Crude fish oil was obtained from the Dardanel Company, located in Çanakkale, Turkey. The walnut shell, date seed, spent coffee and spent black tea were provided from local patisseries and cafés in Çanakkale, Turkey. The sesame hull was obtained from local sesame paste manufacturers (Özen Baharat & Tohum Ltd. Co.) in Çanakkale, Turkey. All chemicals utilized in this study were of analytical grade and were procured from Sigma (USA) and Merck (Darmstadt, Germany).

Various food waste materials, including date seeds, spent coffee, and spent black tea (excluding sesame hulls and walnut shells), were initially dried in a vacuum oven at 45 °C and subsequently ground using a laboratory grinder. The grinding process was not applied to the sesame hull, spent coffee, or spent black tea since particle sizes are sufficient for extraction. To extract polar phenolic compounds, 10±0.2 g of the samples were weighed into an Erlenmeyer flask, and 100 ml of ethanol were added. The sesame hull, walnut shell and date seed were stirred at 24 h, while the spent coffee and spent black tea were stirred at 4 h. The stirring process was carried out at room temperature and in darkness without repetition. Following the extraction process, the mixture was filtered through filter paper, and the filtrate was centrifuged at 5000 rpm for 10 minutes. After centrifugation, the clear upper liquid phase was collected. Subsequently, at least 90% of the solvent was evaporated using a rotary evaporator at 45 °C under vacuum conditions. The remaining portion was then reconstituted to a final volume of 10 ml with ethanol, and this solution was used as a stock solution, which was stored at -18 °C.

The total polar phenolic content (TPC) and antioxidant capacity of the extracts obtained from different food wastes were determined according to the Folin-Ciocalteu method (Singleton and Rossi, 1965) and the radical-α, α-diphenyl-β-picrylhydrazyl (DPPH) assay (Benvenuti et al., 2004), respectively. The total phenolic contents and DPPH assay results (50% inhibition concentration, IC₅₀) of the fish oil samples are given in Table 1. Both analyses were used to determine the BHA equivalents of the polar extracts for the addition level of the fish oil.

The fish oil samples, both without additives and with the addition of BHA and the BHA:Tocopherol (TCL) mixture (BHA:TCL), served as control samples. It needs to be noted that the permissible limit for BHA in edible oils, as defined by the Codex Alimentarius (General Standard for Food Additives CODEX STAN 192-1995, 2017), is set at 200 ppm. For this reason, the IC₅₀ values of 200 ppm BHA were measured, and the amount of polar extracts that provided inhibition equivalent to the IC₅₀ value of 200 ppm BHA was calculated. The addition levels are presented in Table 1. The prepared fish oil samples were stored at temperatures of 25 and 35 °C for a duration of 60 days to assess the impact of the added extracts on the oxidative stability of the oil.

The acid (Ca 5a-40), K₂₃₂ (Ti 1a-64) and peroxide (Cd 8-53) values were measured according to AOCS (Firestone, 2004). The viscosity values were determined using with Brookfield viscosimeter (model DV II + Pro with Rheocalc software, Brookfield Eng. Lab., Inc., MA, USA) equipped with a LV-SC4-18 spindle at 25 °C, and the results were expressed as (cP). The total carotenoid and chlorophyll values were determined according to the method detailed by Minguez-Mosquera et al. (1991). In accordance with these methods, 7.5±0.1 g of oil samples were dispensed into tubes and subsequently topped up to 25 ml with cyclohexane. The resulting mixture was then subjected to spectrophotometric measurements at wavelengths of 470 and 670 nm using a spectrophotometer (Shimadzu UV-1800; Japan). Throughout the storage period, various parameters, including K₂₃₂ (Conjugated Dien, CD), acid value, peroxide value, and viscosity, were monitored. Additionally, the assessment of total carotenoid and chlorophyll content was conducted on both freshly prepared and stored samples.

The thermal behavior of the fish oil samples was determined by thermogravimetric analysis (TGA) using a thermogravimetric analyzer (TGA 4000, Perkin-Elmer, USA). For thermogravimetry and derivative thermogravimetry (TG/DTG) measurements, the samples weighing 5-10 mg were placed into the TG pan. The samples were then heated from 20 to 700 ºC at a rate of 10 ºC/min under dry air, and the weight loss and deterioration temperatures were determined using the TGA software (Pyris Manager).

The present study was planned duplicate, and all analyses were performed in triplicate. The data obtained from the samples were presented as “Mean±Standard Deviation (SD)”. The data were evaluated with the MINITAB (Minitab 16v) statistical software using Analysis of Variance (ANOVA), similarities and differences among samples were determined according to Tukey’s multiple test at p ≤ 0.05.

The total phenolic content (TPC) and 50% inhibition concentration (IC₅₀) values for the additives used in the fish oil are shown in Table 1. The TPC of the extracts obtained from natural waste/by-products ranged from 200.60 to 2156.00 mg/L. The date seed extracts showed the highest TPC, while the spent tea extracts presented the lowest value. The TPC for samples containing 200 ppm of BHA, BHA:TCL (1:1, w: w), and TCL were measured at 135.63 mg/L, 84.38 mg/L, and 25.31 mg/L, respectively. The IC₅₀ values for the natural extracts ranged from 6.64 to 123.30 µg/mL, with their antioxidant activity ranked in descending order as follows: date seed > walnut shell > spent coffee > spent black tea > sesame hull. In comparison, the IC₅₀ values for 200 ppm of BHA and the BHA:TCL (1:1; w:w) solution were determined as 166.20 µg/mL and 272.21 µg/mL, respectively. These results demonstrate that all extracts derived from natural waste and by-products exhibited higher TPC and antioxidant activity when compared to both 200 ppm of BHA, BHA:TCL (1:1; w:w), and TCL. Yang et al. (2014) reported that walnut shell extracted using different solvents (water, chloroform, methanol, ethanol, ethyl acetate, and n-butanol) and the highest TPC were observed in the sample extracted with ethyl acetate (200.40 mg GAE/g). Afifi et al. (2017) reported that the TPC of date seed powder ranged from 7.63 to 71.72 mg GAE/100 g, depending on the extraction conditions. In a study conducted by Hwang et al. (2019), the TPC values for acetone and ethanolic extracts of spent coffee were reported as 51.79 and 80.36 mg GAE/g extract, respectively. The same study reported % inhibition values for 50, 100, and 200 ppm concentrations of spent coffee extracts, ranging from 22.06 to 91.66, while the % inhibition values for BHT varied between 79 and 96.44 at the same concentrations, respectively. Bravo et al. (2013) reported the antioxidant capacities of the spent coffee extracts obtained from different extraction process measured with ABTS and DPPH method to range from 15.31-152.64 and 5.02-82.40 µmol Trolox/gdm. These variations between both literature findings and our own results can be explained by the different extraction conditions, solvents and type of waste materials used. Another reason for the difference between the findings can be the different antioxidant capacity measurements used such as radical scavenging activity method (DPPH), total phenolic compound (TPC), trolox equivalent antioxidant capacity determination (TEAC) using ABTS radical cation, cupric ion reduction capacity (CUPRAC), ferric-reducing antioxidant power (FRAP) or oxygen radical absorbance capacity (ORAC) (Shahidi and Zhong, 2015). These findings in the literature align closely with our own results, especially regarding the % inhibition obtained from the DPPH assay.

The acid (AV) and peroxide values (PV) for both the control and antioxidant-added groups of fish oil, stored at both 25 and 35 °C for a duration of 60 days are presented in Figures 1 and 2, respectively. All additive levels of the extracts from natural sources were determined as the amount equivalent to 200-ppm BHA activity, which is the permissible amount for edible oils according to Codex Alimentarius (2017) (Table 1). The AV of the control samples stored at 25 and 35 °C ranged between 4.42-8.15 and 4.56-8.57 mg KOH/g, respectively. The AV of all fish oils increased during the storage period, while the antioxidant-added samples showed lower AV (Figure 1). The samples stored at 25 °C exhibited slightly lower acid values (AV) than those stored at 35 °C. However, it is important to note that these differences were determined to be statistically significant. The samples supplemented with date seed and walnut shell extracts consistently displayed lower acid values (AV) compared to the other samples at the conclusion of the storage period, and this trend was observed at both temperatures. There were statistically significant differences among the samples in terms of AV (p ≤ 0.05). The differences between the extracted fish oil samples may be explained by the different AV values of the added extracts. As is well known, the lipolysis and hydrolysis reactions of triglycerides cause an increased amount of free fatty acid as well as the acid values of the oils. In addition, a synergistic effect between oxidation and hydrolysis reactions occurred. The extraction process, treatments (refining, frying etc.) and storage conditions affect the free fatty acid and peroxide values of the oils (O’brien, 2008). Therefore, another reason for the difference between acid values can be the differences between the oxidative stability of the fish oil samples. In other words, this situation can be explained by the different resistance of the added antioxidants against oxidation.

One of the most important oxidation indicators of edible oils is peroxide value, and PV shows initial oxidation level as well as providing information about the storage conditions of oils (Iqbal et al., 2008). The PV values for the control samples stored at 25 and 35 °C ranged between 5.46-35.21 and 10.89-44.49 meq O₂/kg, respectively. The PV of BHA-added samples stored at 25 °C ranged from 10.88 to 17.22, while that of the BHA:TCL-added fish oils ranged from 12.26 to 28.82 meq O₂/kg during the storage period. Similar to the AV of the fish oils, the PV values increased during the storage period. As expected, the PV values of the fish oils stored at 35 °C were higher than the samples stored at 25 °C (Figure 2). Furthermore, the fish oil containing date seed extract exhibited lower PV at both storage temperatures. When comparing the control samples and those supplemented with tocopherol to the natural antioxidant extracts obtained from food waste and/or by-products, it became evident that all of the natural extracts provided superior protection against oxidation compared to both the control samples and those with added tocopherol. However, in comparison to the natural extracts added to fish oils, it was observed that the samples with BHA exhibited greater protection against oxidation than the spent coffee and black tea extracts at both storage temperatures. Additionally, sesame hull extracts showed very similar PV results to BHA at both temperatures. However, the most remarkable findings were that fish oils added with walnut shell and date seed extracts had lower peroxide values than those with BHA addition at both temperatures. As is well known, lipid oxidation is a complex chain reaction basically consisting of beginning, propagation and termination. Temperature, light, moisture, presence of oxygen, physical and chemical properties of the substrate, and the presence of oxidation initiators or catalysts all affect the course of oxidation (O’brien, 2008). Antioxidants are defined as agents that delay and/or prevent the oxidation reaction. There are two main groups -primary and secondary antioxidants. The primary antioxidant group comprises free radical scavengers and chain-breaking agents; while the secondary group is involved in the deactivation of metals, inhibition of the breakdown of lipid hydroperoxides to unwanted volatile products, and the regeneration of primary antioxidants (Koleva et al., 2002). Therefore, the differences among the samples mentioned above can be explained by the fact that antioxidants themselves work out their protective properties at different stages of the oxidation process and through different mechanisms.

Figure 3 presents the conjugated diene (CD) values, and Figure 4 presents the viscosity values of the fish oil samples. The CD value is a significant parameter for assessing the degree of degradation in oils and evaluating the effectiveness of the antioxidants employed (Iqbal et al., 2008). The average of the CD values for the fresh samples was 2.27. Consistent with the AV and PV findings, the CD values showed an increase during the storage period at both temperatures. As expected, the fish oil without any antioxidants showed the highest CD value at the end of the storage period, surpassing both the BHA: TCL- and BHA-supplemented samples. Furthermore, in line with the BHA: TCL- and BHA-added fish oils, the fish oils enriched with natural extracts displayed lower CD values than the control samples (those without antioxidants) (Figure 3). These observations align with previous research. For instance, Luther et al. (2007) reported that black raspberry seed extracts significantly reduced the degradation of n-3 polyunsaturated fatty acids, akin to our results. Similarly, Uçak (2018) found that fish oil samples containing propolis (at 500 and 1000 ppm concentrations) had lower PV and CD values compared to both control samples and those supplemented with BHT, which is consistent with our findings. Hwang et al. (2019) reported that the CD and PV values for fish oils with different concentrations added with spent coffee extracts varied between 0.51-4.87 mmol/L and 6.02-8.37 meq O₂/kg during a 14-day storage period. Additionally, the fish oils treated with walnut shell and date seed extracts exhibited lower CD values than those treated with both the control samples and those added with BHA or BHA:TCL. Furthermore, the fish oils treated with sesame hull extracts displayed CD values which are similar to those treated with BHA at both storage temperatures. The AV, PV and CD results proved that sesame hull, walnut shell and date seed extracts could be used as natural antioxidants instead of BHA and tocopherol against oil oxidation. The fish oils supplemented with spent coffee and black tea extracts demonstrated effectiveness in preventing oil oxidation, although their efficacy was slightly less pronounced compared to the other extracts. These variations in effectiveness may stem from differences in the efficiency of the extraction process. Factors such as extraction temperature, duration, and the choice of solvent can significantly influence the antioxidant activity of the extracts. Consequently, it is possible that optimizing these extraction parameters could further enhance the antioxidant activities of sesame hull, walnut shell, and date seed extracts. Recently, in a study conducted by Hwang et al. (2019), the antioxidant activity of spent coffee extracts was evaluated in comparison to BHT in soybean and fish oils. Their research revealed that acetone extracts exhibited greater antioxidant activity when contrasted with ethanolic extracts, and that an escalation in the addition levels of acetone extracts correlated with improved antioxidant activity. Furthermore, the researchers posited that spent coffee extracts had the potential to serve as natural antioxidants for omega-3 oils and fish oils within the same study. These findings in the literature corroborate the outcomes of our own investigation.

The viscosity values of all fish oils were significantly influenced by the additives used, storage time, and temperature, with statistical significance at the p ≤ 0.05 level. These viscosity values are presented in Figure 4. Notably, the control and fish oils supplemented with BHA: TCL were more susceptible to changes in viscosity compared to fish oils with BHA and natural extracts at both storage temperatures. Conversely, the fish oils treated with sesame hull extracts exhibited viscosity levels which were similar to those treated with BHA, while fish oils treated with walnut shell and date seed extracts demonstrated lower viscosity values. The observations mentioned above underscore that natural extracts derived from sesame hull, walnut shell, and date seed were at least as effective as BHA in preserving oil viscosity, in addition to mitigating oxidation.

As is known, fish cannot synthesize carotenoids, but can obtain them from natural or commercial feeds containing carotenoids. Carotenoids also affect the oxidative stability of oils due to their antioxidant activities (Selim et al., 2021). The total carotenoid (TCar) contents in the fresh and stored fish oils are given in Table 2. According to Table 2, the TCar contents in the samples stored at 25 °C ranged from 0.38 mg/kg to 0.52 mg/kg, while those of the samples stored at 35 °C ranged from 0.38 mg/kg to 0.50 mg/kg. Selim et al. (2021) reported that Tcar values for mackerel waste oils were 128.34 mg/g and 98.12 mg/g for sardine waste oil. The disparities between the findings in the literature and our own results may be attributed to several factors, including variations in fish species, geographic regions, dietary compositions, and other environmental factors. The fish oils stored at 35 °C showed a higher loss in carotenoid content than the samples stored at 25 °C. The highest loss in carotenoid content during storage was observed in the fish oil without antioxidants at both storage temperatures. Compared to the control samples, it was observed that the loss in carotenoids was less in the fish oils with natural extracts at both temperatures. The total chlorophyll (Tchl) contents in the fresh and stored fish oil samples are given in Table 3. As seen in Table 3, the Tchl values for the fish oils with and without additives ranged from 4.81 mg/kg to 4.50 mg/kg during the 60-day storage period. Similar to our findings, Koning (1999) reported that the Tchl contents of anchovy and pilchard oils ranged from 2.0 to 37.0 mg/l. Both storage time and temperature were significantly effective on the chlorophyll contents in the fish oils (p ≤ 0.05). The highest Tchl loss was observed in the control samples at both temperatures. Although the changes in the treated fish oil samples may have appeared minor, they were statistically significant. Notably, the fish oils treated with natural extracts, especially when compared to BHA-treated fish oils, exhibited higher total carotenoid (Tcar) values at the conclusion of the 35 °C storage period. When both the “Tcar” and “Tchl” results of the samples are considered together, it becomes evident that natural extracts possess greater antioxidant activity than both BHA and tocopherol. Additionally, these natural extracts are effective in preventing content loss, particularly in oils stored at elevated temperatures.

The thermal decomposition process of the fish oil samples is presented in Figure 5. The initial temperatures for the thermal decomposition (Ton) of the samples obtained from the thermograms were significantly different from each other. The control sample (171.26 °C) presented lower Ton values, while the BHA-added fish oil (196.44 °C) had quite similar Ton to the fish oil added with date seed extract (194.97 °C). Similar results for hoki and tuna oils were reported by Tengku-Rozaina and Birch, (2016). In addition, Dweck and Sampaio (2004) reported that the Ton values of olive and corn oils were higher than the fish oils examined in this study. The thermal decomposition of the fish oil samples occurred in three distinct stages (Figure 5). The samples exhibited significant differences in temperature ranges spanning from 300 to 375 °C, 375 to 450 °C, and 450 to 600 °C. However, no thermal changes were observed in any of the fish oils at temperatures of 600 °C and above. Different oils may have different decomposition temperatures and weight loss rates, depending on their fatty acid composition. Nonetheless, the disparities observed among the samples can be attributed to the use of different additives, despite the fact that the oils used in this study were the same. When examining the weight loss of the fish oils, it became evident that the thermal stability of the fish oil samples followed the order: fish oil with BHA > fish oil with BHA: TCL > fish oil (FBH > FBT > FO). Conversely, the fish oils supplemented with natural extracts exhibited higher weight loss temperatures and lower weight loss rates compared to both the control and FBT samples. Nevertheless, the thermal stability of the fish oils treated with natural extracts exhibited a different trend, with the following order observed: fish oil with date seed extract > fish oil with walnut shell extract > fish oil with sesame hull > fish oil with spent black tea > fish oil with spent coffee (FDS > FWS > FSH > FST > FSC). In alignment with the PV and CD results, the TGA results indicated that the fish oil sample treated with date seed extract (FDS) demonstrated the closest antioxidant activity to that of BHA. The outcomes of this study reaffirm that extracts with high phenolic content can enhance the oxidative stability of fish oils. These findings provide further evidence of a linear relationship between the total phenolic component content and antioxidant capacity.