High freeze-thaw stability of Pickering emulsion stabilized by WPI-carboxymethyl cellulose particles and its effect on frozen pork patties as a fat substitute
Huiyun Y. Zhang, Xinling Li, Huaibin Kang
TL;DR
This study shows that a specific Pickering emulsion can replace fat in pork patties, improving their stability and quality during freezing and thawing.
Contribution
A novel Pickering emulsion using WPI-CMC particles is shown to effectively substitute fat in frozen meat products while maintaining quality.
Findings
WPI-CMC emulsions with 0.6% CMC showed excellent freeze-thaw stability due to a dense interfacial film.
Substituting 75% pork fat with the emulsion reduced thawing loss and oxidation, preserving texture stability.
100% fat substitution impaired performance due to excessive protein cross-linking.
Abstract
This study investigated the freeze-thaw (FT) stability of whey protein isolate-carboxymethyl cellulose (WPI-CMC) Pickering emulsions (PEs) and their performance as fat substitutes in pork patties under repeated FT cycles. The PE containing 0.6% CMC exhibited superior FT stability, characterized by uniform droplet size, low creaming index, and phase separation resistance, resulting from a dense interfacial film that mitigated ice crystal damage. In pork patties, substituting 50%–75% of pork fat with PEs significantly improved quality during FT cycles. After five cycles, the 75% substitution group exhibited significantly enhanced water-holding capacity, as evidenced by 41.5% lower thawing loss and 30.2% reduced expressible moisture, along with inhibited lipid and protein oxidation, demonstrated by 45.5% lower TBARS and 38.6% fewer carbonyl groups, thereby maintaining superior texture…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6
Figure 7- —Major science and technology project of Henan Province
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsProteins in Food Systems · Meat and Animal Product Quality · Microencapsulation and Drying Processes
Introduction
Frozen pork patties, a typical minced meat gel product, have long been favored by consumers for their convenient consumption, rich flavor, and wide price accessibility (Pan et al., 2021; Zhang et al., 2024). To meet consumer demands for juiciness and tenderness, traditional frozen pork patties typically contain a high proportion of animal fat. However, this fat is rich in saturated fatty acids, and excessive intake has been linked to an increased risk of cardiovascular disease and obesity (Boada et al., 2016; Maki et al., 2021). Frozen pork patties depend on low-temperature storage for long-distance transportation and shelf-life extension; however, they are frequently subjected to multiple freeze-thaw (FT) cycles caused by temperature fluctuations during processing, storage, and distribution (Zhang et al., 2023a, b, c). Pork patties with high animal fat content may experience noticeable quality changes during FT cycles. Since animal fat remains almost solid at low temperatures and has limited flexibility, it is less effective at filling the gaps within the minced meat gel structure (Peyronel et al., 2010). This inability to buffer structural stress results in uncontrolled ice crystal growth, where large ice crystals pierce muscle fibers, disrupt cell membranes, and cause irreversible water loss. Meanwhile, fats and proteins undergo co-oxidation, accelerating lipid oxidation and protein denaturation (Chen et al., 2022). These changes collectively reduce the product’s juiciness, harden its texture, and impair its sensory quality, ultimately diminishing commercial value. Thus, reducing animal fat content while controlling FT cycles-induced quality loss has become a critical opportunity and challenge for the frozen meat processing industry.
Pickering emulsions (PEs), stabilized by solid nanoparticles at the oil-water interface, present a promising strategy for developing healthier food formulations. Compared to conventional emulsions, PEs form highly stable interfacial films that effectively mimic the textural properties of fat while allowing for the incorporation of unsaturated oils (Joseph et al., 2025; Sun et al., 2022). This offers a viable pathway to reduce saturated fat content without compromising sensory quality, thus aligning well with current nutritional health trends. For instance, lentil protein-stabilized PEs have replaced fat in sausages with no sensory compromise (Galvão et al., 2024), while nano-cellulose PEs reduced cooking loss and enhanced springiness in emulsified meats (Ji & Wang, 2023). Modulating the interfacial structure of emulsion droplets is a critical strategy to enhance the performance of PEs. Specifically, composite nanoparticles, such as those formed by proteins and polysaccharides, show significant potential in stabilizing these emulsions and improving their functionality.
Whey protein isolate (WPI), a well-known protein in the food industry, has been extensively studied for its emulsifying properties (Fan et al., 2021). It can adsorb at the oil-water interface, creating a stable layer that prevents droplet coalescence. However, like other single-protein-stabilized PEs, WPI-stabilized emulsions may encounter challenges under certain environmental conditions (Yan et al., 2020). For example, changes in pH, temperature, or ionic strength can disrupt the protein’s structure at the interface, leading to emulsion instability (Ma et al., 2025). Carboxymethyl cellulose (CMC), a widely used anionic polysaccharide, has gained significant interest in the food industry due to its unique functional properties. CMC is renowned for its excellent water-binding capacity, film-forming ability, and thickening effect, making it highly suitable for various food applications (Ma et al., 2025). Studies have shown that CMC can interact with proteins, potentially modifying the interfacial behavior of protein-based emulsions (Feng et al., 2024; Sun et al., 2024). Nevertheless, the combined impact of CMC and proteins on FT stability in meat matrices remains underexplored.
Therefore, this study utilizes WPI and CMC as raw materials to prepare high FT stable PEs by optimizing CMC concentrations to explore the optimal preparation conditions. Meanwhile, the particle size distribution, visual characteristics, microscopic morphology, and creaming index of the PEs prepared under different CMC concentrations following FT cycles were measured to study their FT stability. Furthermore, the optimally stable WPI-CMC PEs were incorporated into pork patties as a partial substitute for pork back fat at varying replacement levels. Thawing loss, expressible moisture, color, texture profile, lipid and protein oxidation, and water distribution during FT cycles were systematically evaluated to elucidate the effect of highly freeze-thaw-stable WPI-CMC PEs on the FT stability of frozen pork patties. This study lays a foundation for the development of food-grade WPI-CMC composite Pickering particles and offers a promising approach to creating green, healthy fat substitutes. It also offers theoretical support for the application of PEs as fat substitutes during the freezing process of pork patties, facilitating their commercial development and practical utilization.
Materials and methods
Materials
WPI with a protein purity exceeding 90% was procured from Inner Mongolia Yili Industrial Group Co., Ltd. (Hohhot, China). CMC with a degree of substitution of 0.9 and an average molecular weight of approximately 250 kDa was supplied by Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). The sunflower seed oil was supplied by Xinxing Edible Oil Co., Ltd. (Guangzhou, China). Fresh porcine tissue was provided by Shuanghui Food Co., Ltd. (Luoyang, China). The samples were collected 48 h post-slaughter. The lean muscle tissue, carefully dissected from the Longissimus lumborum muscle (with a sample size of n = 18), was found to have a moisture content of approximately 72%, a protein content around 22.5%, and a lipid content of about 3%. As for the back adipose tissue, it contained around 90% lipid, 7.5% moisture, and 0.5% ash. All analytical-grade chemicals utilized in subsequent analyses were sourced from Macklin Biochemical Co., Ltd. (Shanghai, China).
Preparation of Pickering emulsion stabilized by WPI/CMC complex nanoparticles
Pretreated WPI solutions were prepared following Wang et al. (2024) with minor modifications. WPI powder (2%, w/v) was dispersed in ultrapure water, stirred for 2 h, and refrigerated overnight at 4 °C. The solution was adjusted to pH 12 with 1 M NaOH, heated at 95 °C for 30 min, cooled, and centrifuged at 12,000 rpm for 10 min to collect the supernatant. Meanwhile, CMC powder (1%, w/v) was dissolved in ultrapure water by stirring for 4 h, followed by overnight refrigeration at 4 °C. The pretreated WPI solution was mixed with CMC solutions to obtain final CMC concentrations of 0%, 0.2%, 0.4%, 0.6%, and 0.8% (with a constant WPI concentration of 1%, w/v). The pH of the mixtures was adjusted to 3 using 2 M HCl and stirred at 900 rpm for 1 h to promote complexation. Finally, sunflower seed oil was added to the WPI/CMC mixture at an oil-to-water ratio of 1:1. The mixture was pre-mixed at 200 rpm for 10 min and then homogenized at 15,000 rpm for 3 min at 25 °C using an FSH-2 A homogenizer to form stable PEs.
Freeze-thaw stability of the emulsion
The FT stability of the emulsion was analyzed following the procedure reported by Zhu et al. (2020a, b). Emulsion samples were frozen at − 20 °C for 24 h, then thawed at room temperature for 2 h, and this cycle was repeated three times.
Particle size distribution
Static light scattering was employed to analyze the particle size distribution of PEs at 25 °C. Measurements were conducted using a BeNano 90 Zeta analyzer (Baite Instruments, Dandong, China), based on the methodology described by Zhu et al. (2020a, b). The refractive indices of oil and water were set at 1.47 and 1.33, respectively. Samples were diluted with deionized water until the diffraction cell obscuration exceeded 10%.
Visual and microscopic analysis
Visual changes of the emulsion during FT cycles were observed. Fluorescent inverted microscopy (ECLIPSE Ts2R, Nikon, Shanghai) analyzed morphological changes before and after cycles. Nile blue (0.1%, 30 µL) and Nile red (0.01%, 20 µL) were added to PEs, incubated for 20 min. A 5-µL aliquot was placed on a slide, covered, and sealed with nail polish for observation at excitation wavelengths of 633 nm (Nile blue) and 488 nm (Nile red).
Creaming index
The creaming index was determined to evaluate the gravitational stability of PEs during FT cycles. Emulsion samples (5 mL) were transferred to graduated centrifuge tubes (10 mL) and subjected to the same FT cycles as described in Sect. 2.3. After each cycle, the tubes were allowed to stand at room temperature for 1 h to ensure complete phase separation. The height of the creamed layer (Hc, mm) at the top of the tube (characterized by oil droplet aggregation) and the total emulsion height (Ht, mm) were measured. The creaming index was calculated using the following formula:
\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\text{Creaming }}{\text{Index }}\left( \% \right)\,=\,{\text{Hc/Ht }}\times 100$$\end{document}Preparation of pork patties with WPI-CMC stabilized Pickering emulsion
Pork patties were prepared by substituting pork back fat with PEs at 0%, 25%, 50%, 75%, and 100% (composition in Table 1). Lean meat and back fat were diced (approximate edge length of 30 mm), ground, and mixed with salt, chilled water (4 °C), and PEs via stepwise chopping (2 min, 1 min, 1 min). Patties (2 cm thickness, 8 cm diameter) were formed, wrapped in low-density polyethylene bags, and divided into five groups (n = 20). For FT cycles, patties were frozen at − 30 °C for ≥ 20 h (internal temperature ≤ − 18 °C), stored at − 18 °C for 7 days, and thawed at 4 °C for 12 h. Samples were analyzed after 0, 1, 3, and 5 cycles (F0–F5), with four replicates per group.
Table 1. Recipe for Preparing pork patties using Pickering emulsion stabilized by WPI-CMC nanoparticles as fat substitutesIngredient(g)R1R2R3R4R5Lean meat6060606060Pork backfat1813.594.50Pure water2015.5116.52Pickering emulsions09182736Salt22222Note R1, R2, R3, R4, and R5 correspond to the replacement of pork back fat with Pickering emulsion at levels of 0%, 25%, 50%, 75%, and 100% respectively. WPI, whey protein isolate; CMC, carboxymethyl cellulose
Physicochemical analyses
Thawing loss
Following the approach proposed by Pan et al. (2021), thawing loss assessment commenced when the internal temperature of frozen pork patties reached 4 °C during thawing. The thawing loss was quantified using the following formula:
\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\text {Thawing loss}\,(\%) =\:\:\frac{{\mathrm{M}}_{\mathrm{1}}\mathrm{-}{\mathrm{M}}_{\mathrm{2}}}{{\mathrm{M}}_{\mathrm{1}}}\times100\%$$\end{document}where M_1_ stands for the sample mass (g) before freezing treatment, and M_2_ refers to the sample mass (g) after the thawing process.
Expressible moisture
Based on the experimental procedure reported by Zhang et al. (2023a, b, c), 10-g patty samples were centrifuged at 1760 × g for 10 min at 4℃ to determine expressible moisture. The calculation formula for expressible moisture content (%) is as follows:
\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\text {Expressible moisture}\, (\%) = \:\frac{{\mathrm{M}}_{\mathrm{3}}\mathrm{-}{\mathrm{M}}_{\mathrm{4}}}{{\mathrm{M}}_{\mathrm{3}}}\times100\%$$\end{document}In this formula, M_3_ corresponds to the sample mass prior to centrifugation, while M_4_ represents the mass of the sample after centrifugation.
Color measurement
The color of pork patties before and after FT cycles was measured using a Minolta Chroma-meter^®^ CR-400 at room temperature. The instrument was calibrated with a standard white plate (CIE-Lab*: L*=97.49, a*=−0.21, b*=0.67). Each patty was cut into three slices, and CIE-Lab* values (L*, a*, b*) were measured at five random points per slice (8-mm aperture, D65 illuminant, 2° observer angle). Three replicates were taken per point, with averages calculated for each slice.
Texture profile analysis
Thawed patties were cooked in a preheated oven at 180 °C for 10 − 12 min until internal temperature reached 80 °C, then cooled. Cylindrical samples (2 cm diameter, 1 cm height) were cut from the center, and Texture Profile Analysis (TPA) was performed using a TA-XT Texture Analyzer with a P/50 probe (5 cm diameter). The two-cycle compression test parameters: pre-test/test speed 2 mm/s, post-test speed 5 mm/s, trigger force 5 g, 50% strain. Force-time data were used to calculate hardness, springiness, cohesiveness, and chewiness (n = 3).
Lipid oxidation
Lipid oxidation was evaluated by thiobarbituric acid reactive substances (TBARS) assay adapted from Kerth & Rowe (2016). Briefly, 5 g homogenized patty was mixed with 15 mL 7.5% TCA, homogenized at 14,000 rpm for 1 min, and centrifuged at 15,000×g for 15 min at 4 °C. Two mL of supernatant was mixed with 2 mL of 0.02 mol/L TBA, heated in a boiling water bath for 30 min, cooled, and the absorbance at 532 nm was measured. TBARS values were calculated from an MDA standard curve and expressed as mg MDA/kg patty.
Protein oxidation
Protein oxidation was evaluated by carbonyl content measurement using a modified protocol (Rysman et al., 2016). Briefly, 3.0 g homogenized patty was dispersed in 30 mL 0.6 mol/L NaCl in 20 mmol/L phosphate buffer (pH 6.5) via homogenization (10,000 rpm, 30 s). Proteins were precipitated with 10% TCA, centrifuged (12,000×g, 10 min), and washed twice with ethanol: ethyl acetate (1:1). Carbonyl groups were derivatized with 20 mmol/L DNPH in 2 mol/L HCl (1 h, dark, 25℃), precipitated with 20% TCA, and washed thrice. Pellets were dissolved in 6 mol/L guanidine hydrochloride (pH 6.5, 30 min). Protein concentration (280 nm) and carbonyl absorbance (370 nm) were measured, with results expressed as nmol carbonyl/mg protein (ε = 22,000 L/mol·cm).
Water distribution
Water distribution in pork patties was analyzed using a low-field nuclear magnetic resonance (LF-NMR) spectrometer (MesoMR23-60 H-I, Suzhou Niumag Analytical Instruments, China) operating at 23 MHz. Fresh and freeze-thawed patties were cut into cylindrical samples (diameter 20 mm, thickness 10 mm) and placed into 10-mm-diameter glass tubes for measurement. The transverse relaxation time (T₂) was determined via the Carr-Purcell-Meiboom-Gill (CPMG) pulse sequence with parameters including 8000 echoes, an echo time of 0.5 ms, a repetition time of 2000 ms, and 16 scans per sample. T₂ spectra were decomposed into three components using Chenomx NMR Suite 9.4 software: bound water (T2b, 0–10 ms) tightly associated with macromolecules, immobilized water (T21, 10 − 100 ms) trapped within the protein-lipid network, and free water (T22, > 100 ms) with high mobility. The relative peak areas of these components were normalized to total water content and expressed as percentages to assess water state transitions.
Statistical analysis
To evaluate the FT stability of WPI-CMC PEs and their impact on the physicochemical properties of frozen pork patties as fat substitute, three independent batches of samples were prepared. All measurements (including emulsion characterization and patty quality analyses) were repeated in triplicate, and results are reported as mean values ± standard error (SE). A completely randomized design was used, with treatment groups, FT cycles, and their interactions defined as fixed effects, while batch replications were treated as random effects. Statistical analyses were performed using the General Linear Models (GLM) procedure in Statistix 8.1 software (Analytical Software, St. Paul, MN, USA). Main effects and interactions were assessed via analysis of variance (ANOVA), and Tukey’s test was used for post-hoc comparisons to identify significant differences among means at p < 0.05.
Results and discussion
Freeze-thaw stability of Pickering emulsions
Droplet size distribution
The particle size in PEs significantly impacts its physical properties, including appearance, texture, and stability (Heidari-Dalfard et al., 2025). A narrower droplet size distribution and smaller mean droplet size generally correlate with enhanced emulsion stability (Zhang et al., 2021). Figure 1 illustrates the changes in droplet size distribution of the PE stabilized by WPI/CMC complex nanoparticles with different CMC concentrations, both (A) and after (B) FT cycles. For the initial emulsions, as the CMC concentration increased from 0% to 0.8%, the mean droplet size initially decreased significantly (p < 0.05), from 426.8 ± 18.5 nm at 0% CMC to 352.3 ± 12.7 nm at 0.2% CMC, and further declined to 301.5 ± 9.8 nm at 0.4% CMC, reaching a minimum of 278.2 ± 8.3 nm at 0.6% CMC. However, upon increasing the CMC concentration to 0.8%, the mean droplet size rose significantly (p < 0.05) to 339.7 ± 11.2 nm. The minimal droplet size observed at 0.6% CMC can be attributed to enhanced electrostatic interactions between WPI and CMC, which improved interfacial stability and promoted finer emulsification. As reported by Huang et al. (2024a, b), CMC induces the unfolding of WPI through electrostatic interactions between the positively charged WPI at acidic pH and the negatively charged carboxyl groups of CMC, as well as hydrogen bonding between the amino groups of WPI and the hydroxyl groups of CMC. This structural rearrangement leads to an increase in β-sheet content and greater exposure of hydrophobic amino acids, thereby enhancing interfacial adsorption capacity. Consistent with this observation, Huang et al. (2024a, b) demonstrated that an optimal CMC concentration of 0.5% enhances the surface hydrophobicity of WPI, primarily through inducing conformational changes and structural unfolding of the protein, thereby promoting the formation of smaller and more uniformly distributed emulsion droplets. At this optimal concentration, the droplet volume exhibited a single-peak distribution, indicating a relatively uniform droplet size. Conversely, excessive CMC causes the aggregation of CMC chains, interfering with the uniform distribution of WPI/CMC complex nanoparticles at the oil-water interface, resulting in larger droplets and a broader droplet size distribution. The increased entanglement of CMC molecules reduces the mobility of the nanoparticles, hindering their ability to efficiently stabilize the emulsion droplets (Zhou et al., 2023).
Fig. 1. Droplet size distribution of WPI-CMC Pickering emulsions prepared with different amounts of CMC before (A) and after (B) three freeze-thaw cycles
After the FT cycle treatment, significant variations in mean droplet size and distribution patterns were observed across groups as shown in Fig. 1(b), with most emulsions exhibiting a multimodal distribution pattern, consistent with the general trend in other polysaccharide-protein stabilized emulsions (Wu et al., 2025). Specifically, the control group showed a drastic increase in mean droplet size from the initial 426.8 nm to 5486.3 ± 192.7 nm, with a multimodal distribution, indicating severe aggregation and phase separation. The WPI/CMC-0.2 group had a slightly smaller mean size (5198.7 ± 178.4 nm) but retained a multimodal distribution, with limited alleviation of aggregation. The WPI/CMC-0.8 group also saw a surge in size to 4987.6 ± 156.3 nm, accompanied by a multimodal distribution due to excess CMC-induced flocculation. However, PEs formulated with CMC concentrations of 0.4% and 0.6% still predominantly displayed unimodal distributions. The WPI/CMC-0.4 group exhibited a moderate size increase to 986.5 ± 47.8 nm with a unimodal distribution, maintaining structural integrity. Among them, the 0.6% CMC emulsion retained the smallest mean size (947.3 ± 39.6 nm) with a narrow unimodal distribution, and the volume of its highest peak was lower compared to that at 0.4%. At the optimal CMC concentration, the electrostatic and hydrogen bonding interactions between WPI and CMC are well-balanced, forming a dense and stable layer of WPI/CMC complex nanoparticles at the oil-water interface. This tightly packed layer effectively separates the droplets, preventing aggregation during the FT cycles and maintaining the emulsion’s stability (Heidari-Dalfard et al., 2025).
Visual appearance
The visual appearance of the WPI and WPI/CMC-stabilized PEs in their states before FT treatment and after three FT cycles was depicted in Fig. 2. At the initial stage, all emulsions presented a uniform creamy-white color and homogeneous state, indicating the effective formation of stable emulsions. This outcome aligns with previous research, which demonstrated that appropriate emulsifier systems can maintain the stability of PEs by adsorbing at the oil-water interface (Marhamati et al., 2021; Ravera et al., 2021).
Fig. 2. Morphological appearance of WPI-CMC Pickering emulsions prepared with different amounts of CMC before and after three freeze-thaw cycles. F: Freeze-thaw cycle (F0: before cycles, F3: cycle 3)
However, after three FT cycles, significant disparities in stability emerged among the samples. The WPI emulsion experienced severe phase separation, characterized by a distinct water phase at the bottom and aggregated emulsified matter at the top. This phenomenon can be attributed to the disruption of protein structure at the oil-water interface during repeated FT processes, leading to droplet coalescence and emulsion breakdown (Zhang et al., 2022; Zhu et al., 2017). For the WPI/CMC emulsions, the WPI/CMC-0.2 and WPI/CMC-0.4 emulsions showed noticeable signs of instability, with the former exhibiting liquid separation at the top and the latter developing a thin layer of clear liquid due to stratification. Although these emulsions received some protection from CMC addition, they were still unable to fully resist the destabilizing effects of FT cycles. The WPI/CMC-0.8 emulsion also showed a weak sign of stratification, indicating a reduction in its stability compared to the pre-FT state. Notably, the WPI/CMC-0.6 emulsion stood out as the most stable among all samples after FT treatment. It maintained a relatively uniform appearance with only a negligible stratification tendency, significantly outperforming other counterparts. This superior stability can be attributed to the optimal interaction between WPI and CMC at this specific concentration, which effectively withstood the mechanical stress induced by FT cycles. The combination likely enhanced the integrity of the oil-water interface, preventing droplet aggregation and phase separation more efficiently than other formulations.
Microstructure
Figure 3 presents the microstructural images of WPI and WPI/CMC-stabilized PEs under an optical microscope, before and after three FT cycles. Initially, droplets in all samples are uniformly distributed with minimal aggregation. This indicates that the emulsions are in a relatively stable state. The stabilizing components, either WPI alone or in combination with CMC, effectively prevent the premature coalescence and flocculation of droplets. This aligns with the typical behavior of well-formed emulsions before exposure to external stressors.
Fig. 3. Microstructure changes of WPI-CMC Pickering emulsion prepared with different amounts of CMC before and after three freeze-thaw cycles (Magnification × 200). F: Freeze-thaw cycle (F0: before cycles, F3: cycle 3)
After three FT cycles, significant changes are observable. In the WPI sample, the droplets exhibit substantial coalescence and flocculation. The ice crystals formed during the freezing process have a detrimental impact on the protein film at the droplet interface, as reported in previous studies (Sun et al., 2023). These ice crystals disrupt the stability of the protein-based interface, leading to the aggregation of droplets. As flocculation progresses, the droplets cluster tightly, resulting in more pronounced agglomeration. Among the WPI/CMC samples, different degrees of droplet aggregation are evident. The WPI/CMC-0.2 and WPI/CMC-0.8 samples exhibit a significantly higher degree of droplet coalescence and flocculation compared to their state before the FT process. However, the WPI/CMC-0.4 and WPI/CMC-0.6 samples demonstrate relatively milder aggregation phenomena. The interaction between WPI and CMC may potentially cause the unfolding of WPI molecules via electrostatic or other forces, thereby exposing hydrophobic amino acids and enhancing the protein’s adsorption capability at the oil-water interface (Huang et al., 2024a, b). This would reduce the interfacial tension and contribute to better resistance against the destabilizing effects of FT cycles. In particular, the WPI/CMC-0.6 sample shows a relatively lower degree of droplet aggregation after three FT cycles, indicating that this specific ratio of WPI and CMC is more effective in maintaining the integrity of the emulsion droplets during FT processes.
Creaming index
The creaming index serves as a crucial metric for evaluating the stability of PEs stabilized by WPI and CMC. This index not only reflects the separation tendency of the emulsion phases but also provides insights into the effectiveness of the stabilizing agents. Figure 4 showcases the influence of FT cycles on the creaming index of these PEs at varying CMC concentrations. At the onset, all emulsions presented a creaming index beneath 4%, signifying fairly good initial stability within an acceptable threshold. Nevertheless, as the number of FT cycles increased, a significant rise (p < 0.05) was observed in the creaming index of all PEs. This can be mainly attributed to ice crystal formation during the freezing process. When ice crystals form, they disrupt the delicate equilibrium at the oil-water interface. The mechanical stress exerted by these ice crystals not only perturbs the interfacial film created by the WPI/CMC complex but also interferes with the electrostatic and steric interactions that are crucial for maintaining droplet stability. Consequently, local droplet coalescence and precipitation are expedited, which in turn results in the observed increase in the creaming index (Cai et al., 2023; Ravera et al., 2021).
Fig. 4. The creaming index of WPI-CMC Pickering emulsions prepared with different amounts of CMC. The means with different uppercase letters (A-D) within the same treatment indicate significant difference (p < 0.05); the means with different lowercase letters (a-d) within the same F-T cycle indicate significant difference (p < 0.05). F: Freeze-thaw cycle (F0: before cycles, F1-F3: cycle 1 to cycle 3)
As depicted in Fig. 4, the creaming index of WPI/CMC-stabilized PEs exhibited a trend of first decreasing and then increasing with the increment of CMC concentration. After three FT cycles, the creaming index was at its nadir when the CMC concentration was 0.6%. Compared to PEs prepared with 0%, 0.2%, 0.4%, and 0.8% CMC concentrations, the creaming index of the PE with 0.6% CMC decreased by 48.6%, 35.2%, 29.8%, and 23.5%, respectively. The results indicate that appropriate CMC induces conformational changes in WPI, enhancing droplet surface hydrophobicity to form stronger oil-water interfacial membranes. CMC also increases continuous phase viscosity, and after thawing, promotes flocculation to form a 3D network that prevents emulsion delamination (Hou et al., 2024; Jia & Zhang, 2024), consistent with prior protein-polysaccharide synergy studies (Babu et al., 2024). However, excess CMC forms a dense network restricting droplet mobility, while high concentrations cause bridging flocculation and droplet aggregation, reducing freeze-thaw stability and increasing creaming (Cui et al., 2021; Shehzad et al., 2024).
Physicochemical properties of frozen pork patties
Water holding capacity
The WHC of pork patties, a crucial parameter for texture and juiciness, was evaluated by thawing loss and expressible moisture measurements (Fig. 5). Before FT treatment, different fat substitution levels with WPI-CMC PEs already affected the WHC of pork patties. Compared with the control group, the fat-substituted groups exhibited lower expressible moisture, attributed to the gel-like network formed by WPI-CMC PEs that interacted with patty proteins to enhance system stability. During FT cycles, all treatment groups exhibited a consistent increase in thawing loss and expressible moisture (p < 0.05), though the extent of change varied significantly with the fat substitution level. The control group showed the most pronounced deterioration, with thawing loss increasing from 3.91% to 11.3% and expressible moisture from 3.17% to 12.6% across multiple FT cycles, reflecting a marked decline in WHC. This decline was primarily attributed to ice crystal growth during frozen storage damages muscle cell ultrastructure, causing tissue distortion and the release of intracellular water that remains trapped in the extracellular space post-thaw (Yin et al., 2025). Repeated FT cycles further exacerbate this effect by promoting ice crystal recrystallization, which weakens muscle fiber integrity, impairs capillary water retention, and enhances moisture loss (Wan et al., 2023). Concurrently, freezing-induced denaturation and aggregation of myofibrillar proteins disrupt protein-water interactions, directly reducing their water-binding capacity (Zhang et al., 2023a, b, c).
Fig. 5. Effects of Pickering emulsion as fat substitutes on thawing loss (A) and expressible moisture (B) of pork patties during freeze-thaw cycles. R1, R2, R3, R4, and R5, correspond to the addition of Pickering emulsion at levels of 0%, 25%, 50%, 75%, and 100% respectively. The means with different uppercase letters (**A–**D) within the same treatment indicate significant difference (p < 0.05); the means with different lowercase letters (a–d) within the same F-T cycle indicate significant difference (p < 0.05). F: Freeze-thaw cycle (F0: before cycles, F1-F5: cycle 1 to cycle 5)
After three and five FT cycles, pork patties with 50–75% fat substitution by WPI-CMC PEs showed significantly higher WHC than the control (p < 0.05). The WPI-CMC network acts as a physical barrier, restricting ice crystal growth by occupying matrix space and minimizing water loss during thawing. The 75% substitution group exhibited the lowest thawing loss (6.59%) and expressible moisture (8.80%) after five cycles, with no significant difference from the 50% group (p > 0.05). This is because the WPI-CMC complex in PEs forms a dense interfacial film, a phenomenon consistent with Ding et al. (2023) who found that WPI-CMC complexes form a thicker interfacial layer via hydrophobic amino acid adsorption, which resists ice crystal damage during freezing. However, 100% substitution weakened WHC due to WPI-CMC aggregate formation (Zhang et al., 2024), which disrupted the native protein network and enhanced moisture loss during FT cycles.
Color
Color is a pivotal quality attribute for meat products, profoundly influencing consumer purchase intentions (Tomasevic et al., 2021). Table 2 shows the effect of fat substitution with PEs on pork patty color stability during FT cycles. The L* values increased with higher fat substitution, likely due to smaller oil droplets in WPI-CMC PEs enhancing light reflection in the meat matrix (de Souza Paglarini et al., 2019). All groups showed decreasing L* values during FT cycles, with the steepest decline in the control group due to ice crystal damage to the protein-fat network and reduced light reflectivity (Li et al., 2022). The 75% substitution group exhibited the best retention of L* values, as the PEs enhanced water-holding capacity, effectively reducing surface moisture and minimizing light scattering during thawing. These results align with previous studies on emulsion-stabilized WHC in frozen meats (Zhang et al., 2023a, b, c) (Table 3).
Table 2. Effects of fat substitution with Pickering emulsions stabilized by WPI-CMC nanoparticles on the color stability of pork patties during freeze-thaw cyclesColorTreatmentsFreeze-thaw cycles0135 L* R153.7 ± 0.6^Ca^53.1 ± 0.5^Cab^51.9 ± 0.8^Db^50.6 ± 0.7^Dc^R254.5 ± 0.5^Ca^53.8 ± 0.6^BCab^52.7 ± 0.7^CDb^51.4 ± 0.7^CDc^R355.5 ± 0.4^Ba^54.7 ± 0.5^Ba^53.6 ± 0.6^BCb^52.3 ± 0.6^BCc^R456.4 ± 0.4^Aa^55.8 ± 0.6^Aa^54.5 ± 0.5^ABb^54.1 ± 0.5^Ab^R557.1 ± 0.4^Aa^56.5 ± 0.4^Aa^55.2 ± 0.5^Ab^53.3 ± 0.6^ABc^ a* R115.8 ± 0.5^Aa^14.8 ± 0.2^BCb^13.6 ± 0.3^Bc^12.5 ± 0.3^Dd^R215.6 ± 0.4^Aa^14.7 ± 0.4^Cb^14.4 ± 0.5^ABb^13.6 ± 0.3^Cc^R315.5 ± 0.5^Aa^15.3 ± 0.4^ABab^14.6 ± 0.5^Abc^14.5 ± 0.3^Bc^R415.7 ± 0.3^Aa^15.6 ± 0.4^Aa^15.2 ± 0.7^Aa^15.1 ± 0.3^Aa^R515.5 ± 0.3^Aa^14.8 ± 0.4^BCb^14.4 ± 0.4^ABb^13.3 ± 0.3^Cc^ b* R18.1 ± 0.3^Cc^8.9 ± 0.4^Cc^10.2 ± 0.4^BCb^11.5 ± 0.6^ABa^R28.3 ± 0.3^Cc^9.1 ± 0.4^BCbc^9.6 ± 0.3^Cb^10.6 ± 0.5^BCa^R39.2 ± 0.3^Ba^9.6 ± 0.3^BCa^9.7 ± 0.4^BCa^9.9 ± 0.3^Ca^R410.0 ± 0.4^Aa^10.1 ± 0.4^Ba^10.6 ± 0.5^ABa^10.6 ± 0.4^BCa^R510.3 ± 0.4^Ab^11.1 ± 0.5^Aab^11.2 ± 0.4^Aa^11.9 ± 0.6^Aa^Note R1, R2, R3, R4, and R5 correspond to the replacement of pork back fat with Pickering emulsion at levels of 0%, 25%, 50%, 75%, and 100% respectively. WPI, whey protein isolate; CMC, carboxymethyl cellulose. Means ± SE marked with different uppercase letters (A-D) in a column or with different lowercase letters (a-d) in a row signify significant differences (p < 0.05)
Table 3. Effects of fat substitution with Pickering emulsions stabilized by WPI-CMC nanoparticles on the texture properties of pork patties during freeze-thaw cyclesTexture propertiesTreatmentsFreeze-thaw cycles0135Hardness (N)R128.4 ± 0.5^Aa^29.7 ± 0.9^Aa^25.2 ± 0.7^Bb^18.9 ± 0.8^Cc^R223.1 ± 0.5^Cb^25.6 ± 0.5^Ca^22.8 ± 0.6^Cb^20.2 ± 0.5^Cc^R324.5 ± 1.2^Bb^26.9 ± 0.7^Ba^24.8 ± 0.6^Bb^22.7 ± 0.3^Bc^R427.4 ± 0.9^Aab^28.9 ± 0.8^Aa^27.1 ± 0.3^Abc^25.6 ± 1.2^Ac^R528.5 ± 0.3^Ab^29.9 ± 0.4^Aa^22.6 ± 0.7^Cc^19.6 ± 0.4^Cd^SpringinessR10.80 ± 0.02^Aab^0.82 ± 0.01^Aa^0.79 ± 0.02^Bb^0.75 ± 0.01^Bc^R20.80 ± 0.02^Ab^0.83 ± 0.02^Aa^0.81 ± 0.02^ABab^0.79 ± 0.02^Ab^R30.78 ± 0.02^Ab^0.82 ± 0.01^Aa^0.81 ± 0.01^Aa^0.80 ± 0.01^Aab^R40.80 ± 0.02^Ab^0.83 ± 0.01^Aa^0.81 ± 0.01^Aab^0.81 ± 0.02^Aab^R50.80 ± 0.02^Aab^0.83 ± 0.02^Aa^0.80 ± 0.01^ABb^0.76 ± 0.01^Bc^CohesivenessR10.65 ± 0.02^Aab^0.67 ± 0.03^Aa^0.61 ± 0.03^Ab^0.46 ± 0.03^Cc^R20.60 ± 0.01^Ba^0.60 ± 0.03^Ba^0.59 ± 0.03^Aa^0.53 ± 0.01^Bb^R30.60 ± 0.02^Ba^0.62 ± 0.02^Ba^0.59 ± 0.03^Aa^0.54 ± 0.03^ABb^R40.63 ± 0.02^Aa^0.62 ± 0.02^Bab^0.60 ± 0.01^Abc^0.57 ± 0.02^Ac^R50.53 ± 0.03^Ca^0.54 ± 0.03^Ca^0.50 ± 0.01^Bab^0.47 ± 0.02^Cb^Chewiness (N)R117.8 ± 0.4^Aa^18.1 ± 0.4^Aa^14.4 ± 0.3^BCb^10.7 ± 0.7^Cc^R216.0 ± 0.4^Ba^16.7 ± 0.9^Ba^15.0 ± 0.2^ABb^12.2 ± 0.4^Bc^R315.4 ± 0.4^Ba^15.2 ± 0.4^Cab^15.0 ± 0.4^ABab^14.6 ± 0.3^Ab^R417.5 ± 0.6^Aa^16.3 ± 0.9^BCab^15.6 ± 0.8^Ab^15.1 ± 1.0^Ab^R515.3 ± 0.3^Ba^15.7 ± 0.7^BCa^14.1 ± 0.4^Cb^12.9 ± 0.4^Bc^Note R1, R2, R3, R4, and R5 correspond to the replacement of pork back fat with Pickering emulsion at levels of 0%, 25%, 50%, 75%, and 100% respectively. WPI, whey protein isolate; CMC, carboxymethyl cellulose. Means ± SE marked with different uppercase letters (A-C) in a column or with different lowercase letters (a-d) in a row signify significant differences (p < 0.05)
Fat substitution with PEs initially had no significant impact on the redness of pork patties, as myoglobin-mediated color remained consistent across samples. However, a* values declined in all groups with increasing FT cycles, primarily due to the oxidation of myoglobin to metmyoglobin (MetMb), which was triggered by ice crystal-induced muscle cell damage and the subsequent release of oxidases (Wang et al., 2021a, b). The control group exhibited the most pronounced decrease in a* values, whereas the fat-substituted groups showed less severe reductions. Notably, the 75% fat substitution group maintained the highest a* values throughout the FT cycles. This preservation of redness can be attributed to the ability of PEs to inhibit protein oxidation and limit ice crystal growth, thereby maintaining tissue integrity and reducing MetMb formation, aligning with the observed antioxidant effects of emulsions in pork sausages (Liu et al., 2023).
The b* values demonstrated a positive correlation with substitution levels, showing a significant increase (p < 0.05) as the emulsion proportion rose, which gradually shifted the patty color toward a more yellow hue. This finding differs from Li et al. (2023), who observed lower b* values in PE-stabilized pork sausages; these variations can likely be attributed to differences in emulsion systems, interactions among meat components, and processing techniques. During FT cycles, the control group and the 25% substitution group exhibited notable increases in b* values, aligning with color degradation caused by MetMb and lipid oxidation (Pan et al., 2021; Wang et al., 2021a). In contrast, the 50–75% substitution groups maintained stable b* values (p > 0.05), suggesting that the PE components may have played a protective role by inhibiting oxidative degradation under FT stress (Cheng et al., 2024).
Texture analysis
The texture properties of pork patties, including hardness, springiness, cohesiveness, and chewiness, were analyzed to examine the effects of PE substitution levels and FT cycles. Before FT treatment, 25% and 50% substitution groups showed lower hardness, cohesiveness, and chewiness than the control (p < 0.05), attributed to weakened lipid-mediated protein interactions from reduced fat content, softening the matrix (Zhao et al., 2023). At a 75% substitution level, the hardness was restored to levels comparable to those of the control sample (p > 0.05). Cohesiveness and chewiness exhibited a similar trend, with values approaching those observed in the control group (p > 0.05). This recovery can be attributed to the formation of a dense gel-like network by the WPI-CMC emulsion, wherein complex nanoparticles adsorb at the oil-water interface and crosslink with myofibrillar proteins, thereby reinforcing the structural matrix (Jiang et al., 2023). The emulsion network effectively mimics the key functions of the lipid matrix, leading to an overall improvement in texture. With 100% substitution, the hardness increased further, whereas cohesiveness and chewiness showed a sharp decline. This suggests that an excessive amount of emulsion disturbs the balance of the protein network. The over-crosslinking of WPI-CMC complexes leads to a stiffer matrix, which compromises its flexibility. Unlike the other textural properties, springiness showed little variation across all substitution levels, remaining consistently around 0.80. This suggests that springiness is less influenced by fat replacement and is more dependent on inherent structural elements that remain unaffected by the WPI-CMC PE within the tested range.
After one FT cycle, hardness and springiness temporarily increased, likely due to freezing-induced protein aggregation that tightened the matrix (Wang et al., 2021a, b), while cohesiveness and chewiness remained largely unchanged. However, after three to five cycles, the textural quality of all samples started to decline, with the extent of degradation differing based on the level of substitution. The substitution groups within the 50%–75% range exhibited high stability across all measured texture properties. Compared to other groups, their hardness, springiness, cohesiveness, and chewiness declined more gradually. This improved performance can be attributed to the optimized WPI-CMC network, which effectively balances interfacial stabilization and matrix viscoelasticity (Peng et al., 2024), thereby suppressing ice crystal growth during freezing and minimizing damage to protein structures. In comparison, the control group and the 100% substitution group showed faster degradation rates. The control group does not have a protective emulsion network, making it more susceptible to damage from ice crystal formation. At the same time, the excessive amount of emulsion in the 100% substitution group interferes with normal protein interactions, thereby reducing its resistance to FT stress.
Lipid oxidation
Lipid oxidation significantly impacts the visual and sensory quality of pork patties. As shown in Fig. 6A, TBARS measurements during FT cycles revealed that freshly prepared PE-substituted patties had lower TBARS values than the control (p < 0.05). TBARS content increased in all groups with increasing FT cycles, but PE-substituted samples consistently showed lower levels than the control (p < 0.05). The control had the highest malondialdehyde (MDA) content, while higher PE substitution suppressed MDA formation despite the oxidation-prone unsaturated fatty acids in PE-stabilized sunflower oil. This oxidative stability stems from the combined effects of reduced lipid content and multiple PE-mediated antioxidant mechanisms. The control’s high MDA levels reflect unprotected animal fat oxidation, whereas PEs inhibit oxidation through multiple mechanisms, including forming a steric barrier at oil-water interfaces to shield lipids from oxygen and pro-oxidants (Cai et al., 2023), increasing emulsion viscosity to limit free radical and metal ion mobility (Zhu et al., 2020a, b), and reducing total lipid content as oxidation substrates. These findings align with studies showing PEs outperform traditional emulsifiers in oxidative stability (Li et al., 2023; Rezaee & Aider, 2023), confirming that PE-stabilized sunflower oil effectively mitigates lipid oxidation in pork patties during FT cycles by integrating physical protection, kinetic inhibition, and substrate reduction.
Fig. 6. Effects of Pickering emulsion as fat substitutes on lipid oxidation (A) and protein oxidation (B) of pork patties during freeze-thaw cycles. R1, R2, R3, R4, and R5, correspond to the addition of Pickering emulsion at levels of 0%, 25%, 50%, 75%, and 100% respectively. The means with different uppercase letters (A**–D) within the same treatment indicate significant difference (p < 0.05); the means with different lowercase letters (a–**c) within the same F-T cycle indicate significant difference (p < 0.05). F: Freeze-thaw cycle (F0: before cycles, F1-F5: cycle 1 to cycle 5)
Protein oxidation
The carbonyl content serves as a key indicator for assessing the oxidative status of pork patties (Estévez, 2011). As illustrated in Fig. 6B, the carbonyl content across all samples increased significantly (p < 0.05) with the number of FT cycles. The control group exhibited the most pronounced increase, reaching 2.97 nmol/mg protein after five FT cycles. In comparison, groups in which animal fat was partially replaced with PEs showed a marked reduction in carbonyl formation during the FT process. Moreover, higher fat replacement ratios were associated with reduced carbonylation levels (p < 0.05). The group with the highest replacement ratio consistently exhibited relatively lower carbonyl content at each FT stage. Repeated FT cycles enhanced protein oxidation by damaging muscle structure and exposing proteins to oxidative stress (Leygonie et al., 2012). In the control group, protein oxidation resulted in a noticeable increase in carbonyl content. In contrast, the groups with PE substitution exhibited significantly reduced carbonylation, a protective effect attributable to two complementary mechanisms. First, WPI-CMC nanoparticles form a stable barrier at the oil-water interface, effectively shielding proteins from oxidative damage (Cai et al., 2023). Second, WPI neutralizes harmful free radicals (Corrochano et al., 2018), while CMC chelates pro-oxidative metal ions, thereby inhibiting oxidation (Tran et al., 2017). With higher levels of PE substitution, the concentration of these protective elements increases, leading to even greater reduction in protein oxidation. As a result, incorporating WPI-CMC PEs offers effective protection against protein oxidation during FT cycles, supporting the preservation of meat quality.
Water distribution
Low-field nuclear magnetic resonance (LF-NMR) was used to investigate the distribution and mobility of water in pork patties. In LF-NMR analysis, the transverse relaxation times T2b (0–10 ms), T21 (100–500 ms), and T22 (1000–5000 ms) correspond to bound water tightly associated with macromolecules, immobilized water within the protein network, and free water, respectively. As shown in Fig. 7A, repeated FT cycles increased T2b and T21 relaxation times, indicating enhanced water mobility, attributed to freezing-induced ice crystal growth damaged cellular structures, disrupted protein networks, and reduced water-binding capacity via non-covalent interactions (Pan et al., 2021). Prolonged T22 suggested widened muscle fiber spacing, facilitating water migration out of the matrix (Wang et al., 2021a, b). During the first FT cycle, T21 values showed no significant differences among substitution groups (p > 0.05). However, by the third and fifth cycles, 50%–75% substitution groups exhibited significant leftward shifts in T2b and T21 compared to the control (p < 0.05), indicating reduced water mobility. PEs restricted ice crystal growth and protein denaturation, thereby maintaining muscle structure and inhibiting water migration (Zhang et al., 2023a, b, c).
Fig. 7. Effects of Pickering emulsion as fat substitutes on the T2 relaxation times (A) and the P2 (B) of pork patties during freeze-thaw cycles. R1, R2, R3, R4, and R5, correspond to the addition of Pickering emulsion at levels of 0%, 25%, 50%, 75%, and 100% respectively. The means with different uppercase letters (A**–**D) within the same treatment indicate significant difference (p < 0.05); the means with different lowercase letters (a–d) within the same F-T cycle indicate significant difference (p < 0.05). F: Freeze-thaw cycle (F0: before cycles, F1-F5: cycle 1 to cycle 5)
The relative water proportions showed stable P2b across groups (Fig. 7B), consistent with prior findings that bound water is insensitive to structural changes (Kang et al., 2017). After five FT cycles, the P21 level in the control group decreased significantly, while the P22 level showed a notable increase (p < 0.05). During multiple FT cycles, the rate of moisture movement surpassed the formation rate of extracellular ice crystals, resulting in moisture redistribution. Moreover, the denaturation of myofibrillar proteins caused by FT cycles led to a reduction in the protein’s water-binding capacity (Wang et al., 2021a, b). However, samples with fat substitution exhibited a substantial reduction in the amount of free water. Conversely, the percentage of immobilized water increased in these treated samples. These findings demonstrated that fat substitution with high FT stability PEs effectively restricted water migration. After five FT cycles, the P₂₁ level decreased by 5.15% in the control group, compared to reductions of 3.59% in the 25% group, 2.48% in the 50% group, 2.29% in the 75% group, and 3.50% in the 100% fat substitution group. The 100% substitution group showed a larger P21 decline, likely due to excessive PEs disrupting protein network balance and weakening water retention (Zhang et al., 2023a, b, c). These results align with WHC data (Fig. 5), demonstrating that optimal PE substitution stabilizes water distribution by inhibiting ice crystal damage during FT cycles.
Conclusion
This study demonstrates that WPI-CMC PEs serve as effective functional fat substitutes in pork patties, enhancing FT stability through synergistic protective mechanisms. The optimal PE formulation (0.6% CMC) formed a dense interfacial film via WPI-CMC interactions, inhibiting ice crystal damage and droplet coalescence during FT cycles. Substituting 50%–75% of pork fat with PEs improved water distribution, reduced lipid and protein oxidation, and maintained textural integrity. The 75% substitution group exhibited a favorable balance, as its gel-like network effectively immobilized water within the protein matrix, while the WPI-CMC interfacial layer served as a physical barrier against oxygen and pro-oxidants. Excessive substitution (100%) disrupted myofibrillar networks via gel matrix over-crosslinking, highlighting the need for emulsion concentration optimization. Practically, 75% PE substitution offers a viable strategy for low-fat, FT-stable meat products, aligning with consumer demands for healthier, shelf-stable options. Despite these advances, several limitations remain. FT stability was evaluated under controlled laboratory conditions, which may not fully capture the variable temperature fluctuations and extended storage durations typical of real-world supply chains. Moreover, the current findings are specific to pork patties, as the compatibility of PEs with other animal-based meats or plant-based analogs has not been explored. Additionally, sensory properties and consumer acceptability over prolonged storage periods require further investigation. Future research should focus on optimizing PE formulations for diverse freezing environments, examining long-term flavor and nutrient retention, and expanding applications to other meat and plant-based food systems.
The reference list from the paper itself. Each links out to its DOI / PubMed record.