our case, the best correlations were observed at the end of the test, where the fatty acids composition became the dominant factor rather than the bioactive pro- or antioxidants’ ingredients.
To further investigate the relationships between fatty acid composition, the content of bioactive compounds, and the oxidative stability of the analyzed oils, a principal component analysis (PCA) was performed. The results are presented both graphically (
Figure 2) and in tabular form (
Table 7). The PCA enabled dimensionality reduction and the identification of variables with the greatest influence on sample variability and susceptibility to oxidation.
The first two principal components (PC1 and PC2) explained 53.83% of the total variance (34,34% and 19,49%, respectively), indicating a strong explanatory value of the model. Although the first two principal components accounted for only 53.83% of the total variance, PCA was used primarily as an exploratory tool to visualize clustering patterns and general relationships among samples and quality indices. Additional components, while explaining more variance, did not significantly improve the interpretability of the data. As shown in
Table 7, the highest positive loadings for PC1 were observed for oxidation indicators (K
232, K
268), their growth over time (especially after 10 and 14 days), PUFA content, and the Cox value. The strong positive association among these variables confirms their collective contribution to oxidation processes and their mutual interdependence. Particularly high loadings were noted for 14-day K
232 (0.2978) and the Cox value (0.2772), emphasizing their importance in distinguishing the oxidative stability of oils. C18:3, known for its high oxidative susceptibility, also showed relatively high positive loadings on both PC1 and PC2, further supporting its central role in the initiation of primary oxidation reactions. This suggests that C18:3 content may serve as a meaningful predictor of early oxidative changes in oils. In contrast, antioxidant-related variables such as α- and γ-T, carotenoids, and chlorophylls had lower loading values and vectors oriented in the opposite direction to the oxidation indicators (
Figure 2), suggesting their potential protective effect, particularly against the formation of secondary oxidation products (K
268). Notably, α-T showed a strong inverse relationship with the growth rate of K
268, which is consistent with earlier correlation results (r = −0.72 to −0.63).
Furthermore, both the PCA plot and the tabular data highlight the interrelations among bioactive components. A strong positive correlation was observed between carotenoids and total tocopherol content, suggesting potential synergistic antioxidant effects. Overall, the PCA results confirm the complex nature of oil oxidative stability, determined by a combination of pro-oxidant factors (e.g., PUFAs) and antioxidant compounds, along with their interactions.