Reconstruction of fat crystal organization by a phytosterol/γ-oryzanol bigel: Linking structural evolution in blended fat systems to dough architecture and cookie quality Authors: Yuanyuan Peng, Ke Zhou, Lijuan Han, Yan Yang, Junbo He, Chen Chen, Pei Liu, Jiangtao Feng Published: 2026 ABSTRACT AbstractThe crystal organization of plastic fats plays an important role in determining the structural and functional properties of bakery systems. Commercial vegetable butter (VB) is widely used in bakery applications because its crystalline network contributes to dough processing and baked product quality. Although bigels have emerged as promising structured fats with reduced saturated fat content, how bigel incorporation changes the crystal organization of an existing commercial fat matrix and how these changes are reflected in dough and cookie properties remain poorly understood. In this study, butter cookies were used as a model bakery system to investigate changes in the crystal organization of VB after incorporation of a phytosterol/γ-oryzanol-sodium alginate bigel and their association with dough architecture and cookie quality. Crystal morphology, polymorphism, solid fat content (SFC), and thermal behavior of the blended fat systems were characterized, followed by analyses of dough microstructure, rheological properties, texture, water distribution, and cookie quality. Bigel incorporation progressively reduced crystal aggregation and SFC while altering the polymorphic characteristics and thermal behavior of the blended fat systems. Both β- and β′-type structures remained detectable in the BG-VB blends. Dough viscoelasticity and hardness decreased with increasing BG content, whereas water mobility was reduced in association with the hydrogel phase. At the product level, cookie spread ratio decreased from 4.31 to 3.65-3.90 for BG-containing cookies, breaking force increased from 25.63 to 26-30.97 N for BG-containing cookies, and baking loss remained unchanged. Oil migration remained below 1% after complete replacement of VB, while acceptable sensory quality was maintained. The results show that changes in fat crystal organization, SFC and water distribution were associated with changes in dough architecture and cookie properties in the phytosterol/γ-oryzanol-sodium alginate bigel/VB system. These findings provide insights into the relationship between structured fat organization, dough properties and cookie quality, and may support the development of alternative fat systems for bakery applications.Keywords: Bigel, Fat crystal organization, Dough architecture, Cookie quality Introduction Plastic fats play a fundamental role in determining the structure and functionality of bakery products because their crystal organization governs dough processing, gas retention and final product quality (Jacob and Leelavathi, 2007; Pareyt et al., 2009). Commercial vegetable butter (VB) is one of the most widely used plastic fats in industrial baking owing to its desirable crystallization behavior and processing performance. Its functionality largely depends on the formation of a continuous fat crystal network generated by crystalline saturated lipids. Excessive consumption of saturated and trans fats has been associated with an increased risk of cardiovascular disease and other metabolic disorders (Karabulut et al., 2003; Wali et al., 2020). Replacing VB with liquid vegetable oils could improve the nutritional profile of bakery products, but liquid oils cannot provide the continuous crystal network required for dough development and usually lead to excessive spreading, poor shape retention, and undesirable texture during baking (Jacob and Leelavathi, 2007; Mert and Demirkesen, 2016a). Developing healthier fat systems while preserving fat crystal organization therefore remains an important challenge in bakery products. Oleogels have emerged as promising alternatives to conventional solid fats because they convert liquid oils into semi-solid materials without substantially increasing saturated fat content (Singh et al., 2017; Pehlivanoğlu et al., 2018). Among the available oleogelators, phytosterol (Phy) and γ-oryzanol (Ory) can self-assemble into fibrillar tubular structures that effectively immobilize liquid oils while maintaining a high proportion of unsaturated fatty acids (Chen et al., 2017). Although oleogels can establish continuous crystal networks through self-assembled phytosterol-oryzanol tubules, their limited water-binding capacity restricts their application in bakery products where both lipid structuring and water management are essential. Bigels integrate an oleogel phase with a hydrogel phase and therefore combine oil-binding ability with water-holding capacity (Pérez-Monterroza et al., 2022; Francavilla et al., 2023). Recent studies have shown that bigels can be used as fat replacers in bread, spreads, cream analogues, and other food products while maintaining desirable physicochemical properties (Han et al., 2024; Jiang et al., 2024). Several studies have evaluated the application of bigels in cookies. Quilaqueo et al. (2022) reported that beeswax-based bigels produced cookies with hardness comparable to conventional shortening, although the baked products exhibited increased thickness. Nutter et al. (2023) demonstrated that plant-based bigels effectively replaced commercial solid fats in short dough products while maintaining acceptable dough architecture and baking quality. These studies mainly focused on the technological feasibility of substituting conventional fats with bigels. However, the structural implications of adding a structurally different bigel phase into an existing commercial fat matrix are less known. More specifically, the replacement of conventional fat is not necessarily a simple reduction in fat or solid fat content. The addition of a bigel can also change the organization of the pre-existing crystal aggregates, crystalline fraction, polymorphic characteristics, and thermal behavior of the original fat phase. Here, reconstruction of fat crystal organization refers to the observed changes in crystal morphology, SFC, polymorphic characteristics, and thermal behavior following BG incorporation into VB, rather than direct molecular-level reconstruction of individual fat crystals. How these changes in fat crystal organization influence lipid continuity, water distribution, and dough architecture, and how these changes are finally reflected in cookie quality, remain insufficiently understood. In this study, a Phy-Ory-sodium alginate bigel was incorporated into VB at different substitution levels for butter cookie production. The crystal morphology, crystal polymorphism, SFC, and thermal behavior of the blended fat systems were systematically characterized to investigate how BG incorporation altered the organization and characteristics of the crystalline phase and the structural functionality of the fat system. Dough architecture, rheological properties, texture, and water distribution were subsequently analyzed, followed by the evaluation of cookie microstructure and quality attributes. By integrating these structural and functional changes across the fat system, dough and baked product, this study aimed to establish how bigel-induced reconstruction of fat crystal organization propagates across structural levels and contributes to changes in dough architecture and cookie quality. Materials and methods Materials Food-grade canola oil was obtained from Yihai Kerry Group (Shanghai, China) and used without further purification. Phytosterols (Phy) and γ-oryzanol (Ory) (>99% purity) were purchased from Xi'an Bright Biotechnology Co., Ltd. (Xi'an, China). Food-grade sodium alginate (SA) was supplied by Bidepharm Pharmaceutical Technology Co., Ltd. (Shanghai, China). Commercial vegetable butter (VB) used as the reference fat was obtained from Shandong Beiyuan Grain and Oil Food Co., Ltd. (Shandong, China). According to the manufacturer, VB contained refined vegetable oil, water, mono- and diglycerides of fatty acids, sorbitan tristearate, butylated hydroxytoluene (BHT), butylated hydroxyanisole (BHA), food flavoring, potassium sorbate, and β-carotene. Commercial low-gluten wheat flour (protein content: 8.2%) and white granulated sugar were purchased from a local supermarket (Wuhan, Hubei, China). Preparation of bigel-commercia vegetable butter blended fat systems The Phy-Ory bigel (BG) was prepared according to our previous method with slight modifications (Han et al., 2026). Briefly, Phy and Ory (2:3, w/w; total concentration 8 wt%) were dissolved in canola oil to prepare the oleogel (OG) phase, while SA (0.1 wt%) hydrogel (HG) was used as the aqueous phase. The OG and HG phases were subsequently mixed at an OG mass ratio of 80:20 to obtain the BG. For the preparation of BG-VB blends, VB was melted at 60 °C and equilibrated at 35 °C prior to use. BG and VB were mixed at mass ratios of 1:3, 1:1, and 3:1, followed by homogenization at 15,000 rpm for 1 min. The resulting samples were designated as BG1-VB3, BG2-VB2, and BG3-VB1, respectively. All samples were stored at 5 °C for at least 24 h before further analysis. VB, BG-VB blends, and BG were used as blended fat systems for subsequent analyses and cookie preparation. Characterization of blended fat systems Polarized light microscopy (PLM) The crystal morphology of VB, BG, and BG-VB blends was observed using a polarized light microscope (ZEISS Stemi 305, Germany) equipped with a digital camera. Samples were placed on a preheated glass slide, covered with a coverslip. Images were captured after crystallization at 5 °C for 24 h. X-ray diffraction (XRD) The crystalline structures of VB, BG, and BG-VB blends were characterized using an X-ray diffractometer (Empyrean, Malvern Panalytical, Almelo, Netherlands) with Cu Kα radiation (λ = 1.5406 Å, 45 kV, 40 mA). Diffraction patterns were recorded over a 2θ range of 5°-50° at a scanning rate of 5°/min and analyzed using JADE 6.0 software (Materials Data, CA, USA). Solid fat content (SFC) The SFC of samples was determined using a low-resolution pulsed nuclear magnetic resonance spectrometer (Minispec mq20, Bruker, USA). Prior to analysis, samples were heated at 80 °C for 30 min to erase crystal memory and subsequently equilibrated at 0 °C for 90 min. SFC values were then measured at temperatures ranging from 0 to 50 °C with 5 °C intervals after equilibration for 30 min at each temperature. Differential scanning calorimetry (DSC) The thermal properties of samples were analyzed using the DSC Q2000 (TA Instruments, USA). Approximately 5-7 mg of sample was sealed in a Tzero aluminum pan, equilibrated at 5 °C for 30 s, and heated to 90 °C at a rate of 5 °C/min under a nitrogen atmosphere. Preparation and characterization of dough architecture Dough preparation VB, BG-VB blends, and BG were equilibrated at 25 °C prior to use. For each formulation, 30 g of the corresponding fat system was first creamed with 22.8 g powdered sugar using an electric mixer until a homogeneous cream was obtained. Subsequently, 43.6 g sifted low-gluten wheat flour was incorporated, followed by the addition of 3.6 g externally added water. The mixture was gently mixed until a uniform dough was formed. The BG-containing formulations received additional water from the HG phase of BG, and the amount of externally added water was kept constant at 3.6 g for all formulations. Therefore, the total water contents were 3.6, 5.1, 6.6, 8.1 and 9.6 g for VB, BG1-VB3, BG2-VB2, BG3-VB1 and BG, respectively. The formulation of each dough and the corresponding water contents is summarized in Table S1. Cryo-scanning electron microscopy (Cryo-SEM) The microstructure of dough samples was observed using a Cryo-SEM (Quanta 450, FEI, USA). Dough samples were cut into 5 × 5 × 5 mm cubes, rapidly frozen in liquid nitrogen and fractured. The fractured samples were transferred to a cryo-preparation chamber, sublimated at −90 °C for 10 min to remove surface ice, and sputter-coated with gold for 60 s. Images were acquired at 5 kV. Rheological properties The viscoelastic properties of doughs were measured using a Kinexus Pro + rheometer (Malvern Instruments, UK) equipped with a 40 mm parallel plate geometry and a gap of 1 mm. A strain of 0.05% was selected as a small-amplitude deformation based on the conditions reported for dough rheological measurements by Sun et al. (2021). Frequency sweep measurements were performed at 25 °C over a frequency range of 0.1-10 Hz at a constant strain of 0.05%. Textural analysis Dough hardness was determined using a texture analyzer (TMS-Pro, FTC, USA) equipped with a P/2 cylindrical probe. Dough samples were molded into cylindrical shapes (50 mm diameter × 40 mm height) and compressed to 50% strain. The pre-test, test, and post-test speeds were set at 1.0, 0.5, and 1.0 mm/s, respectively, with a trigger force of 0.05 N. Hardness was defined as the maximum compression force. Five replicates were performed for each sample. Low-field nuclear magnetic resonance (LF-NMR) Water distribution in dough samples was analyzed using a low-field nuclear magnetic resonance spectrometer (NMI20-040 V-I, Niumag Corporation, China). Approximately 10 g of dough was placed in a 40 mm NMR tube and analyzed using the Carr-Purcell-Meiboom-Gill (CPMG) pulse sequence (Lu et al., 2021). The measurement parameters were set as follows: echo time (TE), 0.25 ms; waiting time (TW), 1500 ms; number of echoes, 3000; and number of scans, 4. T2 distributions were calculated using the instrument software. All measurements were performed in triplicate. Preparation of cookies and evaluation of structural and quality attributes Cookie preparation The prepared doughs were rested at 5 °C for 20 min prior to shaping. Dough sheets were rolled to a uniform thickness of 6 mm, allowed to rest for 5 min, and cut using a circular cutter with a diameter of 58 mm. The cookies were baked in a preheated oven at 160 °C for 20 min. After baking, the cookies were cooled at room temperature for 30 min and stored in sealed polyethylene bags until further analysis. Scanning electron microscopy (SEM) The microstructure of cookies was observed using a scanning electron microscope (SEM, Quanta 450, FEI, USA). Cookies were manually fractured to expose a natural cross-section, mounted on aluminum stubs, sputter-coated with gold, and observed at an accelerating voltage of 5 kV. The fracture surface was selected for observation to evaluate the internal porous structure of cookies. Physical and baking properties The baking loss of cookies was determined according to Liu et al. (2024) with slight modifications. The weight of cookies before (M1) and after baking (M2) was recorded. The baking loss was calculated using Eq. (1). The color of cookies was measured using a colorimeter (HunterLab UltraScan VIS, Shanghai Xinlianchuang Electronic Co, Ltd). Prior to measurement, the instrument was calibrated using a standard white calibration plate. The color parameters L*, a*, and b* were recorded at three random locations on the cookie surface, and the average values were reported. The spread ratio was determined according to AACC Method 10-52. Cookie diameter (D) was measured in four different orientations by aligning and rotating two cookies by 90° between measurements. Cookie thickness (H) was obtained by stacking two cookies in different orders and measuring the total height four times. The spread ratio was calculated as D/H. Texture analysis Cookie breaking force was measured using a texture analyzer (TMS-Pro, FTC, USA) equipped with a three-point bending rig (Yang et al., 2022). The distance between the two supporting points was fixed at 40 mm, and compression was applied at a speed of 120 mm/min until fracture occurred. Five replicates were performed at room temperature for each sample. Oil migration Oil migration was determined according to the method of Sylwia Onacik-Gür and Bikowska (2020). Five layers of filter paper were placed in a Petri dish and weighed (M0). Cookies with an initial weight of M1 were placed on the filter paper and stored at 25 °C for 24 h. The final weight of the Petri dish and filter paper was recorded as M2. Oil migration was calculated using Eq. (2). Sensory evaluation A consumer sensory evaluation was conducted to assess the acceptability of the cookies following a previously reported method with slight modifications (Aksu and Seim, 2025). Sixty untrained consumers (30 males and 30 females, aged 18-45 years) from the School of Food Science and Engineering, Wuhan Polytechnic University, participated in the study. All participants provided written informed consent before participation. Cookies were prepared one day before testing, packaged in sealed bags, and stored at room temperature. During evaluation, samples were presented in a randomized order and identified using randomly generated three-digit codes. Assessors were asked to evaluate appearance, color, flavor, texture, and overall acceptability using a five-point hedonic scale (1 = dislike extremely, 5 = like extremely). Drinking water was provided for palate cleansing between samples to minimize carry-over effects. The sensory attributes and corresponding scoring criteria used in this study are summarized in Table 1. Table 1Sensory attributes and scoring criteria used for cookie evaluation.Table 1AttributeDescriptionScaleScore definitionAppearanceVisual appearance, shape regularity and overall attractiveness of cookies1-51 = extremely poor, 5 = excellentFlavorOverall aroma and flavor quality of cookies1-51 = extremely poor, 5 = excellentColorUniformity and attractiveness of cookie surface color1-51 = extremely poor, 5 = excellentMouth feelPerception of texture and eating quality during mastication1-51 = extremely poor, 5 = excellentOverall acceptabilityOverall consumer preference and acceptance1-51 = extremely poor, 5 = excellent Sensory attributes and scoring criteria used for cookie evaluation. Statistical analysis All tests were performed in triplicate or more, and results are expressed as mean ± standard deviation (SD). The data were analyzed by one-way analysis of variance (ANOVA) using SPSS software (IBM SPSS statistic 19) and significant differences were obtained by Duncan multiple comparisons. The level of significance was set at P < 0.05. Normality and homogeneity of variance were verified before analysis. Results and discussion Reconstruction of fat crystal organization in BG-VB blended fat systems Crystal morphology The crystal morphology of VB, BG and BG-VB blends was observed by PLM, and the micrographs are shown in Fig. 1. VB exhibited densely packed spherulitic crystal clusters composed of numerous small crystals, forming a continuous crystalline network. In contrast, BG showed a relatively sparse network with elongated birefringent structures associated with the self-assembled Phy-Ory network (Tang et al., 2021). These structures are indicated by the red arrows in Fig. 1. Polarized light microscopy micrographs of commercial vegetable butter (VB), BG1-VB3, BG2-VB2, BG3-VB1, and phytosterol/γ-oryzanol-sodium alginate bigel (BG). Blue arrows indicate representative packed spherulitic crystal clusters in VB, while red arrows indicate representative elongated structures associated with the self-assembled Phy-Ory network in BG. The crystal morphology changed gradually after BG was incorporated into VB. BG1-VB3 and BG2-VB2 showed less compact spherulitic aggregates than VB, together with elongated birefringent structures that became more apparent with increasing BG content. In BG3-VB1, the spherulitic aggregates were less prominent, while the elongated structures were more widely distributed. Increasing BG incorporation progressively altered the spatial organization of the original VB crystal network, with densely aggregated spherulitic structures becoming less prominent and the elongated structures associated with the Phy-Ory network becoming more evident. Crystal polymorphism The crystal polymorphism was characterized by XRD, and the diffraction patterns are presented in Fig. 2. VB showed diffraction features at 4.52, 4.33, 4.20 and 3.86 Å, corresponding to the coexistence of β and β′ polymorphs. The β′ form is commonly associated with plastic fat systems because of its fine crystal structure and desirable processing properties (Khiabani et al., 2020). X-ray diffraction patterns of VB, BG1-VB3, BG2-VB2, BG3-VB1, and BG. BG exhibited a broad diffraction peak at 4.53 Å, indicating that β-type structures were predominant. No distinct diffraction feature was observed at approximately 3.83 Å for BG. In contrast, the BG-VB blends showed diffraction features at approximately 4.56, 4.35, 4.19 and 3.83-3.85 Å, corresponding to both β- and β′-type structures. The diffraction intensity decreased progressively with increasing BG content, which was consistent with the PLM observations showing less prominent crystal aggregates. The PLM and XRD results indicate that BG incorporation changed both the spatial organization and polymorphic characteristics of the fat crystal phase in the blended systems. The β- and β′-type structures remained detectable in the BG-VB blends, while the diffraction features associated with the original VB crystalline phase became less pronounced as the BG content increased. The decrease in diffraction intensity was accompanied by a decrease in SFC with increasing BG incorporation, indicating a lower proportion of ordered crystalline structures. Solid fat content The SFC profiles of VB, BG and BG-VB blends are shown in Fig. 3A. The SFC of all samples decreased continuously with increasing temperature because of crystal melting. VB maintained the highest SFC throughout the measured temperature range, whereas BG exhibited the lowest values owing to its high proportion of liquid oil. Solid fat content profiles (A) and differential scanning calorimetry thermograms (B) of VB, BG1-VB3, BG2-VB2, BG3-VB1, and BG. Bigel incorporation progressively reduced the SFC of VB. The SFC decreased as the substitution level increased from BG1-VB3 to BG3-VB1. This quantitative decrease complements the morphological and structural changes identified by PLM and XRD. The less aggregated crystal morphology and lower diffraction intensity were accompanied by a lower proportion of solid material over the measured temperature range. The decrease in SFC was consistent with the reduced contribution of the original VB crystalline network and the increasing proportion of the less crystalline BG phase. Thermal behavior The melting behavior of the fat systems was evaluated by DSC, and the thermograms are shown in Fig. 3B. VB exhibited a distinct endothermic melting feature at approximately 44-45 °C, corresponding to the melting of the crystalline fat phase. BG showed a broad and weak melting feature at approximately 55-57 °C, which is associated with the melting of the Phy-Ory crystal network and is consistent with previous reports on Phy-Ory oleogels. The BG-VB blends showed broad and overlapping melting profiles that differed from those of the individual VB and BG samples. A clear independent melting peak at approximately 55-57 °C, as observed for BG, was not resolved in the blends. Instead, the melting feature associated with the original VB phase tended to occur at lower temperatures as the BG proportion increased, while the high-temperature melting contribution observed for BG became less distinct in the blends. These changes indicate that BG incorporation modified the melting behavior of the blended fat systems, but the individual thermal contributions of VB and BG could not be clearly separated from the DSC profiles. The DSC results were consistent with the changes observed by PLM, XRD and SFC. Increasing BG incorporation was accompanied by changes in crystal morphology, diffraction characteristics and crystalline fraction. These changes are consistent with a progressive reorganization of the fat crystal network in the blended fat systems. The observed changes in crystal organization may have contributed to the changes in dough architecture discussed below. Changes in dough architecture associated with fat crystal organization reconstruction Cryo-SEM microstructure The Cryo-SEM images showed differences in dough architecture among the three fat systems (Fig. 4). The VB dough exhibited a relatively compact and continuous matrix with a fine and relatively uniform surface morphology. In BG2-VB2, the matrix appeared more heterogeneous, with larger discrete structural features and more visible fracture regions. The BG dough also showed a heterogeneous matrix with larger irregular structural features and more open regions than VB. These morphology differences were associated with the changes in the fat systems observed in Section 3.1 and suggest that changes in the fat system were accompanied by changes in dough matrix organization. Cryogenic scanning electron microscopy micrographs of doughs prepared with VB, BG2-VB2, and BG. Rheological properties The doughs showed G′ > G″ throughout the tested frequency range, confirming predominantly elastic behavior (Fig. 5A and B). Both moduli decreased progressively with increasing BG content, with VB showing the highest values and BG the lowest. This trend followed the decrease in SFC of the corresponding fat systems. The lower crystalline fraction may have reduced the structural contribution of the fat phase to dough viscoelasticity. Changes in lipid distribution and water mobility may also have contributed to the observed response (Han et al., 2024; Steinkellner et al., 2025). Rheological, textural and water distribution properties of doughs prepared with different VB-BG blends. Storage modulus G′ (A), loss modulus G″ (B), hardness (C), and transverse relaxation time (T2) distributions determined by low-field nuclear magnetic resonance (D). Texture Dough hardness followed the same trend as the rheological properties (Fig. 5C). VB showed the highest hardness (0.576 N), whereas BG showed the lowest value (0.272 N); the BG-VB blends showed intermediate values. The lower hardness was consistent with the lower crystalline contribution of the fat system and the changes in dough matrix morphology observed after BG incorporation (Jacob and Leelavathi, 2007). Changes in lipid distribution and the hydrogel phase may also have contributed to the reduced matrix resistance. Water distribution The LF-NMR results are presented in Fig. 5D. Three relaxation populations were observed in all samples, corresponding to bound water T21, immobilized water T22 and free water T23 (Tang et al., 2019). Compared with VB, the T22 and T23 peaks gradually shifted toward shorter relaxation times as the BG content increased. The shorter relaxation times indicate lower water mobility within the dough. This change was associated with the SA hydrogel phase, which can retain water through its three-dimensional network. Because the externally added water was kept constant at 3.6 g for all formulations, the water introduced through the hydrogel phase resulted in higher total water contents with increasing BG substitution. The total water contents increased from 3.6 g for VB to 5.1, 6.6, 8.1 and 9.6 g for BG1-VB3, BG2-VB2, BG3-VB1 and BG, respectively. The increased water availability and the water-retaining effect of the hydrogel phase may both have contributed to the changes in water mobility observed by LF-NMR. BG incorporation was also associated with changes in dough matrix morphology, as observed by Cryo-SEM. The lower water mobility did not result in stronger dough. Doughs with higher BG contents exhibited lower G′ values and hardness despite the lower water mobility, indicating that lower water mobility did not correspond to higher dough strength. The rheological and textural changes therefore cannot be attributed to water immobilization alone. The lower SFC may have reduced the structural contribution of the crystalline fat phase to dough strength, while the hydrogel phase was associated with lower water mobility. The increased total water content may also have affected interactions among water, flour components and the fat phase. The changes in water mobility occurred together with changes in fat-phase crystallinity and dough matrix morphology, suggesting that water distribution and fat-phase structure may both have contributed to the observed changes in dough properties. Changes in cookie structure and quality associated with dough architecture Cookie microstructure and baking performance The influence of BG incorporation on cookie microstructure was examined by SEM, and the representative images are shown in Fig. 6A. All samples exhibited a porous structure, but the morphology and distribution of the pores differed among the formulations. The VB cookie contained relatively large and irregular pores, with several collapsed cavities. The BG2-VB2 cookie showed more numerous relatively small pores with a more uniform distribution. The BG cookie showed fewer pores in some regions, together with larger irregular voids and more continuous matrix regions. All formulations underwent the same creaming procedure. Differences in the ability of the fat systems to incorporate or stabilize air during creaming may also have contributed to the observed differences in pore structure. Internal microstructure (A) and appearance (B) of cookies prepared with VB, BG2-VB2, and BG. Internal microstructure was observed by scanning electron microscopy. The appearance and baking characteristics are presented in Fig. 6B and Table 2. All cookies maintained an intact shape after baking. Cookie diameter decreased with increasing BG substitution, and the spread ratio decreased from 4.31 for VB to 3.65-3.90 for BG-containing cookies. The lower spread ratio was accompanied by changes in pore morphology and matrix organization observed by SEM. Changes in SFC, fat-phase organization, the hydrogel phase and the increased total water content may have jointly contributed to the reduced spreading during baking. The additional water may also have affected the hydration and thermal behavior of flour components, including gluten proteins and starch. Similar results have been reported for bakery products prepared with structured fat replacers (Pareyt et al., 2009). Table 2Color parameters, baking loss and spread ratio of cookies prepared with different VB-BG blends.Table 2SampleL*a*b*Baking loss (%)Spread ratio (D/H)VB50.63 ± 0.81d7.40 ± 0.36ab22.47 ± 0.42b8.58 ± 0.00a4.31 ± 0.13aBG1-VB352.13 ± 3.12cd6.47 ± 1.24bc22.47 ± 0.67b8.79 ± 0.02a3.90 ± 0.12bBG2-VB255.47 ± 1.03bc8.00 ± 0.20a24.40 ± 0.95a8.82 ± 0.03a3.65 ± 0.03cBG3-VB157.23 ± 2.54ab6.03 ± 0.57c23.10 ± 1.15ab8.93 ± 0.03a3.79 ± 0.18bcBG60.27 ± 1.27a4.10 ± 0.60d21.83 ± 0.91b9.04 ± 0.04a3.82 ± 0.09bcVB, commercial vegetable butter; BG, phytosterol/γ-oryzanol-sodium alginate bigel; L*, a*, and b*, color parameters; D/H, diameter-to-thickness ratio. Different lowercase letters within the same column indicate significant differences (P < 0.05). Color parameters, baking loss and spread ratio of cookies prepared with different VB-BG blends. VB, commercial vegetable butter; BG, phytosterol/γ-oryzanol-sodium alginate bigel; L*, a*, and b*, color parameters; D/H, diameter-to-thickness ratio. Different lowercase letters within the same column indicate significant differences (P < 0.05). Cookie color also changed with BG incorporation. The L* value increased from 50.63 in VB cookies to 60.27 in BG cookies, while the a* value decreased from 7.40 to 4.10. No significant difference was observed in the b* value except for BG2-VB2. The lighter appearance of BG-containing cookies was mainly associated with the lower contribution of β-carotene from VB. The higher water content of the BG-containing doughs may also have contributed to the lighter surface color by delaying the increase in surface temperature during baking and reducing the extent of Maillard browning. Surface browning mainly results from Maillard reactions and caramelization during baking (Sciarini et al., 2013). Baking loss ranged from 8.58% to 9.04%, with no significant differences among the samples. A similar observation has been reported for bakery products prepared with bigels, where incorporation of a hydrogel phase did not markedly alter baking loss (Hou et al., 2024). Mechanical properties and oil retention The breaking force of cookies increased from 25.63 N for VB cookies to 30.97 N for BG cookies, while the BG-VB blends showed intermediate values (Table 3). BG3-VB1 was significantly higher than VB and BG1-VB3. The BG-containing cookies also showed lower spread ratios and different pore morphologies from the VB cookies. The increased water content introduced by the BG hydrogel phase may have affected dough structure during baking and contributed to the higher breaking force. Changes in fat-phase organization, water distribution and dough structure may also have contributed to the higher breaking force of the BG-containing cookies (Mert and Demirkesen, 2016b). The differences in matrix morphology observed by SEM may also be related to the differences in cookie fracture behavior. Table 3Breaking force and oil migration of cookies prepared with different VB-BG blends.Table 3SampleBreaking force (N)Oil migration (%)VB25.63 ± 1.96c0.47 ± 0.00cBG1-VB326.00 ± 1.51bc1.73 ± 0.00bBG2-VB226.40 ± 1.97bc3.48 ± 0.00aBG3-VB129.17 ± 1.26ab1.76 ± 0.01bBG30.97 ± 2.00a0.89 ± 0.01cVB, commercial vegetable butter; BG, phytosterol/γ-oryzanol-sodium alginate bigel. Different lowercase letters within the same column indicate significant differences (P < 0.05). Breaking force and oil migration of cookies prepared with different VB-BG blends. VB, commercial vegetable butter; BG, phytosterol/γ-oryzanol-sodium alginate bigel. Different lowercase letters within the same column indicate significant differences (P < 0.05). Oil migration was 0.47% for VB and 0.89% for BG, with no significant difference between the two samples. The BG-VB blends showed higher oil migration, with BG2-VB2 reaching 3.48%. The higher oil migration in the BG-VB blends may be related to the less aggregated crystal structure and altered lipid-phase organization. Sensory quality The sensory evaluation results are presented in Fig. 7. All samples received acceptable scores for the five evaluated attributes. The flavor score of VB cookies was the highest at 4.13, whereas BG-containing cookies received slightly lower scores. The VB contained food flavoring, whereas no flavor additive was used in the preparation of BG. This difference in formulation may have contributed to the difference in flavor scores. Sensory evaluation of cookies prepared with VB, BG1-VB3, BG2-VB2, BG3-VB1, and BG. The appearance and color scores increased with increasing BG substitution, which agreed with the instrumental color measurements and the more regular shape of BG-containing cookies. Mouthfeel scores ranged from 3.38 to 3.50, with no obvious difference among samples. The overall acceptability score increased from 2.88 for VB cookies to 4.38 for BG cookies. This increase was mainly reflected in the higher appearance and color scores. Conclusions A Phy-Ory-SA bigel altered the crystal organization of the VB, as reflected by changes in crystal aggregation, polymorphic characteristics, SFC and thermal behavior. The blended fat systems showed both β- and β′-type crystal structures, while the contribution of the original VB crystalline phase decreased with increasing BG content. These changes in the fat phase were associated with lower dough viscoelasticity and hardness, while the hydrogel phase was associated with lower water mobility. The changes in the fat phase were associated with changes in cookie structure and quality. Increasing BG content reduced dough spreading during baking, altered the internal pore structure, and increased breaking force. Complete replacement of VB maintained oil migration below 1% and acceptable sensory quality. The present results indicate that changes in crystal morphology, polymorphic characteristics, crystalline fraction and water distribution were associated with changes in dough and cookie properties in the Phy-Ory-SA bigel/VB system. The specific contribution of each polymorph to cookie quality remains to be clarified. Controlling the crystalline structure and solid fat functionality of structured fats may be useful for developing alternative fat systems for bakery applications. Future studies should further investigate the storage stability and flavor evolution of bigel-based bakery products during long-term storage. Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Acknowledgments This research was supported by National Natural Science Foundation of China (CN, 32072134). Handling Editor: Professor Alejandro G.Marangoni Footnotes Supplementary data to this article can be found online at https://doi.org/10.1016/j.crfs.2026.101564. Supplementary data The following is the Supplementary data to this article. Multimedia component 1 References Associated Data Supplementary Materials Multimedia component 1