Tea polyphenol–soy lecithin–beeswax composite oleogelator for soybean oil oleogel: multiscale structural characterization and performance validation in bakery matrices

Abstract

BACKGROUND

Unsaturated fatty acids based oleogels could present higher health benefits than shortening in baked goods, but poor oxidation stability limits their application. Regrettably, natural antioxidants, such as polyphenols, cannot be dispersed directly in grease‐based systems. Herein, the composite gelator (TP–SL–B) was self‐assembled by emulsifying with tea polyphenol (TP), soy lecithin (SL), and beeswax (B).

RESULTS

Proton (¹H) nuclear magnetic resonance analysis validated that the polar choline head of SL was combined with the phenolic hydroxyl group of TP through intermolecular interaction. Based on this stabilized structure, the solubility of TP in soybean oil was greatly enhanced, and thus TP uniformly distributed oleogels (TPOs) were successfully obtained. Although rheological analysis indicated the TPO/shortening blends had weaker deformation tolerance than shortening, replacing 50% shortening with TPO reduced the saturated fatty acid content of cookies by 31.5%. Meanwhile, the peroxide value and sensory evaluation of cookies with half‐replacement of the shortening with TPO were not significantly different from those of the shortening cookies.

CONCLUSION

This study exhibits a strategy for preparing oxidation‐resistant gelators containing natural hydrophilic antioxidants, which has the potential to reduce shortening use in bakery products by partially replacing it with TPO. © 2026 Society of Chemical Industry.

INTRODUCTION

Plastic fat is a mixture formed by the uniform fusion of liquid oil and solid fat, which can impart products with special flavor, texture and taste characteristics.^{1, 2} In particular, the three‐dimensional crystalline network of triglycerides plays a crucial role in the baking field. ³ However, containing saturated or trans fatty acids, traditional plastic fat is related to obesity and cardiovascular disease.^{4, 5} Therefore, many researchers have dedicated time to construct vegetable oil instead of solid fat to reduce saturated fat intake. ⁶ Oleogelation is a simple and healthy way to convert liquid oils into solid‐like gels without chemical modification. ⁷ In brief, the liquid oil is entrapped in a three‐dimensional network structure formed by gelators, with little change in the fatty acid profile.

Nevertheless, the oleogel with high unsaturated fatty acids content shows poor oxidative stability. Natural antioxidants represented by tea polyphenol (TP) exhibit strong antioxidant activity, which can inhibit oxidation reaction by scavenging oxygen free radicals and lipid free radicals, chelating metal ions and inhibiting oxidant enzymes. ⁸ Moreover, TP has various biological activities such as anti‐inflammatory, anti‐cancer and immunomodulation. ⁹ However, the hydrophilicity limits its application in lipids. ¹⁰ In our previous research, a composite gelator containing TP (tea polyphenol–soy lecithin–beeswax, TP–SL–B) was formed by emulsification and lyophilization, with beeswax as gelator and phospholipid as emulsifier. Furthermore, the oleogel prepared by TP–SL–B showed good oxidative stability, with the dispersity of TP significantly improved in the complex system. ¹¹ However, the stabilization mechanism and the interactions between the components in TP–SL–B still need further study, especially the role of lecithin in improving the dispersion of TP in beeswax systems, which is long‐term for expanding the application of hydrophilic substances in oils.

The oleogel with solid‐like structure has broad application prospects in baking products, which requires solid fats to achieve the desired baking characteristics. ¹² The three‐dimensional network structure formed by crystallization or self‐assembly of small molecular gelators shows excellent oil binding properties, and it has been proved that the addition of oleogel is beneficial to reduce the fraction migration of lipids in cookies.^{13, 14} However, it is still difficult to handle and shape the dough owing to the lower viscosity of vegetable oil in oleogel, compared with shortening. Also, the lack of solid fat crystals makes it difficult for the dough to retain bubbles during baking, leading to a less crispy cookie. ¹⁵ Therefore, partial replacement of shortening with oleogel is an effective way to produce more acceptable cookies and expand the application of oleogel.

To this end, the structural properties and formation mechanism of TP–SL–B were investigated by scanning electron microscopy (SEM), ultraviolet (UV) spectroscopy and proton nuclear magnetic resonance (¹H‐NMR) in this study. Then, tea polyphenols uniformed distributed oleogel (TPO) and shortening were mixed at different proportions and added to the cookie formulations. The rheological properties, solid fat content (SFC) and microstructure of TPO and TPO/shortening blends were analyzed and compared with commercial shortening. The quality characteristics of cookies produced by TPO and TPO/shortening blends as the substitute for solid fat were also studied.

MATERIALS AND METHODS

Materials

Soy lecithin (SL) was purchased from Shanghai Macklin Biochemical Co., Ltd (Shanghai, China) and TP with 98% purity was obtained from Hetian Biotechnology Co., Ltd (Hangzhou, Zhejiang, China). Shortening purchased from COFCO Donghai Grain & Oil Industry Co., Ltd (Suzhou, Jiangsu, China), primarily composed of vegetable oils and butter. Beeswax, deuterated methanol, anhydrous methanol, pyrogallic acid, ethanol, hydrochloric acid, ether, petroleum ether, sodium hydroxide–methanol, boron trifluoride–methanol, heptane, sodium chloride, anhydrous sodium sulfate, tetramethylsilane (TMS) and petroleum ether were purchased from Sinopharm Chemical Reagent Co., Ltd (Shanghai, China). The wheat flour, soybean oil (SBO), sugar, eggs and salt were purchased from local stores.

Preparation and micromorphology of TP–SL–B

Beeswax (15 g) and SL (0.9 g) were heated and mixed at 80 °C to form an oil phase solution. TP was dissolved in deionized water (3 mL) to obtain TP aqueous solution with a concentration of 30 mg mL⁻¹. Then the TP solution was emulsified with the oil phase to obtain a water–oil emulsion using a high shear mixer (14 500 rpm, 2 min) under constant temperature (80 °C). The water in the emulsion was removed by lyophilization to obtain the TP–SL–B. Blank gelator (SL–B, made from soy lecithin and beeswax by emulsification) was prepared by the same method without TP addition, ¹¹ and the complex TP–SL was obtained by the same preparation process, but the beeswax was replaced with deionized water.

The interactions in TP–SL–B

Micromorphology

The micromorphology of samples was observed using a scanning electron microscope (SU8100; Hitachi, Tokyo, Japan). Freeze‐dried samples were fixed on an aluminum stub with conductive double‐sided adhesive and coated with gold under vacuum. The examination was performed at an accelerating voltage of 3.0–5.0 kV. Five random sites per sample were imaged, with the clearest image ultimately selected that accurately represented the overall characteristics of the sample.

Ultraviolet (UV) spectrum of TP–SL

The UV absorption spectrum of samples was scanned by a spectrophotometer (TU‐1810; Beijing Purkinje General Instrument Co., Ltd, Beijing, China). TP, SL and TP‐SL were dissolved in anhydrous methanol, respectively, with equal concentration according to the concentration ratio of TP and SL in TP–SL. The UV scanning range of the three solutions was 190–600 nm. ¹⁶

Proton (¹H)‐NMR of TP–SL

TP, SL and TP–SL (10–50 g kg⁻¹) were dissolved with deuterated methanol (99.9%, containing 0.03% v/v TMS) in 2.5 mm NMR tubes, respectively. The ¹H‐NMR of samples was measured using a Bruker Avance AVIII 400 MHz spectrometer (Bruker Instrument Co., Ltd, Billerica, MA, USA). All samples were tested at 25 °C with a resonance frequency of 400 MHz. ¹⁷

Preparation and properties of TPO and TPO/shortening blends

Preparation of TPO and TPO/shortening blends

The SBO and gelators (80–50 g kg⁻¹) were mixed and heated at 80 °C for 15 min, then cooled in refrigerator at 4 °C overnight to form the oleogel. Specifically, TPO and the TP‐free oleogel were prepared from TP–SL–B and SL–B, respectively. TPO and shortening were heated at 80 °C and mixed with different replacement ratios, after cooling in a refrigerator at 4 °C overnight, they were kneaded with low speed stirring in an ice bath at 0 °C to form TPO/shortening blends. Two different proportions of blends were obtained: O50/S50 (50% TPO + 50% shortening), O75/S25 (75% TPO + 25% shortening). All samples were placed in a 25 °C thermostat and stored at least 24 h before testing.

Polarized light microscopy

The crystal microstructure of lipid samples was observed using a DM2700P Leica microscope (Leica Microsystems, Wetzlar, Germany) equipped with a charge‐coupled device (CCD) camera. About 8 μL of sample was placed on a preheated glass slide and covered with a coverslip. Then, stored overnight in a refrigerator at 4 °C. The polarized light micrographs (×200, ×20 objective len and ×10 eyepiece) were taken at room temperature.

Solid fat content

The SFC of lipid samples was determined by low‐resolution pulsed nuclear magnetic resonance (pNMR; Niumag Electronic Technology Co., Ltd, Shanghai, China). According to AOCS method Cd (No. 16b‐93, 1989) to pretreat samples. After equilibrating for 30 min, SFC measurement was officially started at intervals of 5 °C in the range of 20–60 °C.

Rheology

The rheological properties of lipid samples were measured by a DHR‐3 rotary rheometer (TA Instruments, New Castle, PA, USA) with the 40 mm diameter parallel plate geometry and 1000 μm gap. In order to simulate the mixing conditions of dough in the mixer, samples were treated with the same stirring speed and time. ¹⁸ The sample was continuously sheared for 600 s at a constant shear rate of 1900 s⁻¹ to analyze the influence of shear stress in the dough mixing on its rheological properties. In the linear viscoelastic region, the constant strain was 0.01%, the dynamic viscoelasticity of the sheared and unsheared samples in the frequency range of 0.01–100 Hz was measured.

Preparation and evaluations of cookies

Production process of cookies

The different kinds of cookies were prepared using SBO, TPO, O75/S25, O50/S50, and shortening, respectively. The cookie dough recipe on an 100 g wheat flour basis contained 60 g shortening, 15 g sugar, 20 g powdered sugar, 50 g egg and 1 g salt. Initially, the shortening was whipped with a kitchen mixer on speed 1 for 2 min in a bowl. Then sugar and powdered sugar were added to the shortening and stirred at speed 1 for 3 min. The egg was added over three additions and stirred at speed 1 for 2 min each time until it was fully mixed and creamed. After scraping down the mixture from the sides of the bowl, the sieved wheat flour was added and mixed well to form the cookie dough. The dough was equally divided into smaller quantities with a mass of 13 g and pressed into a fixed shape with a mold. After baking in an oven at 175 °C for 17 min, the cookies were cooled for 60 min and then put into polyethylene bags and stored at 25 °C until analysis.

Geometric characteristics

The geometric characteristics of cookies were evaluated according to AACC approved method (10‐54, AACC, 2000). The average diameter (D, in millimeters) and thickness (T, in millimeters) were calculated by measuring six cookies with a calliper. The spread factor was calculated by the ratio of diameter to thickness (D/T).

Texture analysis

The texture of cookies was tested using a three‐point break technique with a texture analyzer (TA‐XT Plus; Stable Microsystems, Godalming, UK). The hardness and fracturability of the sample were calculated from the force–time curve, and the maximum peak breaking force was referred to as the hardness of the sample. The distance at the point of break was the resistance of the sample to bend and was recorded as fracturability of the sample. ¹⁹ For instance, a sample that broke at a very short distance had a high fracturability. The test parameters were set as: pre‐test speed: 1 mm s⁻¹, test speed: 1 mm s⁻¹ and post‐test speed: 2 mm s⁻¹. The cookies were pressed down for 5 mm with a trigger force of 50 g.

Determination of oil migration

The oil migration of cookies was determined using the method as described by Onacik‐Gur and Zbikowska ¹⁴ with some modifications. Five layers of paper was placed in a Petri dish and the total weight was recorded as M ₀. The cookies with mass of M ₁ were put on the paper and placed in an incubator at 30 °C for 24 h. The weight of the paper and Petri dish was measured again, and the weight was recorded as M ₂. The oil migration was calculated as follows:

$$\text{Oil migration}\left(\%\right)=\left(M2-M0\right)/M1\times 100$$ (1)

Fatty acid composition

The fatty acid composition of the cookies was analyzed using an Agilent 7820A gas chromatograph (Agilent, Santa Clara, CA, USA) equipped with a hydrogen flame ionization detector. The oil of cookies was extracted and methylated to be detected. Specifically, 0.5 g sample, 100 mg pyrogallic acid, 2 mL 95% ethanol, a few pieces of boiling stones, and 4 mL of water were transferred to a 250 mL flat‐bottomed flask. Next, 10 mL 8.3 mol L⁻¹ hydrochloric acid solution was added and mixed thoroughly. Then, the flask was placed in a water bath at 75 °C for 40 min to aid hydrolysis, subsequently, the flask was removed and cooled to room temperature. Next, 10 mL 95% ethanol and 50 mL mixed ether–petroleum ether solution were added, then the ether layer was extracted. The extract was collected into a flask and concentrated to dryness using a rotary evaporator. The residue was the fat extract. Then, 8 mL 2% sodium hydroxide–methanol solution were added to the fat extract, and the mixture was refluxed at 80 ± 1 °C in a water bath until all oil droplets disappeared. Then, 7 mL 15% boron trifluoride–methanol solution was added to the top of the reflux condenser, and refluxing was continued for a further 2 min, then heating was stopped and the mixture was rapidly cooled to room temperature. Subsequently, 10 mL of heptane and saturated sodium chloride aqueous solution were added, then, the mixture was left to stand to separate into layers. Next, 5 mL of the upper n‐heptane extraction solution was pipetted into a 25 mL test tube, and 3 g anhydrous sodium sulfate was added. The upper solution was pipetted into the sample vial and injected into a capillary column (60 m × 0.25 mm × 0.25 μm, Thermo Fisher, Waltham, MA, USA). The temperature program was set according to the method of Yuan et al. (2020), ²⁰ and the proportions of various fatty acids were expressed as the weight percentage (%, w/w) of total fatty acids.

Peroxide value monitoring

Initially, to assess the oxidation trends of cookies during long‐term storage, the peroxide value test under accelerated oxidation was carried out. The cookies were put into zippered polyethylene bags and stored at 60 °C for 35 days, while a batch of cookies were taken out every 7 days for determination. The cookies were crushed and extracted with petroleum ether for more than 12 h, and filtered through a funnel filled with anhydrous sodium sulfate. After petroleum ether in the filtrate was removed using a rotary evaporator, the residue was the sample to be tested. The peroxide value of oil extracted from the cookies was determined by the method of AOCS Cd 8‐53.

Sensory evaluation of cookies

According to the scoring standards in Table 1, sensory evaluation was conducted on cookie samples prepared with different types of fats. The products were assigned random numerical codes. Ten panelists (five males and five females, aged 20–30) were recruited, and received brief training prior to sensory evaluation. Using a nine‐point scale, the cookie samples were scored for morphology, color, flavor, texture (internal structure), mouthfeel, and overall acceptability.

Table 1.

Reference for sensory evaluation

Evaluation criteria	Assessment standards	Total score
Morphology	Rich roasted aroma, no off‐flavors (7–9); strong roasted aroma, minimal or slight off‐flavors (4–6); weak roasted aroma, noticeable off‐flavors (1–3).	9
Color	Surface uniformly yellow, no over‐browning or whitening (7–9); surface color relatively even, deep yellow or light white (4–6); surface color uneven, significant over‐browning or whitening (1–3).	9
Flavor	Rich roasted aroma, no off‐flavors (7–9); strong roasted aroma, minimal or slight off‐flavors (4–6); weak roasted aroma, noticeable off‐flavors (1–3).	9
Texture	Uniform and dense porous structure, almost no holes (7–9); relatively uniform porous structure, small holes present (4–6); porous structure uneven/not dense, large holes present (1–3).	9
Mouthfeel	Crisp texture, smooth, not sticky (7–9); moderately crisp or smooth, slightly sticky (4–6); not crisp, rough texture, sticky (1–3).	9
Overall acceptability	Highly liked (7–9); moderately liked (4–6); disliked (1–3).	9

Statistical analysis

Statistical analysis was completed with Origin 9.0 and SPSS 20 software. All experiments (except for the sensory test, which was conducted on ten panelists) were performed with triplicate samples and the data were presented as mean ± standard deviation. The one‐way analysis of variance (ANOVA) and Tukey's test was carried out to analyze the significant difference among samples, and the level of confidence was 95% (P < 0.05).

RESULTS AND DISCUSSION

Micromorphology of TP–SL–B

In our previous study, the composite gelator embedded with TP was obtained after lyophilization to remove water, which could improve the dispersibility of TP in the oleogel. ¹¹ In order to further analyze the structural characteristics of composite gelator, SEM analysis was carried out to observe the surface micromorphology of the samples (Fig. 1). TP particles were spherical structures with obvious wrinkles on the surface (Fig. 1(a)). This was because TP was easy to swell by water absorption, and the outer water film was broken after freeze‐drying treatment, resulting in surface shrinkage. Both SL and beeswax were amorphous structures with smooth surface (Fig. 1(b),(c)). However, the surface of the blank gelator SL–B was attached to some lamellar structures (Fig. 1(d)). This could be due to the fact that the hydrophobic chains of SL combine with beeswax, while the hydrophilic groups are exposed to form a small flaky structure. Compared with SL–B, TP–SL–B had more lamellar structures on the surface and were connected with each other (Fig. 1(e)). It might be that the polar groups of SL interacted with polar TP strongly, which made hydrophilic components exposed and formed more connected lamellar structure. ²¹

Figure 1.

SEM images of tea polyphenol (a), soy lecithin (b), beeswax (c), SL‐B (d) and TP–SL–B (e). TP–SL, tea polyphenol–soy lecithin; TP–SL–B, tea polyphenol–soy lecithin–beeswax.

Ultraviolet (UV) spectrum of TP–SL

Based on the earlier assumptions, the possible interaction mechanism between SL and TP was further explored by spectroscopy techniques, to reveal the critical role of SL for improving the TP dispersion in beeswax. UV spectra of TP, SL and TP–SL are presented in Fig. 2(a). The absorption peaks of SL were very weak. TP had two characteristic absorption peaks (230 nm and 273 nm, caused by phenyl acyl and C=O chromophore, respectively),^{22, 23} of which the maximum absorption wavelength appeared around 230 nm. In the UV spectrum of TP–SL, the characteristic absorption peaks of SL disappeared, while the maximum absorption wavelength was similar to TP, as the peak positions did not shift. This might be caused by the signal of lecithin being overwhelmed by the stronger signal background of TP. However, the absorption peak intensity of TP–SL was slightly lower than that of TP. These results indicated that although the chemical bonds were not formed in the TP–SL, the chromogenic groups of TP were partially obscured, and the electron transition on the conjugated structure in TP was hindered. Specifically, it can be inferred that SL is covered on the phenolic group of TP through intermolecular interaction, ²⁴ which requires further structural characterization.

Figure 2.

UV spectra (a) and ¹H‐NMR spectra (b) of TP, SL and TP–SL. TP, tea polyphenol; SL, soy lecithin; TP–SL, tea polyphenol–soy lecithin.

Proton (¹H)‐NMR analysis of TP–SL

As shown in Fig. 2(b), the interaction sites between TP and SL could be further explored by ¹H‐NMR spectroscopy. The chemical shifts of characteristic peaks for TP, SL and TP–SL are presented in Table 2. But the specific peak attribution could not be obtained from ¹H‐NMR spectrum because TP and SL were mixtures. ²⁵ And the ¹H‐NMR spectrum of TP–SL was not a simple superposition of the two components. The majority of signal peaks appearing in the spectrum of the TP–SL were attributed to SL, while most of the proton signals of TP were shifted, weakened or even disappeared. Specifically, the signals of aromatic protons and phenolic hydroxyl protons (δ 6.967 and δ 6.523) in TP were obviously weakened and moved to the low field by 0.004 ppm and 0.006 ppm, respectively. In addition, the proton signals at the polar segment of SL also changed. The signal peak of the choline part (δ 3.247) weakened and had an upfield shift of 0.012 ppm, while the signal peak of the glycerol part (δ 5.36–5.22) shifted to the low field by 0.011 ppm, 0.013 ppm and 0.015 ppm, respectively. However, the proton signal (δ 1.45–1.19) of the fatty acid chain remained unchanged, implying that the fatty acid chains were not involved in the reaction. Based on the these results, it could be inferred that TP–SL was formed by the weak intermolecular interaction. Given their chemical structures, it can be further inferred that the interactions mainly included hydrogen bonds between the phenolic hydroxyl group in TP and the polar segment of SL.^{26, 27, 28} In summary, the SL played a role of connecter in TP–SL–B. The hydrophobic chains of SL could combine with beeswax, and the polar groups might interact with TP, which would allow TP to be embedded on beeswax and be dispersed in the oleogel.

Table 2.

Chemical shift of crucial peaks in proton nuclear magnetic resonance (¹H‐NMR) spectra

Sample	Chemical shift of crucial peaks (ppm)
TP	6.967 (s, 14H), 6.523 (s, 10H), 5.978 (s, 11H), 5.047 (s, 2H), 4.989 (s, 5H), 4.775 (s, 3H), 3.908 (s, 1H), 3.352 (d, J = 6.0 Hz, 1H), 3.333 (d, J = 3.4 Hz, 12H), 2.996 (d, J = 17.3 Hz, 6H), 2.859 (d, J = 17.4 Hz, 7H)
SL	5.44–5.36 (m, 4H), 5.36–5.22 (m, 3H), 4.460 (dd, J = 12.0, 3.2 Hz, 1H), 4.295 (ddt, J = 7.1, 4.6, 2.6 Hz, 1H), 4.24–4.14 (m, 1H), 4.14–4.01 (m, 2H), 3.70–3.61 (m, 2H), 3.247 (s, 6H), 2.816 (dt, J = 14.0, 6.1 Hz, 3H), 2.350 (dt, J = 10.7, 7.4 Hz, 4H), 2.15–2.01 (m, 6H), 1.621 (p, J = 7.1 Hz, 4H), 1.45–1.19 (m, 32H), 0.926 (td, J = 6.8, 3.5 Hz, 6H)
TP–SL	5.44–5.34 (m, 3H), 5.38–5.31 (m, 2H), 5.34–5.21 (m, 1H), 4.46 (dd, J = 12.0, 3.2 Hz, 1H), 4.29 (ddp, J = 7.0, 5.2, 2.7 Hz, 1H), 4.22–4.17 (m, 1H), 4.15–3.96 (m, 2H), 3.69–3.61 (m, 2H), 3.23 (s, 6H), 2.81 (dt, J = 14.0, 6.0 Hz, 3H), 2.35 (dt, J = 10.6, 7.4 Hz, 4H), 2.15–2.01 (m, 6H), 1.63 (q, J = 7.2 Hz, 4H), 1.45–1.29 (m, 31H), 1.07–0.92 (m, 1H), 0.94 (s, 1H), 0.92 (dd, J = 7.0, 3.5 Hz, 4H), 0.37 (s, 1H)

Note: ‘s’, ‘d’, ‘t’, ‘q’, ‘m’ represents single‐peak, double‐peak, triple‐peak, quadruple‐peak, and multiple‐peak, respectively. The multi‐letter combination represents a corresponding combination of peak types. TP, tea polyphenol; SL, soy lecithin; TP–SL, tea polyphenol–soy lecithin.

Characteristics of TPO and TPO/shortening blends

Crystal morphology of TPO and TPO/shortening blends

In order to explore the feasibility of TPO as a substitute for shortening, the properties of TPO and TPO/shortening blends with different replacement ratios (50:50 and 75:25) were investigated. Microstructure of the TPO, TPO/shortening blends and shortening was evaluated by polarized light microscopy (Fig. 3). The internal morphology of TPO was of needle‐like crystals while the crystal form of shortening was dendritic. Given that it is mainly composed of liquid fat, the needle‐like crystals in TPO were likely formed by the complex of TPs and beeswax. The difference in the types of crystals might be due to the different crystalline platelet orientations and crystallization rates of beeswax and triglyceride. ¹⁹ When TPO was mixed with shortening, in addition to needle‐like crystals, triglyceride and beeswax crystals were also densely gathered together as clusters. Furthermore, with the increase of shortening content, the size and number of spherical crystal units increased. ²⁹ This is beneficial for enhancing the plasticity of blends, when high‐crystalline shortening is involved. ¹⁸ However, the organoleptic evaluation of the application to bakery products is still debatable, because it is related to the degree of crystallinity and SFC at a specific temperature.

Figure 3.

Polarized micrographs of TPO and TPO/shortening blends. Scale bar: 50 μm. TPO, tea polyphenol uniformly distributed oleogels; O75/S25, 75% TPO + 25% shortening; O50/S50, 50% TPO + 50% shortening.

Solid fat content of TPO and TPO/shortening blends

The SFC analysis is necessary to reflect the proportion of solid components in fat at a certain temperature. In the process of dough preparation, the fat with high SFC lacks enough liquid oil to provide sufficient aeration, while the fat with low SFC has no ability to keep air until the mixing is completed. ³⁰ Therefore, the appropriate SFC is essential for the dough to shorten or to develop air cells. ³¹ In Fig. 4(a), it was obvious that the SFC values of TPO were much lower than those of commercial shortening at 20 °C. These results suggest that the SFC in the TPO system primarily originated from the crystallization of beeswax, while SBO did not undergo solidification. Instead, it remained uniformly encapsulated as small droplets within the crystalline network formed by beeswax, constituting a gel‐like system. The blends with higher SFC could be obtained by adding shortening into TPO, which was consistent with the conclusion obtained by polarized micrographs (Fig. 3). TPO and TPO/shortening blends were almost converted to liquid at around 40 °C. Thus with the melting point of fat being close to the temperature of the human mouth this might mean that the samples exhibit a smooth and delicate taste.^{32, 33} However, actual taste cannot be directly predicted from SFC alone. We have evaluated specific sensory characteristics in the baked products discussed later.

Figure 4.

Characteristics of TPO and TPO/shortening blends. (a) Solid fat content. (b) Frequency sweep curves of unsheared flour mixed with TPO or TPO/shortening blends. (c) Frequency sweep curves of sheared flour mixed with TPO or TPO/shortening blends. Microscopic images were captured at 200× magnification. Elastic modulus (G′) is represented by filled symbols, and viscous modulus (G′′) is represented by empty symbols. TPO, tea polyphenol uniformly distributed oleogels; O75/S25, 75% TPO + 25% shortening; O50/S50, 50% TPO + 50% shortening.

Rheological properties of TPO and TPO/shortening blends

The crystallization of the gelator formed a space‐filling network to limit the flow of liquid oil, and this solid‐like structure confers gels with shear sensitivity. ³⁴ When the applied stress exceeds a certain range, their dynamic viscoelasticity is changed substantially due to the deformation of the structure. ³⁵ In the process of making cookies with TPO or shortening, it is necessary that they are whipped and mixed with sugar, eggs and flour. In order to analyze the effect of whipping and stirring on rheological behavior of TPO and TPO/shortening blends, the dynamic viscoelasticity of samples was determined after the steady shear, which was performed to simulate dough mixing conditions, and compared with those of unsheared samples.

The frequency sweep test results are shown in Fig. 4(b),(c). The elastic modulus (G′) was higher than the viscous modulus (G′′) for all the samples, meaning a relatively stable solid‐like structure with good frequency independence was formed. But the modulus of all samples was reduced after shearing, indicating that the gel structure was partially destroyed during the dough mixing process. Although the shear resistance of TPO was weaker compared with shortening, the mechanical strength of the samples gradually increased with the increasing proportion of shortening under shear state. Generally, a larger crystal content tended to form a gel structure that exhibited higher G′ and G′′ values, ³⁶ which could be explained by polarized micrographs (Fig. 3). From the rheological results, there is still a large disparity between TPO and shortening, but the strategy of TPO partially replacing shortening is still worth considering, for the sheared O50/S50 blends had better deformation tolerance than that of unsheared. ³⁷

Characteristics of cookies

Geometric characteristics, texture properties and oil migration of cookies

The appearance of cookies produced by SBO, TPO, shortening and TPO/shortening blends respectively are shown in Fig. 5. Wheat flour forms an expandable membrane rather than an elastic network, which causes lateral expansion of cookies during baking, and the spread factor can be used to characterize the geometric characteristics and aeration effect of cookies. ³⁸ As shown in Table 3, the spread factor of cookies increased with the increasing ratio of oleogels to shortening. In other words, the spread factor of cookies is negative correlated to the SFC of lipids. Although the fat with low SFC could aerate well when mixed with the dough, it could not keep the bubbles in the mixed dough. ³⁰ Therefore, due to the rupture of air bubbles in the baking process, the dough gradually spread out, forming cookies with a larger diameter and smaller thickness. The addition of TPO and TPO/shortening blends instead of SBO increased the aeration capacity and decreased the spread factor, which improved the geometric characteristics of cookies.

Figure 5.

Photographs of appearance and slices of cookies produced with soybean oil (a), TPO (b), O75/S25 (c), O50/S50 (d) and shortening (e). TPO, tea polyphenol uniformly distributed oleogels; O75/S25, 75% TPO + 25% shortening; O50/S50, 50% TPO + 50% shortening.

Table 3.

Geometric characteristics, texture properties and oil migration of cookies prepared with SBO, TPO, O75/O25, O50/S50 and shortening

Samples	The spread factor	Hardness (g)	Distance to fracture (mm)	Oil migration (%)
SBO	6.93 ± 0.05a	1786.56 ± 174.88a	1.38 ± 0.04a	3.18 ± 0.67a
TPO	6.74 ± 0.08b	1222.38 ± 85.34b	1.19 ± 0.08b	2.17 ± 0.17a
O75/S25	6.72 ± 0.05b	1427.99 ± 45.29b	0.99 ± 0.06c	0.69 ± 0.03b
O50/S50	6.57 ± 0.05b	1201.22 ± 104.14b	0.92 ± 0.05cd	0.40 ± 0.09b
Shortening	6.09 ± 0.04c	1204.53 ± 129.42b	0.75 ± 0.15d	0.28 ± 0.08b

Note: Data presented as mean ± standard deviation of three replications. Means with different lowercase letters in the same column are significantly different (P < 0.05). SBO, soybean oil; TPO, tea polyphenol uniformly distributed oleogels; O75/S25, 75% TPO + 25% shortening; O50/S50, 50% TPO + 50% shortening.

Texture is one of the important physical properties of cookies. As shown in Table 3, the cookies prepared with SBO had low‐fracture with a deep breaking distance, and their hardness was much higher than those prepared with shortening. The crispy texture of cookies is mainly affected by flour and lipids. In the preparation of dough, liquid oil is dispersed in the dough in the form of liquid drops during mixing, which cannot cover all the surface of the flour. Flour particles mainly composed of starch and protein are easy to contact with water, which promotes the formation of a gluten network and forms harder cookies. ²⁶ TPO and TPO/shortening blends could better wrap around the surface flour particles to reduce the crosslinking of gluten and improve the aeration ability, which was conducive to the formation of crispy cookies.

Furthermore, oil migration will accelerate the oxidative rancidity and affect the storage and sensory quality of cookies. It could be seen from Table 3 that the cookies with SBO had the largest oil migration, while TPO reduced the oil migration by solidifying the vegetable oil, but the effect was not significant. Due to the synergistic effect of beeswax crystals and solid fat crystals, the oil migration in cookies prepared by the TPO/shortening blends decreased effectively. Combining texture properties and oil migration, the cookies prepared by O50/S50 were not significantly different from shortening cookies, so the ratio of 50% oleogel and 50% shortening has the potential to replace 100% shortening products.

Fatty acid composition of cookies

The fatty acid composition of cookies was determined and presented in Table 4. The main fatty acids of cookies with TPO were oleic acid, linoleic acid, palmitic acid and linolenic acid. However, palmitic acid (42.59%) was the highest fatty acid component in cookies with shortening, followed by oleic acid (36.73%) and linoleic acid (11.75%). With the increase of the substitution amount of TPO, the content of saturated fatty acids decreased proportionally, thereby increasing the ratio of unsaturated fatty acids and saturated fatty acids from 0.98 to 4.57, which improved the nutritional value of cookies (edible oils with suitable health benefits is approximately 7.16). ³⁹ Though TPO slightly increased the content of saturated fatty acids for beeswax contained long‐chain fatty acids, fatty alcohols and esters formed by alkanols and alkanoic acids, the fatty acid composition of cookies with TPO was still similar to cookies with SBO. The fatty acid composition of SBO did not change because the addition of beeswax was tiny, and beeswax was combined with SBO physically in TPO. This result is similar to the findings from Sobolev et al., ⁴⁰ that less beeswax addition did not change total unsaturated fatty acid content in liquid oils obviously: 88.6% in pure sunflower oil and 87.5% in 6% wax adding.

Table 4.

Fatty acid composition of cookies prepared with SBO, TPO, O75/O25, O50/S50 and shortening

	SBO	TPO	O75/S25	O50/S50	Shortening
C12:0	—	—	0.05 ± 0.02b	0.09 ± 0.01ab	0.16 ± 0.03a
C14:0	0.06 ± 0.01d	0.07 ± 0.00d	0.24 ± 0.04c	0.42 ± 0.02b	0.77 ± 0.04a
C15:0	—	—	0.03 ± 0.01b	0.04 ± 0.01ab	0.07 ± 0.00a
C16:0	11.69 ± 0.14d	12.83 ± 0.23d	20.33 ± 0.27c	28.07 ± 0.30b	42.59 ± 0.30a
C16:1	0.16 ± 0.01b	0.18 ± 0.01ab	0.36 ± 0.03a	0.25 ± 0.05ab	0.32 ± 0.05ab
C17:0	0.09 ± 0.01a	0.09 ± 0.00a	0.11 ± 0.01a	0.14 ± 0.02a	0.19 ± 0.03a
C18:0	4.10 ± 0.10c	4.19 ± 0.10bc	4.66 ± 0.16bc	5.38 ± 0.33ab	6.51 ± 0.30a
C18:1	24.74 ± 0.09d	24.62 ± 0.12d	27.58 ± 0.02c	31.27 ± 0.27b	36.73 ± 0.58a
C18:2	53.08 ± 0.20a	51.87 ± 0.33a	41.9 ± 0.15b	31.03 ± 0.21c	11.75 ± 0.12d
C18:3	5.34 ± 0.06a	5.22 ± 0.11a	4.00 ± 0.15b	2.64 ± 0.10c	0.49 ± 0.02d
C20:0	0.27 ± 0.02a	0.27 ± 0.01a	0.26 ± 0.01a	0.26 ± 0.00a	0.24 ± 0.01a
C20:1	0.13 ± 0.02a	0.16 ± 0.02a	0.14 ± 0.01a	0.15 ± 0.00a	0.12 ± 0.01a
C22:0	0.25 ± 0.01ab	0.28 ± 0.02a	0.21 ± 0.01b	0.14 ± 0.00c	0.07 ± 0.01d
C24:0	0.09 ± 0.00b	0.23 ± 0.02a	0.13 ± 0.00b	0.14 ± 0.00b	—
Saturated	16.55 ± 0.03d	17.95 ± 0.32d	26.02 ± 0.05c	34.68 ± 0.00b	50.59 ± 0.68a
Unsaturated	83.45 ± 0.02a	82.05 ± 0.33a	73.98 ± 0.05b	65.32 ± 0.00c	49.41 ± 0.68d

Note: Data presented as mean ± standard deviation of three replications. Means with different lowercase letters in the same column are significantly different (P < 0.05). SBO, soybean oil; TPO, tea polyphenol uniformly distributed oleogels; O75/S25, 75% TPO + 25% shortening; O50/S50, 50% TPO + 50% shortening.

Peroxide value variations of cookies

The use of oleogel instead of shortening can improve the fatty acid composition and nutritional value of cookies. But the high content of unsaturated fatty acids makes cookies more susceptible to oxidation during storage. The accelerated storage experiment at 60 °C was used to observe the change of peroxide value of cookies prepared with TPO and TPO/shortening blends. Initially, the peroxide value of all samples was very low and there was no significant difference. With the extension of storage time, the peroxide value of all samples increased at different rates (Fig. 6(a)). Comparing the peroxide value growth curves of SBO cookies to TP‐free oleogel cookies, it was found that oleogel could reduce the generation of oxidation products through solidifying oil, but the effect was not obvious. The oxidation of TP‐free oleogel cookies accelerated significantly in the late storage period, and the peroxide value was close to SBO cookies at the end of storage. However, the peroxide value of cookies prepared with TPO reduced by 33.5% as compared to cookies containing SBO, indicating that the load of TP could improve the stability of the vegetable oil‐based cookies. The addition of shortening could also reduce the formation of peroxide in cookies, which was mainly related to the fact that the saturated fat contained in shortening was not easy to oxidize. It showed that the usage of TPO instead of 50% shortening in cookies not only increased the content of unsaturated fatty acids in cookies, but also maintained good oxidative stability, which has no significant difference with the cookies prepared from shortening.

Figure 6.

Characteristics of cookies. (a) Peroxide value of cookies during the storage period at 60 °C. (b) Sensory evaluation of cookies. SBO, soybean oil; TFO, tea polyphenol‐free oleogel; TPO, tea polyphenol uniformly distributed oleogels; O75/S25, 75% TPO + 25% shortening; O50/S50, 50% TPO + 50% shortening.

Sensorial profile analysis of cookies

Instrumental analysis is helpful to identify some sensory parameters, but it cannot fully express the complexity of human sensory system. The consumer hedonic scores of cookies on the attributes of appearance, color, flavor, structure, taste and overall acceptance are shown in Fig. 6(b). All the cookies presented scores around seven in terms of color, structure and taste. The score of cookies prepared with TPO and SBO was slightly lower than that of shortening cookies, but the score of cookies made with TPO/shortening blends reached the level of shortening cookies. And the score exceeded the central value of the scale, which indicated that consumers were satisfied with these properties of cookies containing oleogel in general. Furthermore, the appearance score of cookies prepared with TPO was above eight, which was significantly higher than that of shortening cookies. This related to the good moldability of beeswax, TPO could be used to produce cookies with complete shape and consistent thickness. The appearance of food is the first impression of sensory properties, which is conducive to improving the overall acceptance. The lower flavor score of cookies made with SBO and TPO could be due to the fact that SBO does not have a milk flavor, which would lead to weak baking aroma. This might be the reason for the decline in their overall impression score. In summary, shortening cookies had the highest overall acceptance score, but the score of cookies prepared from O50/S50 was closest to that of shortening cookies, which presented good acceptance in consumers.

CONCLUSION

In this study, a composite gelator was formed with beeswax, SL and TP. The polar segment of SL was combined with the phenolic hydroxyl group of TP through intermolecular interaction, and the two non‐polar fatty acid chains were interlinked with beeswax. The SFC and network structure of TPO and TPO/shortening blends were weaker than shortening, but the cookies prepared by O50/S50 exhibited acceptable physicochemical properties, such as aeration ability, geometric properties, texture and oil migration, and they were the most similar to the shortening cookies from the sensory perspective. Most importantly, the addition of SBO in TPO elevated the unsaturated fatty acid content of the cookies and TPs inhibited the oxidation of lipids, that gave the cookies higher nutritional value while enhancing storage stability. These results indicated that the emulsification method enabled the preparation of a gelator with hydrophilic antioxidant to produce oleogels with antioxidant properties, and the partial replacement of shortening with oleogels is a more effective way to improve the unsaturated fatty acids content in cookies without destroying the quality.

CONFLICT OF INTEREST

The authors confirm that they have no conflicts of interest with respect to the work described in this manuscript.

DATA AVAILABILITY STATEMENT

Research data are not shared.