Lipid oxidation, volatile, and taste-active compounds in Yu-Xiang shredded pork during processing

Abstract

This study employed a comprehensive analytical approach to systematically investigate the dynamic changes of flavor compounds in Yu-Xiang shredded pork across five critical processing stages: raw pork, marination, initial stir-frying, seasoning addition, and final sauce incorporation. Results demonstrated that lipid oxidation intensified throughout processing, as evidenced by significantly increased thiobarbituric acid reactive substances (TBARS) values (p < 0.05). Analysis revealed 14 medium- and long-chain fatty acids, with total free fatty acid concentration showing a progressive increase during thermal stages. Gas chromatography-mass spectrometry (GC–MS) identified 100 volatile compounds, including 19 key aroma-active components with odor activity values (OAVs) > 1. Both marination and thermal processing significantly enhanced volatile formation. The incorporation of Yu-Xiang sauce proved crucial for developing characteristic aroma profiles. Nucleotide analysis indicated the inosine monophosphate (IMP) pathway as the primary degradation route. Sixteen free amino acids were quantified, showing marked increases in umami-enhancing aspartic and glutamic acids. These findings reveal the synergistic chemical pathways driving flavor formation in Yu-Xiang shredded pork, offering a mechanistic basis for industrial quality standardization.

Graphical abstract

Highlights

• •GC-MS revealed rising volatile diversity and abundance across processing stages.
• •Thermal processing triggers lipid oxidation and protein degradation, shaping basic aroma and taste.
• •Exogenous seasonings introduce terpenoids and sulfur compounds, enriching the flavor.
• •5′-IMP increased from 218.15 to 491.27 mg/100 g (p < 0.05), indicating IMP as the main degradation pathway.
• •Marinating and stir-frying enhanced umami taste via accumulation of Asp and Glu.

1. Introduction

Yu-Xiang shredded pork is one of the most representative dishes in Sichuan cuisine and has gained widespread popularity in many East and Southeast Asian countries due to its distinctive flavor characteristics. For Yu-Xiang shredded pork, shredded pork is stir-fried with a characteristic blend of pickled peppers, garlic, ginger, and vinegar. This combination yields the so-called “fish-fragrant” flavor, named after Sichuan fish-based dishes. The unique appeal of Yu-Xiang Shredded Pork lies in its sophisticated balance of salty, sweet, sour, spicy, umami, and aromatic elements. This complex flavor profile arises from both the artful blending of seasonings and a series of heat-driven chemical reactions during stir-frying, including lipid oxidation, Maillard reaction, and thermal degradation pathways (Fu et al., 2022; Wojtasik-Kalinowska et al., 2023). A thorough understanding of these flavor formation mechanisms is essential not only for culinary appreciation but also for facilitating the industrial production of traditional dishes.

With the development of the food industry, the industrial production of traditional dishes has become an emerging trend (Yao et al., 2025). The growing demand for ready-to-eat foods has accelerated the industrial transformation of traditional dishes like Yu-Xiang Shredded Pork. However, maintaining authentic flavor and taste in large-scale production remains challenging. A scientific analysis of flavor-forming substances and their dynamic changes during processing is crucial for optimizing industrial protocols, developing effective flavor retention technologies, and establishing quality control standards.

Previous studies have employed various analytical approaches to characterize the flavor profile of Yu-Xiang Shredded Pork. For example, Chen, Cai, et al. (2024) used physicochemical measurements and intelligent sensory analysis to evaluate and compare the characteristics of Yu-Xiang shredded pork made from different ingredients. Ping et al. (2025) investigated the effects of different reheating methods on the odor and taste characteristics of Yu-Xiang shredded pork, demonstrating that microwave heating is the most effective method for retaining the original flavor. Wang et al. (2025) employed a multi-omics approach to decipher the flavor chemistry of reheated Yu-Xiang shredded pork, and systematic deconstruction of flavor variations in reheated dishes is crucial for optimizing their sensory quality.

Despite these advances, significant research gaps remain. Most studies have focused on either the final product or individual seasonings, with limited attention to the dynamic evolution of flavor compounds throughout the cooking process. The impact of critical processing on flavor development remains poorly understood. Furthermore, systematic investigations integrating lipid changes, volatile compound profiles, and taste-active components are notably absent, despite their synergistic roles in overall flavor formation.

To address these research gaps, this study systematically investigates the dynamic changes in lipids, volatile compounds, and taste-active components during the processing of Yu-Xiang shredded pork. The current research aim was to (i) monitor key lipid oxidation indicators—specifically free fatty acids and malondialdehyde (MDA)—across different processing stages; (ii) characterize the evolution of volatile compound profiles using gas chromatography–mass spectrometry; and (iii) quantitatively analyze changes in critical taste-active compounds, including nucleotides and free amino acids. Furthermore, this study seeks to establish meaningful correlations between these chemical transformations and the sensory characteristics of the final product. The findings are expected to offer fundamental insights into flavor formation mechanisms in traditional meat dishes and provide a scientific foundation for optimizing processing techniques, improving product consistency, and standardizing quality in industrial production.

2. Materials and methods

2.1. Materials and sample preparation

To standardize the raw materials, pork tenderloin from commercial white pigs (a typical commercial crossbreed, approximately 7 months old, weighing 90–100 kg, with an intramuscular fat content of approximately 4%) was sourced from Tangshan Shuanghui Food Co., Ltd. The samples were transported under refrigerated conditions (4 °C) and stored at 4 °C for no longer than 24 h prior to processing. All other ingredients and additives required for preparing Yu-Xiang shredded pork were procured from local markets in Beijing, China. The formulations were established based on the local standard “Sichuan Cuisine Cooking Techniques” (SB/T 10946-2012) and preliminary experiments conducted (Table 1). After washing, the tenderloin (300 g) was cut into strips measuring 4 cm × 0.5 cm × 0.5 cm and marinated with the prepared marinade (Table 1) for 20 min. All experiments were performed in triplicate. Carrots, black fungus, and asparagus were cut into fine strips. The traditional stir-frying procedure was as follows: oil was added to a wok and heated. Marinated pork was then added and stir-fried for 120 s at 1300 W (corresponding to an approximate pan temperature of 160–180 °C) until the color changed. Subsequently, the seasonings and ingredients (Table 1) were added and stir-fried for 120 s. Finally, the Yu-Xiang sauce was added, and the mixture was stir-fried for 120 s at 1500 W (approximately 180–200 °C) until the sauce thickened, thus completing the cooking process. The mixture was manually stirred at approximately regular intervals (about every 3 s) to ensure uniformity. The samples were divided into five groups: Y0 (raw pork), Y1 (marinated pork), Y2 (stir-fried until color changed), Y3 (stir-fried with seasonings and ingredients), and Y4 (the final dish).

Table 1.

Dosage of marinade and seasonings.

Class	Name	Dosage (g)
Marinated preparations	Salt	3
	Starch	5
	Cooking wine	10
Seasonings and ingredients	Fungus	30
	Green bamboo shoots	30
	Carrots	30
	Green onion	9
	Ginger	2
	Garlic	3
Yu-Xiang sauce	Red pickled pepper	60
	Bean paste	15
	Monosodium glutamate	3
	Salt	6
	Sugar	24
	Light soy sauce	6
	Dark soy sauce	6
	Vinegar	24
	Starch	18
	Water	30

2.2. Analysis of fatty acid composition

The FFA content was determined using the reported method by Xia et al. (2021) with slight modification. A 5.0 g samples was homogenized with 40 mL chloroform-methanol mixture (2:1, v/v) at 4500 g for 10 s. After standing for 1 h, the homogenate was filtered through two-layer qualitative filter paper. The filtrate was washed with 20% volume of a solution containing 7.3 g/L NaCl and 0.5 g/L CaC12, and centrifuged at 4000 ×g for 15 min. The lower phase was evaporated using a rotary evaporator at 44 °C to harvest the total lipids. Lipids (100 mg) were dissolved in 1.0 mL of chloroform and the resulting solution was loaded onto a pre-conditioned aminopropyl solid-phase extraction (SPE) cartridge (100 mg). The gel cartridge was washed with 2.0 mL chloroform/2-propanol (2,1, v/v) to remove neutral lipids, then the FFAs were eluted by 3.0 mL 2% acetic acid (w/w, dissolved in ether) and dried under nitrogen. For methyl esterification, the FFAs were mixed with 2 mL 14% mass fraction of BF3/methanol and kept at 60 °C for 30 min. Then 2, 2dimethoxypropane was added as the dehydration reagent. After cooling, the solution was mixed with 1 mL H₂O and 1 mL n-hexane, shaken for 5 s and stood for 1 h. The upper phase was collected, dried with nitrogen and re-dissolved in 0.5 mL n-hexane for the following gas chromatography analysis.

Free fatty acids (FFAs) were analyzed by gas chromatography (GC-14B, Shimadzu, Japan) equipped with a flame ionization detector. A 100 mm × 0.25 mm capillary column with a high-polarity polydihydrodiphenylsiloxane stationary phase (film thickness 0.2 μm) was used. Nitrogen was used as the carrier gas, with the flow rate adjusted to 1 mL/min and the injection volume set at 1uL. The injector temperature was maintained at 270 °C, and the detector temperature at 280 °C. The temperature program started at 100 °C (held for 13 min), increased at 10 °C/min to 180 °C (held for 6 min), then rose at 1 °C/min to 200 °C (held for 20 min), and finally increased at 4 °C/min to 230 °C (held for 10.5 min). Under these chromatographic conditions, the fatty acid standard solution and sample solution were injected, with quantification based on peak areas. All fatty acid contents are expressed in g/100 g sample. This method minimizes the confounding effects of moisture loss and prevents artificial increases in fatty acid levels due to changes in oil content during cooking.

2.3. Analysis of thiobarbituric acid reactive substances (TBARS)

The TBARS value was determined based on the reference with modifications (Xia et al., 2021). No external MDA standard was used for calibration; instead, quantification was based on the theoretical millimolar absorptivity coefficient of MDA (ε = 155 L·mol⁻¹·cm⁻¹). The sample (5.0 g) was homogenized with 25 mL of 7.5% trichloroacetic acid solution at 10,000 rpm for 30 s. The homogenate was filtered, and 5 mL of filtrate was mixed with 5 ml of 0.02 mol/L thiobarbituric acid solution. The mixture was vortexed thoroughly, heated in a water bath at 100 °C for 40 min, and then cooled for 1 h. The solution was centrifuged at 5000 ×g for 5 min at 4 °C. The supernatant was then collected, mixed with 1.0 mL chloroform, and layered. The upper layer was obtained for the following analysis. The absorbance of each sample was measured at 532 nm and 600 nm using 96-well plates. TBARS values were calculated using the following formula: TBARS (mg/kg) = (A532 − A600)/155 × 72.06 × 1000/10, where A532 and A600 are the absorbances at 532 nm and 600 nm, respectively; 155 is the millimolar absorptivity of MDA (L·mol⁻¹·cm⁻¹); 72.06 is the molar mass of MDA (g/mol); 1000 converts g to mg; and 10 accounts for sample dilution. All measurements were performed in three independent biological replicates (n = 3).

2.4. Analysis of volatile compounds by GC–MS

The volatile compounds of Yu-Xiang shredded pork at different stages were extracted and quantified by headspace solid-phase microextraction gas chromatograph-mass spectrometer (HS-SPME-GC–MS).

Following a previously published method of Wang, Fan, et al. (2023), a 5 g homogenized sample was transferred to a 20 mL headspace vial. Prior to sample extraction, the 65 μm PDMS/DVB SPME fiber was preconditioned at 250 °C for 30 min according to the manufacturer's instructions. Then, 10 μL of 2-methyl-3-heptanone internal standard solution (0.078 g/L) was added, and the vial was immediately sealed. After equilibrating for 15 min, volatile compounds were collected using a 65 μm PDMS/DVB fiber via automated solid-phase microextraction (SPE) at 60 °C for 30 min. Finally, the analytes were thermally desorbed in the GC–MS injector for 10 min.

Chromatographic separation was performed using an HP-5MS capillary column (30 m × 0.25 mm, 0.25 μm) with helium as the carrier gas at a constant flow rate of 1.5 mL/min in splitless mode. The injector temperature was maintained at 250 °C. The oven temperature program was as follows: held at 40 °C for 3 min, increased at 5 °C/min to 90 °C, then raised at 10 °C/min to 230 °C and held for 7 min.

Mass spectrometric detection employed electron ionization (70 eV) with an ion source temperature of 200 °C and an interface temperature of 250 °C. The filament current was set to 200 μA and the detector voltage to 1.0 kV. Data were acquired in the mass range of 35–350 m/z.

Volatile compounds were identified by comparing mass spectra with the NIST 08/08 s libraries (match factor > 80). Quantification was based on the internal standard method using the following equation:

$$X_i=A_i/A_s\times C_s$$

where X_i denotes the mass concentration (g/L) of the target analyte; C_s represents the mass concentration (g/L) of the internal standard (2-methyl-3-heptanone); A_i and A_s corresponds to the peak area of the target analyte and the internal standard, respectively. Calibration curves for the internal standard and target analytes were prepared over the relevant concentration range, showing good linearity (R² > 0.99) and reproducibility, ensuring reliable quantification.

The aroma activity value (OAV) was calculated as:

$$\mathrm{OAV}=\mathrm{C}/\mathrm{OT}$$

where C is the concentration of the volatile compound in Yu-Xiang shredded pork (μg/kg) and OT is the odor threshold from the literature. Compounds with OAV > 1 were considered aroma-active.

2.5. Analysis of E-nose

A selective electronic nose system was used to monitor the evolution of odor distribution during the thermal processing of Yu-Xiang shredded pork. While GC–MS identifies and quantifies individual volatile compounds, the E-nose enables rapid and holistic evaluation of overall odor profiles, providing complementary insights into aroma changes throughout processing. Sample analysis was conducted using an electronic nose (E-nose) system (PEN3, AIRSENSE Analytics, Germany) featuring ten metal oxide semiconductor (MOS) sensors with defined specificities: W1C (aromatic compounds), W5S (nitrogen oxides), W3C (ammonia and aromatic compounds), W6S (hydrogen), W5C (olefins and aromatic compounds), W1S (methane and hydrocarbons), W1W (sulfides), W2S (alcohols and aromatic compounds), W2W (aromatic and organic sulfides), and W3S (olefins) (Hu et al., 2025).

After sample preparation, 5.0 ± 0.1 g of homogenized sample was placed into a standardized glass sampling container and equilibrated at ambient temperature (25 ± 1 °C) for 60 min. Prior to analysis, the sensor array was calibrated by purging with clean air at a flow rate of 1.0 L/min for 30 min to ensure a stable baseline. Each measurement cycle consisted of a 150-s sampling period and a 120-s sensor cleaning period, with a constant gas flow of 1.0 L/min maintained throughout. All thermal processing stages were repeated three times, and the average sensor responses were used for subsequent data analysis.

2.6. Analysis of free amino acids

The free amino acid composition was analyzed following the method Zhan et al. (2020) with modifications. Briefly, 5.00 ± 0.01 g of sample was homogenized (10,000 rpm, 1 min) with 15 mL of 5% (w/v) trichloroacetic acid solution using a high-speed homogenizer. The homogenate was allowed to stand for 2 h at room temperature (25 ± 1 °C) for complete protein precipitation, followed by centrifugation at 10,000×g for 30 min at 4 °C. The resulting supernatant was filtered through a 0.22 μm aqueous-phase membrane filter (Millipore, USA) prior to analysis. Amino acid quantification was performed on an Agilent 1260 HPLC system equipped with a UV–Vis detector, using an automated online o-phthaldialdehyde (OPA) derivatization. The OPA reagent consisted of o-phthaldialdehyde dissolved in methanol, 0.4 M borate buffer (pH 9.5), and 2-mercaptoethanol as a reducing agent, ensuring efficient derivatization of primary amino acids. Separation was carried out on a ZORBAX Eclipse AAA column (4.6 × 150 mm, 3.5 μm) maintained at 40 °C. The analysis used a detection wavelength of 338 nm, which corresponds to the maximal absorbance of OPA-derivatized primary amino acids (Liu et al., 2017) and a mobile phase gradient between buffer (phase A, 10 mM sodium phosphate dibasic and 10 mM sodium borate, pH = 8.2) and an organic mixture (phase B: acetonitrile/methanol/water, 45:45:10, v/v/v) at a flow rate of 1.0 mL/min. The gradient started at 2% B, increased to 57% B by 13.4 min, then to 100% B by 13.5 min (held until 15.7 min), and was finally brought back to 2% B for column re-equilibration. The injection volume was 3 μl.

2.7. Analysis of nucleotides

Nucleotide were extracted and analyzed by the following procedure to monitor the ATP degradation pathway (ATP → ADP → AMP → IMP → inosine → hypoxanthine), which is critical for understanding the dynamic formation of taste-active compounds during meat processing. First, 4.00 ± 0.05 g of a thoroughly mixed sample was homogenized with 20 mL of pre-cooled 5% (v/v) perchloric acid in an ice-water bath (10,000 rpm, 30 s). The homogenate was then centrifuged at 8000×g for 10 min at 4 °C, and the supernatant was collected by filter paper. The remaining pellet was re-extracted with 10 mL of chilled 5% perchloric acid and centrifuged again. The combined supernatants were neutralized to pH 6.5 ± 0.1 using 1 M potassium hydroxide and diluted to a final volume of 50 mL with HPLC-grade water. Prior to HPLC analysis, 1 mL aliquots were centrifuged again (8000×g, 10 min, 25 °C) and filtered through 0.22 μm nylon membranes.

HPLC analysis was carried out using a Polaris 5C18-PAQ column maintained at 30 °C, with UV detection at 254 nm. An isocratic elution was applied for 28 min using 0.05 M phosphate buffer (pH 6.5) as the mobile phase at a flow rate of 1.0 mL/min. The injection volume was 20 μl.

2.8. Statistical analysis

All data are presented as mean ± standard deviation (n = 3). Each experiment included three biological replicates, and each measurement was repeated three times. Statistical differences among groups were evaluated using one-way ANOVA followed by Tukey's post hoc test (p < 0.05) in IBM SPSS Statistics 26. Relationships between variables were assessed using Pearson correlation analysis in the same software. All figures were prepared using Origin 2021 software.

3. Results and discussion

3.1. Determination of the free fatty acids (FFA) content

The composition and content of fatty acids in pork were critical factors influencing the flavor, nutritional value, and lipid oxidation stability of its processed products. Free fatty acids (FFAs), mainly derived from the hydrolysis of phospholipids and triglycerides, served as important flavor precursors in meat products. They undergo oxidation to generate volatile flavor compounds, such as aldehydes, alcohols, and ketones (Zhang et al., 2020). As summarized in Table 2, a total of 14 fatty acids were identified in the Yu-Xiang shredded pork, comprising 8 saturated fatty acids (SFA), 3 monounsaturated fatty acids (MUFA), and 3 polyunsaturated fatty acids (PUFA). Palmitic acid (C16:0), stearic acid (C18:0), oleic acid (C18:1), and linoleic acid (C18:2) were identified as the predominant fatty acids in Yu-Xiang shredded pork, accounting for over 90% of the total fatty acid content.

Table 2.

Changes in free fatty acids of Yu-Xiang shredded pork at different stages.

FFA(mg/100 g)	Different processing stages
	Y0	Y1	Y2	Y3	Y4
C6:0	0.18 ± 0b	0.22 ± 0.01b	0.24 ± 0.01b	0.24 ± 0.01b	0.22 ± 0.02a
C8:0	0.21 ± 0.01c	0.23 ± 0.01bc	0.25 ± 0.01ab	0.27 ± 0.02a	0.24 ± 0.02bc
C10:0	0.17 ± 0.01c	0.18 ± 0.01bc	0.21 ± 0.01a	0.2 ± 0.01ab	0.21 ± 0.02a
C11:0	ND	ND	0.1 ± 0b	0.11 ± 0.01ab	0.13 ± 0.01a
C12:0	0.1 ± 0c	0.11 ± 0.01c	0.13 ± 0ab	0.14 ± 0.01ab	0.18 ± 0.05a
C14:0	ND	ND	0.1 ± 0.02c	0.23 ± 0.01a	0.16 ± 0.02b
C16:0	4.55 ± 0.53c	4.79 ± 0.56c	5.32 ± 0.27bc	6.43 ± 0.4ab	7.7 ± 0.72a
C18:0	3.02 ± 0.31a	2.65 ± 1.33a	2.97 ± 0.08a	3.78 ± 0.24a	3.01 ± 0.28a
ΣSFA	8.23 ± 0.87b	8.17 ± 1.93b	9.32 ± 0.42ab	11.41 ± 0.72a	11.83 ± 1.14a
C16:1	0.08 ± 0.01d	0.09 ± 0.01cd	0.13 ± 0.01a	0.11 ± 0.01b	0.1 ± 0.01bc
C18:1	1.33 ± 0.19ab	1.02 ± 0.79b	1.71 ± 0.1ab	1.85 ± 0.3a	1.84 ± 0.11a
C22:1	0.1 ± 0.02c	0.13 ± 0.01c	0.09 ± 0.01c	0.25 ± 0.02b	0.45 ± 0.04a
ΣMUFA	1.5 ± 0.22bc	1.23 ± 0.81c	1.94 ± 0.11abc	2.21 ± 0.33ab	2.38 ± 0.16a
C18:2	0.7 ± 0.07c	0.47 ± 0.03c	1.14 ± 0.05b	1.66 ± 0.12a	1.82 ± 0.17a
C18:3	ND	ND	0.09 ± 0.01a	0.09 ± 0.02a	0.08 ± 0.01a
C20:5	0.09 ± 0.01bc	0.06 ± 0.01d	0.07 ± 0cd	0.11 ± 0.01b	0.14 ± 0.03ab
ΣPUFA	0.79 ± 0.08c	0.53 ± 0.03d	1.3 ± 0.06b	1.86 ± 0.15a	2.03 ± 0.2a
ΣUFA	2.29 ± 0.3cd	1.76 ± 0.84d	3.24 ± 0.17bc	4.07 ± 0.48ab	4.42 ± 0.36a
ΣFFA	10.53 ± 1.17c	9.93 ± 2.77c	12.56 ± 0.59bc	15.48 ± 1.2ab	16.25 ± 1.5a

Note: Different lowercase letters indicate significant difference (p < 0.05); Y0 (raw pork), Y1 (marinated pork), Y2 (stir-fried until color changes), Y3 (stir-fried with seasonings and ingredients), and Y4 (finished product with Yu-Xiang sauce). ND indicates not detected.

The composition and content of FFAs demonstrated an overall increasing trend throughout the processing stages. The total FFAs content remained relatively stable at the Y1 stage (9.93 mg/100 g) compared to the raw meat (10.53 mg/100 g), followed by a substantial increase to 16.25 mg/100 g in the final product (Y4). Saturated fatty acids (SFAs) exhibited a general upward trend during processing, with total SFAs content increasing from 8.23 mg/100 g in raw meat to 11.83 mg/100 g in the final product. Medium-chain SFAs (C6:0-C12:0) remained relatively stable, while long-chain SFAs showed more pronounced changes. Notably, palmitic acid (C16:0), the predominant SFAs, increased significantly from 4.55 mg/100 g to 7.70 mg/100 g, consistent with previous findings in stir-fried meat products (Bai et al., 2021). This increase can be attributed to the combined effects of lipid hydrolysis releasing bound fatty acids and moisture loss during thermal processing, leading to a relative concentration effect. Moreover, the higher oxidative stability of saturated fatty acids compared to unsaturated fatty acids may further contribute to their apparent accumulation. In contrast, unsaturated fatty acids (UFAs) exhibited more complex dynamics during processing. The total UFAs content showed an initial decrease from 2.29 mg/100 g to 1.76 mg/100 g at the Y1 stage, followed by a subsequent progressive increase to 4.42 mg/100 g in the final product.

The distinct evolutionary patterns between SFAs and UFAs highlighted their differential roles in flavor formation. Specifically, the chemical instability of UFAs, attributed to the presence of double bonds, renders them more susceptible to thermal oxidation, leading to the generation of various volatile compounds including alkanes, aldehydes, and ketones that significantly contribute to the aroma profile of Yu-Xiang shredded pork. The stage-dependent variations in FFAs can be explained by shifts in the balance between hydrolytic and oxidative processes. During the marination stage (Y0-Y1), an equilibrium between lipid hydrolysis and FFA oxidation maintained relatively stable total FFA levels. In contrast, the thermal processing stages (Y2-Y4) promoted the activation of lipases at moderate temperatures and non-enzymatic thermal hydrolysis at higher temperatures, ultimately driving significant FFAs accumulation as the rate of hydrolysis surpassed that of oxidation (X. Wang, Wang, et al., 2023). This net accumulation indicated that thermal hydrolysis outpaced oxidative consumption during stir-frying. The pronounced increase in UFAs from Y2 to Y4 proved crucial for flavor development. The oxidative degradation of these dominant UFAs generated key volatile compounds: notably, oleic acid decomposition yielded octanal, nonanal, and heptanal, while linoleic acid oxidation produced hexanal and 2-pentylfuran (Chen, Yuan, et al., 2024). Similar trends in FFAs accumulation during thermal processing have been reported in previous studies on pork. Wang, Nie, et al. (2023) reported progressive lipid oxidation and FFA accumulation during pan-heating of pork belly as temperature increased, which aligns with our findings in Yu-Xiang shredded pork. These compounds collectively enhanced the aromatic complexity of Yu-Xiang shredded pork, demonstrating the systematic transformation from lipid metabolism to the development of its characteristic flavor profile.

3.2. Determination of TBARS

The formation of TBARS was an indicator of lipid oxidation. This process adversely affected the flavor, nutritional value, and shelf-life of food products by generating secondary oxidation products, notably MDA, which served as a biomarker for lipid oxidation (Akhtar et al., 2025). As shown in Fig. 1, TBARS values increased significantly during the stir-frying process of Yu-Xiang shredded pork. The initial TBARS value of the raw pork was 0.06 mg MDA/kg, which increased significantly to 0.41 mg MDA/kg after thermal processing. This finding was consistent with a previous study conducted on chicken and duck (Xiong et al., 2020).

Fig. 1.

Changes in TBARS of Yu-Xiang shredded pork at different stages; Note: Different lowercase letters indicate significant difference (p < 0.05); Y0 (raw pork), Y1 (marinated pork), Y2 (stir-fried until color changes), Y3 (stir-fried with seasonings and ingredients), and Y4 (finished product with Yu-Xiang sauce).

The Y0 sample exhibited the lowest TBARS value among all samples. Marination led to a slight increase to 0.08 mg MDA/kg, though this change was not statistically significant (p > 0.05). This minor rise indicated that the addition of NaCl during marination could moderately promote lipid oxidation (Tang et al., 2001). In contrast, a significant increase in TBARS value was observed as early as stage Y2 once heating was initiated, indicating that temperature was the most critical factor influencing lipid oxidation in cooked meat (Hoac et al., 2006). This effect can be explained by temperature-driven free radical chain reactions, in which thermal energy accelerates the initiation of lipid radicals, followed by propagation through lipid peroxidation and decomposition of hydroperoxides into secondary oxidation products such as MDA. Throughout the stir-frying process of Yu-Xiang shredded pork, where temperatures exceeded 150 °C, the TBARS value rose significantly from 0.25 mg MDA/kg at stage Y2 to 0.40 mg MDA/kg at Y4. This pronounced increase suggested that high-temperature heating in the presence of oil considerably promoted the formation of MDA in the dish. The elevated temperature accelerates the breakdown of lipid hydroperoxides into aldehydes and other secondary oxidation products, thereby intensifying TBARS values, which aligns with observations reported in stir-fried meat systems (Bai, You, Wang, & Luo, 2022). In addition, these lipid oxidation-derived aldehydes can further interact with amino compounds through Maillard-type reactions, contributing to the formation of complex aroma compounds and enhancing the overall flavor profile.

3.3. Determination of volatile organic compounds (VOCs) using GC–MS

To better understand the flavor profile of Yu-Xiang shredded pork, volatile compounds from different stages were analyzed using gas chromatography–mass spectrometry (GC–MS). According to the results presented in Table 3, a total of 100 VOCs were identified in Yu-Xiang shredded pork, including 21 alcohols, 21 aldehydes, 5 acids, 14 esters, 4 ketones, 21 hydrocarbons, and 14 heterocyclic compounds. The number of volatile compounds increased progressively across processing stages, with 19 compounds detected in raw meat (Y0), 31 after marination (Y1), 43 after initial stir-frying (Y2), 56 after seasoning addition (Y3), and 75 in the final product (Y4). This clear ascending trend demonstrates that the volatile flavor profile became increasingly complex as processing advanced, with each stage—particularly thermal treatment and seasoning incorporation—contributing to the diversification of aroma compounds.

Table 3.

Changes in volatile compounds during processing of Yu-Xiang shredded pork.

Compound	Y0(μg/kg)	Y1(μg/kg)	Y2(μg/kg)	Y3(μg/kg)	Y4(μg/kg)
Aldehyde
Hexanal	6.85 ± 1.53b	15.35 ± 4.67b	136.36 ± 17.6a	151.43 ± 17.11a	136.38 ± 5.67a
Nonanal	7.44 ± 2.27b	14.36 ± 2.41ab	31.14 ± 7.43a	33.29 ± 11.4a	32.43 ± 9.86a
(E)-2-Nonenal	0.4 ± 0.34b	0.41 ± 0.47b	2.66 ± 0.22a	3.29 ± 0.49a	0.54 ± 0.31b
Decanal	1.14 ± 0.65bc	0.44 ± 0.18c	2.9 ± 0.81ab	4.59 ± 0.8a	1.65 ± 0.68bc
2-Ethylhexanal	0.72 ± 0.72a	0.38 ± 0.05ab	0.77 ± 0.17a	ND	ND
2,4-Dodecadienal	ND	ND	1.29 ± 0.15a	1.11 ± 0.34a	1 ± 0.14a
(E)-2-Decenal	0.18 ± 0.12a	0.11 ± 0.03ab	ND	ND	ND
(E)-2-Undecenal	ND	ND	0.63 ± 0.25a	0.32 ± 0.08b	0.31 ± 0.07b
Dodecanal	ND	ND	0.28 ± 0.17a	ND	ND
Heptanal	ND	0.17 ± 0.1b	ND	0.4 ± 0.17b	2.94 ± 1.75a
(E)-2-Octenal	ND	0.29 ± 0.07d	6.97 ± 0.23c	16.58 ± 1.86a	11.34 ± 1.25b
(E)-2-Heptenal	ND	ND	2.44 ± 1.13ab	2.81 ± 0.4a	0.87 ± 0.72bc
Benzaldehyde	ND	ND	ND	1.45 ± 0.21a	1.96 ± 0.66a
Octanal	ND	ND	6.01 ± 1.27b	17.3 ± 2.54a	16.66 ± 2.85a
4-Ethylbenzaldehyde	ND	ND	0.41 ± 0.11c	4.59 ± 0.8b	6.85 ± 1.53a
Neral	ND	ND	ND	12.33 ± 11.17a	11.34 ± 2.25a
(Z)-2-Decenal	ND	ND	2.93 ± 0.83a	3.06 ± 1.08a	2.39 ± 0.71a
(E,E)-2,4-Dodecadienal	ND	ND	ND	0.5 ± 0.3a	0.22 ± 0.08ab
Furfural	ND	ND	ND	ND	5.83 ± 1.11a
Phenylacetaldehyde	ND	ND	ND	ND	3.03 ± 1.36a
Citronellal	ND	ND	ND	ND	0.19 ± 0.1a
Alcohol
1-Octen-3-ol	1.78 ± 0.56c	5.79 ± 2.6c	38.78 ± 6.75a	27.18 ± 6.23ab	21.43 ± 5.02b
2-Ethyl-1-hexanol	7.35 ± 1.71b	15.2 ± 2.74a	6.48 ± 2.29b	ND	ND
(E)-2-Octen-1-ol	0.44 ± 0.31c	1.03 ± 0.12c	4.59 ± 0.73b	8.29 ± 1.73a	0.98 ± 0.34c
1-Octanol	1.04 ± 0.67c	2.75 ± 0.24c	12.05 ± 1.21a	6.02 ± 2.08b	1.08 ± 0.17c
2-Propyl-1-heptanol	6.16 ± 0.53b	8.04 ± 0.66b	14.54 ± 3.54b	40.9 ± 7.88a	36.77 ± 5.77a
1-Heptanol	ND	0.27 ± 0.06c	1.03 ± 0.2b	2.44 ± 0.38a	ND
2-Butyl-1-octanol	ND	0.16 ± 0.05c	1.03 ± 0.2b	1.11 ± 0.34b	1.73 ± 0.46a
1-Eicosanol	ND	0.49 ± 0.02a	ND	ND	ND
trans-2-Undecen-1-ol	ND	ND	0.23 ± 0.21a	ND	ND
1-Hexanol	ND	ND	ND	0.42 ± 0.18a	ND
2-Ethyl-2-hexen-1-ol	ND	ND	0.29 ± 0.08a	ND	ND
Nerolidol	ND	ND	ND	0.16 ± 0.1ab	0.44 ± 0.31a
Eucalyptol	ND	ND	ND	10.92 ± 2.33ab	22.37 ± 10.42a
α-Terpineol	ND	ND	ND	0.96 ± 0.36ab	4.41 ± 0.95a
6-Methyl-1-heptanol	ND	ND	ND	ND	0.83 ± 0.53a
Linalool	ND	ND	ND	ND	6.51 ± 1.8a
2,4-Dimethyl-3-hexanol	ND	ND	ND	ND	0.2 ± 0.07a
3-Decyn-2-ol	ND	ND	ND	ND	0.18 ± 0.02a
Citronellol	ND	ND	ND	ND	0.2 ± 0.17a
Phloroglucinol	ND	ND	ND	ND	0.12 ± 0.05a
cis-Verbenol	ND	ND	ND	ND	4.38 ± 1.18a
Acid
Nonanoic acid	0.37 ± 0.28bc	0.23 ± 0.16c	0.3 ± 0.09bc	1.69 ± 0.54a	1.25 ± 0.5ab
Octanoic acid	0.25 ± 0.13bc	ND	ND	0.49 ± 0.19b	1.03 ± 0.2a
Dehydroacetic acid	ND	ND	ND	ND	4.74 ± 1.44a
Pterin-6-carboxylic acid	ND	ND	ND	ND	0.15 ± 0.06a
2-Ethylhexanoic acid	ND	ND	ND	ND	0.11 ± 0.04a
Ester
Methyl hexanoate	ND	8.5 ± 1.04b	13.82 ± 3.84a	ND	ND
Ethyl hexanoate	ND	15.36 ± 3.65a	ND	ND	ND
Octyl formate	ND	ND	ND	7.79 ± 0.46a	ND
Allyl 2-ethylbutyrate	ND	ND	ND	9.28 ± 8.67a	ND
Methyl octanoate	1.17 ± 0.29b	1.59 ± 0.3a	ND	ND	ND
Methyl nonanoate	0.85 ± 0.65a	0.61 ± 0.3b	ND	ND	ND
Ethyl octanoate	ND	1 ± 0.41a	ND	ND	ND
Ethyl nonanoate	ND	0.18 ± 0.1b	0.58 ± 0.12a	ND	ND
Allyl tridecyl oxalate	ND	ND	0.72 ± 0.03ab	0.44 ± 0.31b	1.07 ± 0.24a
Allyl 2-ethylbutyrate	ND	ND	0.81 ± 0.31a	ND	ND
Nonyl docosanoate	ND	ND	ND	ND	0.3 ± 0.13a
Methyl salicylate	ND	ND	ND	ND	0.68 ± 0.16a
Methyl 2-ethyldecanoate	ND	ND	ND	ND	0.19 ± 0.1a
Ethacrynate ester	ND	ND	ND	ND	0.14 ± 0.03a
Ketone
2-Heptanone	ND	0.89 ± 0.2b	4.48 ± 1.14a	0.44 ± 0.18b	ND
2-Methyl-3-octanone	ND	ND	23.84 ± 6.49a	ND	ND
2,3-Octanedione	ND	ND	2.53 ± 1.04ab	5.25 ± 2.74a	2.53 ± 1.04ab
Acetophenone	ND	ND	ND	0.82 ± 0.08b	1.93 ± 0.7a
Alkane
Hexadecane	1.48 ± 1.17bc	4.07 ± 1.68a	3.74 ± 0.45b	2.24 ± 0.78bc	0.29 ± 0.06c
3-Methyldecane	0.83 ± 0.3b	0.94 ± 0.16ab	2.19 ± 1.3a	1.3 ± 0.15ab	1.89 ± 0.3ab
2-Methylundecane	ND	ND	0.29 ± 0.06a	ND	ND
2-Pentadecyl-1,3-dioxolane	ND	ND	ND	ND	0.31 ± 0.08a
4-Cyclopropylcarbonyloxytridecane	ND	ND	ND	ND	0.22 ± 0.08ab
Tridecane	0.18 ± 0.12b	0.51 ± 0.05ab	1.26 ± 0.38ab	1.68 ± 0.93a	1.48 ± 0.36a
Nonadecane	0.76 ± 0.08ab	0.46 ± 0.06b	0.53 ± 0.05ab	0.87 ± 0.3a	0.54 ± 0.25ab
6-Methyloctadecane	ND	ND	ND	0.27 ± 0.14a	0.32 ± 0.04a
2,6-Dimethylheptadecane	ND	ND	ND	0.41 ± 0.17a	ND
3-Methyltridecane	0 ± 0a	0.16 ± 0.01a	0.54 ± 0.58a	0.24 ± 0.05a	0 ± 0a
Tetradecane	ND	0.53 ± 0.08b	1.11 ± 0.14a	ND	0.75 ± 0.16b
Alkene
1-Decene	ND	0.37 ± 0.05a	ND	ND	0.43 ± 0.15a
cis-β-Farnesene	ND	ND	ND	0.77 ± 0.28b	1.6 ± 0.6a
Limonene	ND	ND	ND	ND	1.39 ± 0.16a
(E)-5-Methyl-4-undecene	ND	ND	0.22 ± 0.09a	ND	ND
5-Nitro-4-nonene	ND	ND	0.24 ± 0.19a	ND	ND
Caryophyllene	ND	ND	ND	14.19 ± 3.08a	16.72 ± 1.38a
Humulene	ND	ND	ND	0.67 ± 0.37a	0.2 ± 0.03b
α-Farnesene	ND	ND	ND	2.73 ± 0.75a	1.7 ± 0.43b
γ-Terpinene	ND	ND	ND	ND	4.6 ± 0.51a
cis-β-Farnesene	ND	ND	ND	ND	0.24 ± 0.09a
Heterocyclic compound
1-(1,5-Dimethyl-4-hexenyl)-5-methylbenzene	ND	ND	ND	2.02 ± 0.49a	1.62 ± 0.9b
Allyl methyl sulfide	ND	ND	ND	3.07 ± 0.3b	4.26 ± 0.91a
2-Pentylfuran	ND	0.95 ± 0.66c	7.05 ± 3.19bc	13.86 ± 2.8ab	16.44 ± 5.13a
2-Ethyl-6-methylpyrazine	ND	ND	ND	1.12 ± 0.74b	1.89 ± 0.3a
Diallyl disulfide	ND	ND	ND	10.56 ± 2.01b	14.93 ± 0.65a
S-Methyl hexanethioate	ND	ND	ND	28.58 ± 4.82a	30.86 ± 6.19a
6-Methylheptyl vinyl ether	ND	0.19 ± 0.07b	0.62 ± 0.24b	1.64 ± 0.64a	2.02 ± 0.49a
endo-Borneol	ND	ND	ND	1.63 ± 0.49a	1.12 ± 0.74a
Camphor	ND	ND	0.35 ± 0.14ab	0.57 ± 0.33a	0.79 ± 0.13a
2,4-Di-tert-butylphenol	ND	ND	0.33 ± 0.15b	2.24 ± 0.81a	1.59 ± 0.3a
Isoborneol	ND	ND	ND	0.51 ± 0.14a	0.09 ± 0.03b
trans-α-Bergamotene	ND	ND	ND	0.99 ± 0.28b	1.63 ± 0.49a
Geranyl vinyl ether	ND	ND	ND	ND	0.98 ± 0.2a
Dipropyl disulfide	ND	ND	ND	ND	4.33 ± 2.87a

Note: Different lowercase letters indicate significant difference (p < 0.05); Y0 (raw pork), Y1 (marinated pork), Y2 (stir-fried until color changes), Y3 (stir-fried with seasonings and ingredients), and Y4 (finished product with Yu-Xiang sauce). ND indicates not detected.

Volatile aldehydes significantly influence the aroma profile of Yu-Xiang shredded pork due to their low odor thresholds, enabling them to impart distinct aromatic notes even at minimal concentrations. In the Y0 sample, a small quantity of aldehydes was detected due to autoxidation of its lipids (Wang, Wang, et al., 2023). Following marination, the total content and variety of aldehydes increased. It was noted that compounds such as heptanal and (E)-2-octenal were newly generated. This enhancement was likely attributable to salt-induced stimulation of lipid oxidation during marination and the salting-out effect on flavor compounds (Wang, Wang, et al., 2023). Heating initiation at stage Y2 markedly accelerated aldehyde formation, with the number of aldehyde species increasing from 8 to 13 and total aldehyde content rising sharply from 31.5 μg/kg to 194.8 μg/kg. This phenomenon reflected the dominant role of temperature in driving lipid oxidation. Specifically, the oxidation of unsaturated fatty acids such as oleic acid (C18:1) and linoleic acid (C18:2) leads to the formation of characteristic aldehydes, including nonanal and octanal from oleic acid, and hexanal and 2-pentylfuran from linoleic acid oxidation. Throughout the heating process from stages Y2 to Y4, both the diversity and total content of aldehydes continued to increase. It was noted that the contents of hexanal and nonanal did not exhibit a continuous increase with prolonged heating, but rather increased initially and then decreased at stage Y4. This observation was consistent with the findings reported by Bai et al. (Bai et al., 2021). This pattern suggested that moderate heating promoted aldehyde formation, whereas prolonged heating might lead to their further reaction or degradation. Following the addition of seasonings and ingredients during stages Y3 and Y4, a variety of aromatic aldehydes were generated, including phenylethanal (bitter almond aroma), neral (citrus and grassy notes), furfural (almond-like and sweet odors), and citronellal (lemongrass and Sichuan pepper-like scents). Phenylethanal had been identified as a key aroma component not only in Chinese soybean paste (Zhao, Fan, & Xu, 2021). Neral likely derived from red bubble pepper, while furfural was presumably generated through the caramelization of sugars during heating. These compounds, primarily derived from the seasonings, enhanced both the intensity and complexity of the overall aroma profile of Yu-Xiang shredded pork.

In general, alcohols were formed through the decomposition of secondary fatty acid hydroperoxides. Due to their higher odor thresholds, they were generally considered to contribute less significantly to the flavor profile of Yu-Xiang shredded pork compared to aldehydes (Zhao, Sun, et al., 2021). The raw meat (Y0) contained 5 types of alcohols with the lowest total concentration (16.8 μg/kg). After marination (Y1), the number increased to 8 types with a concentration of 33.73 μg/kg, showing a significant increase (p < 0.05). During the heating stages from Y2 to Y4, both the variety and content of alcohols showed a continuous increasing trend. This trend was consistent with observations reported in studies on stir-fried beef topping (Bai, You, Ji, et al., 2022). This could be attributed to the oxidation and thermal degradation of lipids, which generated a substantial quantity of alcohols. Compounds detected across all five stages included 1-octen-3-ol, (E)-2-octen-1-ol, and 1-octanol, which were also present at relatively high concentrations. Among these, 1-octen-3-ol (characterized by a mature mushroom odor) originated from the peroxidation of arachidonic acid. Its concentration peaked at stage Y3 (38.78 μg/kg) and subsequently declined to 21.43 μg/kg by stage Y4. The observed decrease likely resulted from its metabolism into aldehydes, ketones, and other VOCs. Alcohols such as nerol, eucalyptol, and α-terpineol were detected starting from stage Y3 following the addition of spices, contributing floral, woody, and cooling notes to the Yu-Xiang shredded pork. Among these, nerol was a primary volatile component typically found in ginger (Wang et al., 2019). Following the addition of red marinated peppers and doubanjiang at stage Y4, linalool, citronellol, and cis-verbenol were detected, imparting floral and sweet notes, which were likely contributors to the characteristic sweet flavor profile of Yu-Xiang shredded pork. In summary, alcoholic compounds played a crucial role in the formation of the overall flavor profile.

The ester concentrations exhibited non-monotonic kinetic behavior during thermal processing (stages Y1–Y4), imparting a distinct fruity aroma. Owing to their low odor thresholds, they could contribute positively to the overall aroma profile at low concentrations; however, when present excessively, they might impart off-odors and disrupt the aromatic harmony (Carpena et al., 2021). The initial ester content was relatively low in the raw meat (Y0), measuring 2.02 μg/kg. Upon the addition of cooking wine (Y1), the ester concentration increased markedly to 27.24 μg/kg, imparting an initially overpowering and unbalanced odor. During the subsequent thermal processing stages (Y2-Y4), the ester compounds underwent further reactions, degradation, and interconversion, which led to a gradual decrease in their concentration (Wang, Wang, et al., 2023). The final ester content reached 17.51 μg/kg, resulting in a more harmonious aroma for the Yu-Xiang shredded pork. It had been reported that esters were important flavor compounds in fermented seasonings (Liao et al., 2023). The introduction of the Yu-Xiang sauce at the Y4 stage increased some ester species such as allyl 2-ethylbutyrate, nonyl docosanoate, methyl salicylate, and methyl 2-ethyldecanoate, which was likely attributable to the incorporation of doubanjiang.

Acidic compounds, despite their relatively high taste thresholds, played a notable role in shaping the flavor profile of Yu-Xiang shredded pork. From stage Y0 to Y3, the content of acidic compounds remained stable with no significant changes (p > 0.05). However, a marked increase (p < 0.05) was observed following the addition of the Yu-Xiang sauce at stage Y4. This significant rise not only enhanced the overall sourness but also contributed to the characteristic sweet and sour balance that defines the final product's profile.

Heterocyclic compounds, such as pyrazines, furans, and sulfur-containing species, increased considerably during processing. For example, 2-pentylfuran, which contributed bean-like and fruity notes, was first detected during the marination stage and originated from the oxidation of linoleic acid (Flores, 2018). With continued heating, its content exhibited a significant increase (p < 0.05), which was consistent with the findings in an earlier study on stir-fried pork (Wang, Nie, et al., 2023). Pyrazines, particularly 2-ethyl-6-methylpyrazine, contributed characteristically roasted and nutty aromas to Yu-Xiang shredded pork. These compounds first emerged at stage Y3 and accumulated progressively during subsequent thermal processing. Their formation can be attributed to the Maillard reaction. Furthermore, the addition of spices such as ginger, garlic, and chili peppers provided various volatile compounds to the Yu-Xiang shredded pork and enhanced its overall aroma characteristics. Sulfur-containing compounds were primarily derived from garlic (Wang, Wang, et al., 2023). Compounds such as allyl methyl sulfide, diallyl disulfide, and S-methyl thiohexanoate imparted intense garlic and onion odors, contributing a pungent and stimulating flavor that enhanced the overall complexity and harmony of the aroma profile.

Terpenoids represented the largest class of plant secondary metabolites and were common flavoring components in the diet. They possessed low odor thresholds and could impart unique flavor characteristics to meat products. At stage Y3, the addition of ginger and garlic, followed by further processing, led to a significant increase in both the diversity and concentration of terpenoids. Compounds such as caryophyllene, humulene, and α-farnesene were identified, imparting spicy and woody notes. Subsequently, starting from stage Y4, with the incorporation of the Yu-Xiang sauce and the continuous rise in heating temperature, the variety and content of terpenoids and sulfur-containing compounds increased further. γ-Terpinene (citrus and lemon notes) and cis-β-farnesene (woody and floral notes) began to emerge. These observations collectively indicated that the inclusion of aromatic adjuncts such as ginger, garlic, and chili peppers provided multiple sources of volatile compounds to the pork, thereby enhancing the overall complexity and richness of the aroma profile.

To improve interpretability, emphasis was placed on aroma-active compounds (OAV > 1), which were identified as the primary drivers of stage separation. Flavor evolution was visualized via a conventional heatmap (Fig. 2) and an OAV heatmap (Fig. 3), which revealed 19 compounds with OAV greater than 1 in the final product. In the figure, deeper red shades indicated higher relative content of a compound, whereas deeper blue shades represented lower relative content. These results clearly showed that raw meat contained only a limited number of compounds, most exhibiting low odor activity, which likely contributed to its generally unappealing and less harmonious odor. During the marination stage, although the number of reddish areas in the heat map increased slightly—primarily due to esters—most regions remained blue, indicating that the overall aroma remained weak. In the heating stages (Y2 to Y3), the majority of the compounds in the Yu-Xiang shredded pork gradually turned red and intensified in color. The transition from Y3 to Y4 showed considerable divergence in flavor compounds, attributable mainly to the addition of seasonings. This progression demonstrated that as processing advanced, the content, odor contribution, and variety of compounds increased significantly, ultimately forming a well-balanced and characteristic stir-fried aroma. Notably, among the 100 identified volatile compounds, 19 were determined as key aroma contributors (OAV > 1), primarily including aldehydes (e.g., hexanal, nonanal, phenylethanal), selected alcohols (e.g., 1-octen-3-ol), esters, and heterocyclic compounds such as furans and pyrazines. These aroma-active compounds, owing to their low odor thresholds and distinctive sensory characteristics, played dominant roles in defining the overall flavor profile of Yu-Xiang shredded pork.

Fig. 2.

Heat map of volatile flavor compounds; Note: Y0 (raw pork), Y1 (marinated pork), Y2 (stir-fried until color changes), Y3 (stir-fried with seasonings and ingredients), and Y4 (finished product with Yu-Xiang sauce).

Fig. 3.

Heatmap of flavored OAV; Note: Y0 (raw pork), Y1 (marinated pork), Y2 (stir-fried until color changes), Y3 (stir-fried with seasonings and ingredients), and Y4 (finished product with Yu-Xiang sauce).

3.4. E-nose analysis

The E-nose was capable of simulating the human olfactory perception, and demonstrated superior sensitivity in detecting subtle differences between samples compared to conventional sensory evaluation. As shown in Fig. 4, the distinct response patterns of the 10 metal oxide sensors indicated significant changes in volatile compounds throughout processing. The substantially higher total response value in processed samples compared to raw meat confirmed that thermal processing promoted the release and formation of volatile compounds, resulting in greater variety and concentration of aroma-active substances. These findings were consistent with the GC–MS results, demonstrating that the E-nose can effectively capture the overall evolution of aroma profiles during processing. Unlike GC–MS, which provides detailed compound identification, the E-nose offers a rapid and holistic assessment of volatile changes, highlighting its complementary role in flavor analysis.

Fig. 4.

The odor traits changes evaluated by electronic nose; Note: Y0 (raw pork), Y1 (marinated pork), Y2 (stir-fried until color changes), Y3 (stir-fried with seasonings and ingredients), and Y4 (finished product with Yu-Xiang sauce).

Analysis of individual sensor responses provided specific insights into the aroma profile of Yu-Xiang shredded pork. The W5S (sensitive to nitrogen oxides), W1S (sensitive to methane and hydrocarbons), and W1W (sensitive to sulfides) sensors exhibited notably high responses throughout the processing stages. This pattern indicates the presence of substantial amounts of nitrogen oxides, terpenes, and sulfur-containing volatile compounds. Among these, the W5S sensor demonstrated the highest response value, with its signal intensity progressively increasing with extended processing time, suggesting continuous generation of nitrogen oxides during stir-frying. This phenomenon may be attributed to complex biochemical reactions involving proteins (Kong et al., 2024). In contrast, the W5C (sensitive to olefins and aromatic compounds), W3C (sensitive to ammonia and aromatic compounds), and W1C (sensitive to aromatic compounds) sensors exhibited significantly lower responses throughout the processing stages. This can be attributed to their primary sensitivity to aliphatic hydrocarbons, which generally possess higher odor thresholds and consequently elicit weaker sensor responses (Ramírez & Cava, 2007). The response value of the W2S sensor (sensitive to alcohols and aromatic compounds) peaked at the Y1 stage, coinciding with the addition of cooking wine during marination. This increase is likely attributed to the enhanced release of ethanol and other aromatic alcohols from the marinade. As thermal processing progressed, the response subsequently decreased, presumably due to the volatilization of these alcohols—a trend corroborated by GC–MS analysis.

To visualize and comprehensively understand the changes in the aroma characteristics of the dish across different processing and storage stages, Principal Component Analysis (PCA) was applied to perform unconstrained dimensionality reduction on the electronic nose dataset (Fig. 5). The first two principal components (PC1 and PC2) accounted for 94.1% of the total variance (84.8% and 9.3%, respectively). The PCA score plot showed a distinct separation of Yu-Xiang shredded pork samples from five different processing stages, despite some overlap among individual samples. This distribution pattern indicated a progressive transition in odor profiles during thermal processing, with each stage developing distinct aromatic qualities and forming discernible aroma clusters.

Fig. 5.

PCA of electronic nose data; Note: Y0 (raw pork), Y1 (marinated pork), Y2 (stir-fried until color changes), Y3 (stir-fried with seasonings and ingredients), and Y4 (finished product with Yu-Xiang sauce).

3.5. Analysis of free amino acids

The flavor of cooked meat products was largely determined by the content, composition, and odor thresholds of free amino acids (FAAs). FAAs were considered important flavor precursors in meat (Yuan et al., 2021). Catalyzed by endogenous proteases, proteins underwent proteolytic degradation into short peptides. These peptides were subsequently further hydrolyzed into free amino acids (Toldrá, 2006). This proteolytic pathway was primarily responsible for generating compounds in Yu-Xiang shredded pork that were associated with umami, sweet, and bitter tastes (Yin et al., 2023). FAAs exhibited distinct sensory properties. Aspartic acid and glutamic acid were responsible for the umami taste in food, and their concentrations were found to increase markedly during thermal processing, indicating their key role in the formation of the characteristic savory taste of Yu-Xiang shredded pork. Glycine, alanine, serine, threonine, and proline were characterized as sweet-tasting amino acids. Several amino acids, including methionine, leucine, isoleucine, histidine, arginine, and phenylalanine, contributed to bitter notes. These FAAs contributed significantly to the characteristic flavor profile of Yu-Xiang shredded pork.

Changes in the content of FAAs across different stages of Yu-Xiang shredded pork preparation are presented in Table 4. During the cooking process, protein degradation intensified with prolonged heating time, leading to the concentration of most FAAs increasing progressively. The total FAA content demonstrated a progressive upward trend, rising from 911 mg/100 g in the raw meat (Y0) to 1016 mg/100 g after marination (Y1), and further increasing to 1066 mg/100 g (Y2), 1087 mg/100 g (Y3), culminating at 1418 mg/100 g in the final product (Y4). The initial increase during the marination stage (Y0 to Y1) was attributed to the action of endogenous proteases and the physicochemical environment created by the marinade. NaCl disrupted the muscle cell structure by protein denaturation and osmotic stress. This process facilitated the release and activation of endogenous proteolytic enzymes, thereby promoting the hydrolysis of proteins into free amino acids (Ju et al., 2022). The observed increase in FAAs during the subsequent heating stages (Y2-Y4) was primarily driven by continued protein hydrolysis (Lv et al., 2019).

Table 4.

Changes in free amino acids during processing of Yu-Xiang shredded pork.

FAA(mg/100 g)		Different processing stages
		Y0	Y1	Y2	Y3	Y4
UFAA	Asp	2.53 ± 1.79b	3.01 ± 2b	1.87 ± 0.35b	3.89 ± 1.33b	9.2 ± 1.54a
	Glu	3.42 ± 1.3b	11.03 ± 1.82b	10.93 ± 1.78b	10.65 ± 2.01b	158.69 ± 11.82a
SFAA	Ser	6.24 ± 1.59c	6.85 ± 1.33c	2.36 ± 0.71d	11.8 ± 2.37b	25.71 ± 1.77a
	Gly	22.47 ± 5.22a	20.88 ± 5.21ab	21.51 ± 4.51ab	19.13 ± 5.31ab	10.38 ± 8.2b
	Thr	2.34 ± 0.84a	2.66 ± 0.25a	2.6 ± 0.17a	2.63 ± 0.09a	4.62 ± 2.77a
	Ala	513.73 ± 38.74d	631.05 ± 13.39c	698.37 ± 53.48b	695.41 ± 17.02b	839.82 ± 14.04a
	Val	0.63 ± 0.74b	0.56 ± 0.34b	0.74 ± 0.63b	1.2 ± 1.06b	4 ± 1.95a
	Lys	89.08 ± 10.57b	94.95 ± 4.2b	74.02 ± 9.25b	71.79 ± 10b	148.72 ± 23.16a
BFAA	His	63.07 ± 12.9ab	49.03 ± 8.77bc	45.56 ± 4.58c	69.18 ± 7.97a	60.16 ± 5.91abc
	Tyr	88.7 ± 3.78a	74.21 ± 2.12b	76.15 ± 7.13b	68.88 ± 1.42b	55.86 ± 2.41c
	Arg	84.68 ± 4.69a	87.04 ± 20.87a	79.7 ± 30.21a	75.73 ± 29.56a	49.48 ± 19.92a
	Cys-s	14.92 ± 2.6b	16.6 ± 1.57b	36.9 ± 7.1a	36.74 ± 9.98a	31.35 ± 9.51ab
	Met	11.84 ± 0.95a	11.53 ± 2.38a	6.09 ± 5.4a	9 ± 1.11a	12.02 ± 3.12a
	Phe	1.99 ± 0.18bc	2.08 ± 0.09bc	1.66 ± 0.21c	2.31 ± 0.55b	3.14 ± 0.24a
	lle	3.14 ± 2.58ab	2.28 ± 0.69ab	7.2 ± 3.79a	7.72 ± 2.66a	0.12 ± 1.08b
	leu	2.22 ± 1.75ab	2.21 ± 0.98ab	0.72 ± 0.45b	1.13 ± 0.76b	5.24 ± 1.88a

Note: Different lowercase letters indicate significant difference (p < 0.05); Y0 (raw pork), Y1 (marinated pork), Y2 (stir-fried until color changes), Y3 (stir-fried with seasonings and ingredients), and Y4 (finished product with Yu-Xiang sauce).

Concurrently, the content of umami amino acids exhibited a more pronounced increase, with values of 5.94 mg/100 g (Y0), 14.04 mg/100 g (Y1), 12.79 mg/100 g (Y2), 14.54 mg/100 g (Y3), and 167.89 mg/100 g (Y4). The final product was rich in glutamic acid (Glu), with a content of 158.7 mg/100 g, a characteristic likely attributable to the introduction of exogenous fermented condiments (e.g., doubanjiang, pickled peppers). Among all detected FAAs, alanine was present at the highest concentration, a finding consistent with studies on shredded beef, suggesting its role as a key taste-active compound in muscle foods (Chen, Yuan, et al., 2024). In most meat products, Glu (glutamic acid) was a primary contributor to the umami taste. It could interact synergistically with chloride ions to further enhance the umami flavor, similar to the effect of monosodium glutamate (Gunlu & Gunlu, 2014). It was also noteworthy that Asp (aspartic acid) and Glu exhibited significant synergistic interactions with 5′-ribonucleotides, particularly inosine 5′-monophosphate (IMP) and adenosine 5′-monophosphate (AMP) (Yang et al., 2022).

3.6. Analysis of nucleotides

Nucleotides contributed significantly to the development of meat-like flavors. Among them, IMP was the most abundant nucleotide in muscle tissue and a key precursor for flavor formation (Jayasena et al., 2013). As shown in Table 5, the content of umami nucleotides, particularly IMP, exhibited a significant increasing trend throughout the processing of Yu-Xiang shredded pork, culminating in a final concentration that was markedly higher (p < 0.01) than that in the raw meat. This accumulation was considered a primary contributor to the product's enhanced taste quality.

Table 5.

Changes in nucleotides during the processing of Yu-Xiang shredded pork.

Nucleotide (mg/100 g)	Different processing stages
	Y0	Y1	Y2	Y3	Y4
5’-IMP	218.15 ± 17.95c	277.4 ± 18bc	329.79 ± 22.32b	428.62 ± 60.4a	491.27 ± 27.48a
5’-AMP	34.51 ± 14.87c	45.55 ± 25.71bc	50.43 ± 21.15bc	89.78 ± 35.31ab	107.19 ± 18.6a
HX	38.76 ± 4.48ab	48.34 ± 8.16ab	49.36 ± 6.83a	44.92 ± 9.27ab	33.36 ± 9.73b
5’-GMP	15.24 ± 9.13d	21.04 ± 11.19d	123.57 ± 12.12c	209.14 ± 20.22b	271.32 ± 20.79a
IN	ND	2.6 ± 1.52cd	8.74 ± 2.34bc	13.06 ± 4.22b	24.53 ± 6.62a

Note: Different lowercase letters indicate significant difference (p < 0.05); Y0 (raw pork), Y1 (marinated pork), Y2 (stir-fried until color changes), Y3 (stir-fried with seasonings and ingredients), and Y4 (finished product with Yu-Xiang sauce). ND indicates not detected.

The rise in IMP occurred in two distinct phases. An initial increase was observed during marination, with levels rising from 218.15 mg/100 g at Y0 to 277.4 mg/100 g at Y1. This early elevation was likely attributed to the addition of sodium chloride (NaCl), which was known to accelerate the enzymatic degradation of adenosine triphosphate (ATP) into IMP (Hernández-Cázares et al., 2011). A subsequent and more substantial increase took place during thermal processing, with the IMP concentration rising progressively from 329.78 mg/100 g (Y2) to 428.62 mg/100 g (Y3) and reaching 491.27 mg/100 g (Y4). This suggested that continued heating further promoted the reactions leading to IMP formation. The underlying mechanism involved a multi-step enzymatic pathway for ATP degradation: ATP was hydrolyzed to ADP by ATPase, then converted to AMP by kinases, and subsequently deaminated to IMP. IMP could be further metabolized to inosine and hypoxanthine (Bi et al., 2021). Notably, the thermal degradation of IMP to inosine has also been reported to contribute to flavor enhancement (Jayasena et al., 2015), suggesting that both IMP accumulation and its thermal breakdown products influence the overall flavor profile.

External ingredients further enriched the umami characteristics of the dish. The fermented condiments added at stage Y4, namely red marinated peppers and doubanjiang, were rich in inosine monophosphate (IMP) and guanosine monophosphate (GMP) (Que et al., 2025). These ingredients served as a source of exogenous umami compounds. The flavor-active 5′-nucleotides, specifically 5’-AMP, 5’-IMP, and 5’-GMP, interacted synergistically with glutamate derivatives, significantly enhancing the characteristic umami and meaty flavors (Duan et al., 2020). In summary, the IMP pathway served as the predominant route for flavor development in Yu-Xiang shredded pork. The final nucleotide profile arose from two main sources. It combined endogenous enzymatic degradation, accelerated by marination and heating, with the direct incorporation of nucleotides from fermented condiments.

4. Conclusions

This study systematically examined the dynamic evolution of flavor-active compounds and the formation mechanism of characteristic flavor in Yu-Xiang shredded pork. Comprehensive analysis of volatile compounds and flavor precursors across key processing stages revealed significant changes in odor profiles, demonstrating that stir-frying initiates a cascade of chemical transformations. Lipid hydrolysis and oxidation were promoted by high-temperature stir-frying, with TBARS values significantly increasing from 0.06 to 0.41 mg MDA/kg and FFA content rising from 10.53 to 16.25 mg/100 g. Concurrently, proteolysis and thermal degradation led to substantial increases in FAAs and nucleotides, particularly IMP which rose from 218.15 to 491.27 mg/100 g. These taste-active compounds not only enhanced umami perception directly but also participated in Maillard reactions during thermal processing, generating heterocyclic compounds that complemented the lipid-derived volatiles. The incorporation of seasonings at critical stages further enhanced the sensory profile by substantially elevating the diversity and concentration of key aldehydes and alcohols. These processes generated 100 volatile compounds, including 19 key aroma-active components with OAVs >1. These findings provided mechanistic insights into flavor formation during stir-frying of traditional pork dishes, emphasizing the coordinated roles of lipid oxidation, volatile generation, and taste-active compound accumulation. The results offered theoretical guidance for quality optimization and process standardization in industrial production of traditional culinary dishes, while establishing a foundation for future research on flavor chemistry in traditional cooking methodologies.

CRediT authorship contribution statement

Zhuowen Wang: Writing – original draft, Methodology, Data curation, Conceptualization. Zhongqiang Wang: Formal analysis, Data curation. Jinxuan Cao: Project administration, Funding acquisition. Jinpeng Wang: Resources, Methodology, Data curation. Yuemei Zhang: Supervision. Wendi Teng: Conceptualization. Shuai Zhuang: Methodology. Ying Wang: Resources, Funding acquisition, Data curation.

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.

Data availability

Data will be made available on request.