To determine fatty acid profiles, a gas chromatography/flame ionization detector (B420A, BEIFEN, China) was used. The film thickness, length and inner diameter of the BPX70 GC column were 0.25 mm, 60 m, and 0.25 mm, respectively. The stationary and carrier phases were Polymer Biscyanopropylsiloxane silphenylene and nitrogen gas, respectively (Keramat and Golmakani, 2016).

The PV was quantified according to the Official Method (Cd 8-53) of American Oil Chemists’ Society and was expressed as meq O₂ per kg oil (AOCS, 2000).

Reported as a percentage of oleic acid, the free fatty acid content was measured according to the AOCS official method (Cd 3d-63) (AOCS, 2000).

The density of CKO was measured following the AOCS official method (Cc10a- 25) whereby a pycnometer was used (AOCS, 2000).

Measuring the CKO was assisted by Nuclear Magnetic Resonance (NMR, Bruker, NMF 100, Karlsruhe, Germany). Prior to NMR analysis, the samples were melted at 100 °C and then cooled to 80 °C for 5 min. After that, the temperature was decreased to 60 °C and held for 5 min. Finally, the samples were rapidly cooled to 0 °C and t held at this temperature for 1 h. The samples were stabilized at 0, 5, 10, 20, 25, and 30 °C and were kept for 30 min at each temperature prior to NMR analysis (Nejadmansouri et al., 2016).

The refractive index of CKO was determined using a refractometer (RX7000a; Atago, Japan) at 20 °C.

The AOM analysis was carried out according to the AOCS official method (Cd 12-57). Briefly, 20 mL portions of CKOs were poured into the reaction tube and placed in a constant-temperature heater which kept the temperature constant at 97.8 ± 0.2 °C. The total flow rate was adjusted to 2.33 mL/min for each tube. The AOM induction period (IP) (i.e. the time required to reach a PV of 100 meq O₂/kg oil) was calculated according to the standard method of AOM (AOCS, 2000).

The protection factor (PF) was calculated according to eq. (1).

Where AOM IP_a is the AOM induction period (IP) of the CKO samples containing antioxidants (BHT, β-carotene, α-tocopherol, tocopheryl acetate and CEO) and AOM IP_c is the AOM IP of the control sample (Hraš et al., 2000).

The oxidative stability was estimated by measuring the Rancimat IP, using a rancimat 743 apparatus (Metrohm, Switzerland) according to the AOCS method (Cd 12b-92). The tests were carried out with 3.0 ± 0.1 g of CKO. All samples were studied at 80 °C. The temperature of the conductivity tube was kept constant at 21 °C and the air flow rates of motion were set at 20 L/h for each experiment. The Rancimat IP was printed automatically by the apparatus software with a precision of two decimals (AOCS, 2000).

The PVs of CKO samples were worthy of documentation during accelerated storage (Figure 1). The PVs of all CKO samples increased at the early stages of the storage period. Even though the PV of the control group began to decrease after 12 h, the PVs of the samples containing antioxidants began to decrease after 18 h. The PV of the control sample increased much faster and to a higher level than that of the CKO samples as they contained antioxidants and reached 107.93 meq O₂/kg after 12 h. In going beyond the mentioned time, and especially after 18 h, the PVs of samples which contained CEOs extracted by HD, SD, MAHD and MASD reached 89.94, 82.18, 80.20, and 88.45 meq O₂/kg, respectively (Figure 1a). This indicated that the CEOs can delay the rate of hydroperoxide formation. The CEO extracted by MAHD functioned more effectively in reducing the PV of CKO, compared to the function of CEOs extracted by the HD, SD and MASD methods (P < 0.05). This may be attributed to the presence of higher amounts of eugenol in the CEO extracted by the MAHD method. Also, Golmakani et al., (2017) reported that a higher antioxidant activity of CEO extracted by HD-based methods is a result of higher eugenol contents. The antioxidant activity of phenolic compounds comes from their chemical structure and reducing functions. They quench the singlet oxygen, neutralize free radicals and chelate transitional metals (Hosseini et al., 2019b). It has been reported that eugenol can prevent the creation of iron and OH radicals, two known chemical agents which contribute to lipid peroxidation (Nagababu et al., 2010).

Tocopheryl acetate, α-tocopherol, and β-carotene reduced the PV of CKO to a similar extent that BHT did after 8 h of storage (Figure 1b). After 12 h, however, tocopheryl acetate, α-tocopherol, and β-carotene were less effective in their functions than BHT was. Also, Keramat et al., (2017a) reported that β-carotene and BHT were equally effective in reducing the PV of virgin olive oil after 21 days of storage. However, β-carotene showed a pro-oxidant activity after 42 days of storage. They concluded that the differences among the antioxidant activities of different antioxidants can be clearly determined at the latter stages of oxidation. The effect of BHT in reducing the PV of CKO was similar to that of the CEO extracted by the MAHD method. After 12 h of storage at 97.8 °C, BHT and the CEO extracted by the MAHD method reduced the PV of CKO by 31.49 and 25.69%, respectively. Furthermore, Golmakani et al., (2018) reported that O. sanctum essential oil is able to inhibit the oxidation of CKO, quite similar to the action of BHA.

The AOM induction period and protection factor of the CKO samples are presented in Table 3. There was no significant difference (P < 0.05) among the AOM IP values of CKO samples containing the different variants of CEOs. Among different CEO samples, the one extracted by MAHD showed the highest PF. The PFs of CKO samples containing BHT and tocopheryl acetate showed no significant differences compared to the PF of the CEO extracted by MAHD. The AOM IPs of the samples containing the MAHD-extracted CEO, BHT, and tocopheryl acetate were 1.53, 1.54 and 1.51 times higher than that of the control sample. This indicates that the CEO extracted by MAHD can replace the function of BHT. Furthermore, α-tocopherol was observed to be less effective than tocopheryl acetate in extending the AOM IP of CKO. A better performance of tocopheryl acetate was observed, which increased the AOM IP of CKO more sharply than the action of α-tocopherol. This can be due to the fact that the esters of α-tocopherol become less susceptible to degradation at elevated temperatures. According to Wüstenberg et al., (2011), the esterification of natural antioxidants can improve their chemical stability, oxidative stability, and heat tolerance, especially in systems which are characterized by the presence of lipids. Here, the AOM IPs and PFs of CKO samples containing β-carotene and α-tocopherol were similar to those which contained CEOs extracted by the HD, SD, and MASD methods.

The role of any antioxidant in retarding the formation of secondary oxidation products in the CKO can be monitored by the Rancimat method. The Rancimat IP of the CKO samples was automatically recorded and taken as the breaking point of the plotted curves (i.e. the intersection point of the two extrapolated parts of the curve). Per antioxidant, the Rancimat IP values of the CKO samples varied when quantified by the Rancimat method (Figure 2). The Rancimat IP of the control sample was significantly lower (2.79 h) which can be attributed to the presence of high amounts of PUFAs in the CKO, hence the susceptibility to oxidation. All antioxidants that were used herein proved to be effective in increasing the Rancimat IP of CKO in comparison to the control. In this regard, Viuda‐Martos et al., (2010) reported that oregano, thyme, clove, sage, and rosemary essential oils were effective in increasing the Rancimat IP of lard. Here, CKO samples that contained CEOs experienced increases in their Rancimat IP values, especially when the CEO had higher amounts of eugenol. For example, the Rancimat IP was extended by 30.11% in the case of the CKO sample enriched with the CEO extracted by MAHD (which had a eugenol content of 88.14%). On the other hand, the Rancimat IP was extended by 18.28% in the case of the sample containing the CEO extracted by MASD (which had a eugenol content of 78.44%). This shows a slower oxidation rate in samples with higher eugenol contents. The CEOs extracted by the HD and SD methods increased the Rancimat IP of CKO by 22.22 and 20.43%, respectively. BHT and tocopheryl acetate were the most effective antioxidants in reducing the production of secondary oxidation products in CKO. Accordingly, BHT and tocopheryl acetate increased the Rancimat IP of CKO by 55.55 and 44.08%, respectively. The Rancimat method revealed that the Rancimat IP values of samples containing CEO (extracted by MAHD) were significantly lower than those of samples enriched with the BHT and tocopheryl acetate. However, through the AOM assay, no significant differences were found between the AOM IP of samples containing CEO (extracted by MAHD) and those containing BHT and tocopheryl acetate. A strong and significant correlation was not found between the AOM IP values of CKO samples through the AOM and their Rancimat IP values obtained by the Rancimat method (R² of 0.48). These results may be related to the fact that the Rancimat method measures the changes in electrical conductivity caused by the generation of volatile compounds by thermal oxidation (Velasco et al., 2004). Thus, in CKO samples containing CEO, the volatile compounds in the CEO could possibly be measured as secondary oxidation products through the Rancimat method. The Rancimat method is not a reliable method for measuring the antioxidant activity of CEO.

The α-tocopherol and β-carotene increased the Rancimat IP of CKO by 22.94 and 25.09%, respectively, indicating that they were less effective than BHT, tocopheryl acetate, and CEO extracted by MAHD. Similarly, a previous study showed that BHT, Zataria multiflora essential oil, and Thymus vulgaris essential oil were more effective than α-tocopherol and β-carotene in reducing virgin olive oil oxidation (Keramat and Golmakani, 2016; Keramat et al., 2017b).

There were significant differences among the fatty acid profiles of the CKOs before AOM (untreated) and after AOM (Table 4). The PUFA content of all CKO samples, particularly EPA and DHA, decreased under accelerated oxidation conditions. Regarding the control group, the amount of PUFA in the CKO was 31.11%, but it decreased to 17.13% at the end of the accelerated storage period. Furthermore, the amounts of EPA and DHA of the control sample were reduced by 49.32 and 49.97%, respectively, in comparison to the untreated CKO. Natural and synthetic antioxidants markedly reduced the loss in PUFA in CKO samples during accelerated storage. CEOs extracted by HD, SD, MAHD, and MASD methods reduced the PUFA contents in the CKO samples by 40.63, 47.03, 29.67, and 42.46%, respectively. On the other hand, the use of BHT, tocopheryl acetate, α-tocopherol, and β-carotene in CKO samples reduced their PUFA contents by 42.14, 40.50, 39.18, and 39.15, respectively. The CEO extracted by MAHD was the most effective antioxidant, followed by β-carotene and α-tocopherol. The effects of BHT and tocopheryl acetate, as revealed by the AOM assay, were similar to the effect of CEO extracted by MAHD, and they were better than the effects of β-carotene and α-tocopherol. Nonetheless, BHT and tocopheryl acetate were less effective when compared to the CEO extracted by MAHD, and also when compared to β-carotene and α-tocopherol, regarding the preservation of the PUFA in CKO. Similar to our findings, Luther et al., (2007) reported that although the extract of black raspberry seed flour was more effective than the extract of Chardonnay grape seed flour in preserving the PUFA of fish oil, the extract of grape seed flour exhibited a stronger capacity for reducing the overall lipid oxidation in fish oil. These results indicate that the capacity of an antioxidant for preserving a selected fatty acid may not be the same as its ability to reduce the overall lipid oxidation of an oil sample. Here, the CEOs extracted by the HD and SD methods protected the PUFAs of the CKO to the same extent that the BHT and tocopheryl acetate did. Similarly, Golmakani et al., (2018) showed that the PUFA contents of soybean oil samples containing O. sanctum essential oil were similar to the PUFA contents of samples containing BHA during storage at both 45 and 60 °C.