National Technical University of Athens (NTUA)
Μελέτη και βελτιστοποίηση διεργασιών ανάκτησης συστατικών υψηλής προστιθέμενης αξίας από παραπροϊόντα αλίευσης και παραγωγής προϊόντων ιχθυηρών
Abstract
dc:descriptionThe aim of this thesis was the valorization of fish filleting and gutting side streams. The valorization of these raw materials involved the isolation of bioactive compounds, such as polyunsaturated fatty acids and proteins, as well as the development of novel fish-based products. For the recovery of bioactive compounds, conventional solvent extraction using different solvents was studied, in comparison with innovative technologies, such as Pulsed Electric Fields (PEF), High Pressure (HP), and Supercritical Fluid Extraction (SFE), either as pre-treatments or as alternative extraction processes. These products included a refrigerated fresh product (fish croquette) and a shelf-stable dehydrated product (instant fish soup powder).The present thesis was experimentally structured into five thematic sections. In the first section, the chemical composition of the side streams was determined in order to identify the bioactive compounds of interest for recovery from each material. The second section focused on the study of conventional extraction of lipids and proteins from evisceration and filleting by-products of sea bass, respectively. In the third section, the effects of PEF, HPP, and SFE on lipid extraction, as well as the impact of PEF on protein extraction, were investigated. The fourth section examined the development of an appropriate encapsulation system for the recovered fish oils. Finally, in the fifth section, filleting by-products were utilized for the development of a dehydrated fish soup and a fish croquette. For the croquette, different edible coatings were tested and compared with uncoated samples with the aim of extending shelf life. In the dehydrated product, strategies to improve stability were evaluated, including the addition of natural antioxidants or the application of suitable mild thermal pretreatments, designed specifically to promote the formation of antioxidant compounds through Maillard reactions.The evisceration by-products of sea bass were richer in lipids (45.7% dry weight), while the filleting by-products were richer in proteins (45.7%). Therefore, the extraction of lipids was studied in the first by-product, and the extraction of proteins in the second. The gutting by-products were characterized by a higher content of polyunsaturated fatty acids (PUFA) (40.8%) and a lower content of saturated fatty acids (SFA) (17.2%) compared to the filleting by-products, which exhibited almost equal concentrations of SFA and PUFA (26.9% and 27.3%, respectively). The ω-3 fatty acids accounted for 21.9% and 10.5% in the case of evisceration and filleting by-products, respectively.For the kinetic study of lipid extraction, a polar solvent (ethanol) and a non-polar solvent (hexane) were employed. Extractions were carried out at different temperatures (20–50°C), solvent-to-solid ratios (10:1–50:1), and extraction times (1–30 min). Lipid extraction was rapid and reached equilibrium within the first 5 min. Ethanol achieved a higher recovery yield from the evisceration by-products under all tested conditions compared to the non-polar solvent, and was able to recover almost the total lipid fraction (99.6%) either at a solvent-to-solid ratio of 50:1 and 35°C, or at 50°C with ratios higher than 30:1. In the case of filleting by-products, hexane achieved higher lipid recovery, reaching complete extraction at 20°C and a ratio of 50:1. The observed differences between the two by-products and solvents were attributed to the composition of the fatty acids in the raw materials. With regard to the composition of the extracted lipids, the PUFA content increased in both by-products when ethanol was used. Furthermore, extracts from filleting by-products exhibited lower peroxide values (PV) and p-anisidine values (p-AV) compared to those from gutting by-products, due to the significantly higher PUFA concentration in viscera. The use of hexane enhanced oxidation, with PV values up to 2-fold higher and p-AV values up to 45% greater compared to ethanol extracts. All p-AV values remained below 20, which is the maximum oxidation limit established by the Codex Alimentarius for fish oils intended for human consumption. The corresponding PV limit is 5 meqO₂/kgoil, with extracts obtained at 20 and 35°C remaining within this threshold. For the selection of the optimal extraction conditions, both extraction yield and oxidation level were considered. The results indicated that ethanol extraction at a solvent-to-solid ratio of 47:1 and 34°C was the optimal condition, achieving 95% lipid recovery without significant alterations in extract quality.For the extraction of proteins from filleting side streams, the extraction kinetics were investigated under different pH values (1, 3, 9, 11, 13), temperatures (25–75°C), solvent-to-solid ratios (10:1–50:1), and extraction times (1–240 min). Alkaline pH values resulted in significantly higher yields compared to acidic pH values. The total protein content (96.4%) was recovered at pH=13, 75°C, and a solvent-to-solid ratio of 50:1. Extraction yield followed a J-shaped curve as a function of pH, with lower values near the isoelectric point of proteins (pI≈4.5) and higher yields as the pH deviated from this point in both directions. The determination of α-amino nitrogen in the protein extracts was used to assess the degree of protein hydrolysis, with values ranging from 3.3 to 29.2 mgFAN/gprotein. The effect of pH on this parameter also followed the characteristic J-shaped curve, with the highest values observed at extreme pH conditions. Protein content exceeded 90% under alkaline conditions, whereas at pH 3 it ranged between 78.8–85.9% and at pH 1 between 59.4–68.2%. Deviation of pH from the pI increased both water binding capacity (WBC) and oil binding capacity (OBC), with the exception of pH=13, where a decrease was recorded. A similar reduction in OBC was also observed at pH=1, due to the high concentration of inorganic components. In the third section, PEF were applied to fresh gutting by-products, using 100 to 1000 pulses to evaluate the effect of this process on lipid extraction. Treatment with 100 pulses did not result in statistically significant differences compared to untreated samples, whereas all other processing conditions led to an increase in yield of up to 12%. However, increasing the number of pulses above 750 reduced the concentration of unsaturated fatty acids by up to 25.5%, while extracts obtained with more than 500 pulses exceeded the aforementioned oxidation limits set for fish oils. Therefore, the optimal condition was 250 pulses, which increased the maximum yield by 6% without altering the quality of the extract compared to untreated samples. HPP was applied either to fresh or dried samples with the solvent (ethanol) at pressures ranging from 200 to 600 MPa. Application of HPP to fresh by-products did not affect extraction yield, but all extracts exhibited lower PUFA concentrations and increased levels of oxidation. When HPP was applied as an extraction method, yields increased by up to 7.2% compared to untreated samples; however, extracts exceeded oxidation limits under all conditions except at 200 MPa.SFE was studied exclusively on Arctic charr filleting by-products, under different pressures (20–45 MPa) and temperatures (40–80°C). Increasing temperature either enhanced, reduced, or had no significant effect on extraction yield depending on the applied pressure, with up to 77% of the contained lipids recovered at maximum pressure and temperature. Moreover, increasing pressure led to higher lipid concentrations at 80°C, whereas no significant differences were observed among extracts obtained at 40°C. The effect of temperature on astaxanthin concentration also depended on pressure, with maximum values recorded at 45 MPa and 80°C, or at 35 MPa and 40°C. Despite variations in astaxanthin content, all extracts showed comparable antioxidant activity. Finally, SFE extracts presented the lowest PV values (0.7–2.1 meqO₂/kgoil), although increasing temperature led to a rise in PV of up to 72%.Regarding the effect of PEF on protein recovery, treatments were applied to hydrated by-products at different electric field intensities (1.1–2.1 kV/cm) and pulse numbers (50–250). Subsequently, the samples were extracted at pH 13, at temperatures of 25 or 50°C, and at solvent-to-solid ratios of 30:1 or 50:1. Increasing both parameters resulted in higher recovery yields and faster extraction rates. Complete protein recovery (95.9%) was achieved when the sample was treated with a specific energy of 11.0 kJ/kg and extracted at 50°C and ratios of 30:1 or higher—thus under milder temperature and solvent requirements compared to conventional extraction. The protein content of the extracts remained stable, except at the maximum specific energy (11.0 kJ/kg), where a reduction of up to 10.5% was recorded. Moreover, the highest field intensity (2.1 kV/cm) led to increased concentrations of free α-amino nitrogen, while no statistically significant differences were observed between the other two intensities. Finally, WBC and OBC values increased by up to 71.8% and 18.3%, respectively, but only in extracts treated with a specific energy above 5 kJ/kg.Moving on the results of the fourth thematic section, Tween 80, sodium caseinate, and their 1:1 mixture were initially tested as emulsifiers. Sodium caseinate produced emulsions with larger mean droplet diameter (MDD) (252 nm) and higher polydispersity index (PDI) (0.37), in contrast to Tween 80, which resulted in the smallest MDD (212 nm) and PDI (0.23). Ultimately, the mixture of the two emulsifiers was selected as the optimal option, as it combined the rapid adsorption and smaller particle size associated with Tween 80 with the protein’s ability to form a denser and more stable interfacial film around oil droplets. In the subsequent stage, optimization of high-pressure homogenization (HPH) was performed by studying different pressures (40–80 MPa) and homogenization cycles (0–10). Increasing pressure from 40 to 60 MPa reduced MDD and PDI values by up to 25% and 19%, respectively, while no further differences were observed between 60 and 80 MPa. Similarly, MDD and PDI significantly decreased during the first four homogenization cycles, but additional passes did not further reduce these parameters. Although the highest pressure (80 MPa) yielded the greatest encapsulation efficiency (EE), the samples exhibited increased oxidation even after the first cycle. Conversely, at 40 MPa no changes in oxidation were observed after 10 cycles, but EE was the lowest among the tested pressures. Conditions of 60 MPa and four cycles were identified as optimal, since lipid oxidation did not differ significantly compared to unhomogenized samples, while EE reached 84.7%. The next stage investigated the choice of encapsulating matrix by testing different maltodextrin concentrations (0–30%) in the feed solution, with or without 1% sodium alginate or 10% gum arabic. Increasing maltodextrin concentration from 0% to 20% improved EE, though no significant differences were observed between 20% and 30%. The maximum EE values achieved at 20% maltodextrin were 82.9% with gum arabic and 79.1% with sodium alginate, which were selected as the most effective encapsulating carriers. The efficiency of encapsulation was further confirmed by FTIR spectra comparisons between encapsulated and non-encapsulated fish oils. Subsequently, the effect of spray-drying and freeze-drying on fish oil encapsulation was evaluated. Spray-drying produced particles with up to 25% higher EE and a more uniform, spherical morphology, whereas freeze-dried samples showed irregular and porous structures. Despite the superior performance of spray-drying, the process exhibited lower throughput and frequent nozzle clogging; therefore, freeze-drying was selected as the encapsulation method. The final stage addressed the storage stability of encapsulated fish oils at 25 °C under controlled relative humidity (RH) of 43% and 75%. During storage, EE gradually decreased, with the most pronounced changes observed at higher RH. The whiteness index (WI) also decreased over time, consistent with EE trends. According to the maximum permissible p-AV for fish oils, non-encapsulated oil reached the critical value within 7 days, whereas the encapsulated samples with maltodextrin combined with sodium alginate or gum arabic reached this threshold after 55 and 45 days at 43% and 75% RH, respectively.Moving on to the results of the final thematic section, the first product developed from filleting by-products was a dehydrated soup. In the shelf-life study, the effect of water activity (aw, range 0.21–0.53) and temperature (20–50 °C) on lipid oxidation, color changes, and sensory attributes was evaluated. The oxidation rate reached a minimum (0.36 meqO₂/(kgoil٠d)) at aw=0.43. Based on peroxide value limit (20 meqO₂/kgoil) as the shelf-life criterion, the product’s estimated shelf life was 21, 39, 55, and 43 days at aw levels of 0.21, 0.32, 0.43, and 0.53, respectively. Regarding the effect of temperature, increasing storage temperature reduced lipid oxidation rates, with induction phase rates of 0.94, 0.29, and 0.01 meqO₂/(kgoil٠d) at 20, 35, and 50°C, respectively, corresponding to calculated shelf lives of 21, 69, and 2083 days. This trend was attributed to Maillard reactions producing antioxidant compounds. Color variation (ΔE) was most pronounced at the beginning of storage and subsequently stabilized at different points depending on temperature. Sensory evaluation revealed that when color acceptability reached the rejection threshold, ΔE equaled 28. Consequently, the shelf-life limit based on color at 50°C was 20 days, whereas at 20 and 35°C the ΔE never reached the rejection threshold. Therefore, shelf life was determined by color changes at 50°C, while at the two lower temperatures it was limited by lipid oxidation.The production of antioxidant compounds during storage at 50°C was experimentally confirmed, as antioxidant capacity increased by up to 216% after 6 days. Based on these findings, samples were pretreated at 50°C for 6 days and subsequently stored at 20°C and at 21%–53% RH. The oxidation rates of the pretreated samples were 66%–80% lower compared to the corresponding untreated samples, resulting in a significant extension of shelf life by 83–174 days. On the other hand, the addition of natural antioxidants derived from rosemary at the maximum permissible concentration reduced oxidation rates by 75% compared to the samples without antioxidants, leading to a shelf-life extension of 60 and 198 days at 20°C and 35°C, respectively. The addition of the antioxidant did not affect the sensory characteristics, as it was not perceived either in the dehydrated product or in the reconstituted fish soup.The second developed product in the present research was a fish croquette, enriched with the recovered proteins. The use of edible coatings made from chitosan or gelatin was tested to extend the product's shelf life under refrigerated storage (0°C–12°C). To assess spoilage kinetics, the effect of the edible coating on the development of spoilage microorganisms, lipid oxidation, pH, color, texture, and sensory characteristics was examined. Chitosan significantly inhibited microbial growth during storage by extending the lag phase by 2 to 7 days compared to the control samples and reducing the growth rate of the total viable count (TVC) by 18%–59%. On the other hand, gelatin reduced the growth rate by 15%–46%. For calculating the shelf life of the products, a limit of 7 log(CFU)/g for the TVC was used, as proposed for fish products. The results showed that the shelf life of the uncoated samples was 14, 8, 6, and 3 days at 0°C, 4°C, 8°C, and 12°C, respectively, 16, 9, 7, and 5 days, respectively, for the gelatin-coated samples and increased to 23, 14, 9, and 5 days for the chitosan-coated samples, respectively. Pseudomonas spp. was among the dominant spoilage microorganisms in this product, with results consistent with those for TVC. Among other examined microorganisms, Brochothrix thermosphacta, Enterobacteriaceae spp. and Shewanella spp. showed similar behavior in response to chitosan application. Chitosan inhibited their growth almost throughout the shelf-life study. Different results were observed in the case of lactic acid bacteria. In chitosan-coated samples, lactic acid bacteria developed, reaching levels of 5.0–5.5 log(CFU)/g, while in the uncoated samples, growth did not exceed 3.5 log(CFU)/g. The different effect of chitosan on this specific group of microorganisms is explained by the lower initial pH (5.9) compared to the other samples (6.3), which favors the growth of lactic acid bacteria. During storage, a slight increase in the degree of lipid oxidation was observed, with the uncoated samples showing greater oxidation. However, the values remained at low levels throughout the storage period, without any unpleasant sensory characteristics due to oxidation. Regarding the results of the overall change in colour and texture, no clear trend was recorded regarding the effect of coating or temperature or time.The results of this PhD thesis highlight the potential and high efficiency of converting fish industry side streams from low-value by-products into, either high nutritional and added-value components—through green solvents or the application of innovative extraction processes—or into innovative food products with high nutritional value. The applications of the studied aquaculture-derived alternatives align with the principles of the circular economy and the reduction of the volume of generated by-products.
Degree
thesis:*- Grantor dc:publisher
- National Technical University of Athens (NTUA)
- Year dc:date
- 2025
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Semenoglou, Ioanna
- Σεμένογλου, Ιωάννα
Subjects
dc:subject × 15- Παραπροϊόντα ιχθυηρών
- Αξιοποίηση
- Μη θερμικές διεργασίες
- Εγκλεισμός ιχθυελαίων
- Ανάπτυξη προϊόντων
- Fish by-products
- Valorization
- Nonthermal processing
- Encapsulation
- Product development
- Επιστήμη Χημικού Μηχανικού
- Επιστήμες Μηχανικού και Τεχνολογία
- Χημική μηχανική
- Chemical Engineering
- Engineering and Technology
Rights
- Language dc:language
- gre
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/59874