Dynamic changes in quality and flavor compounds of pork tendons during puffing process

Pork tendon, a collagen- and elastin-rich tissue from the pig Achilles tendon region, is valued as a traditional tonic food with high protein, high calcium, and low fat. Fresh tendons are typically dehydrated for storage, but drying compacts the structure and hardens the surface, making rehydration and cooking difficult. Conventional swelling methods (oil-, alkali-, and water-based) each have drawbacks, including potential carcinogens in oil-swelling, nutrient loss and texture damage in alkali-swelling, and time consumption in water-swelling. Puffing, a rapid thermal process leveraging phase change and gas thermopressure, can create porous structures and drive Maillard reactions that enhance flavor. While puffing has been studied in plant-based materials, there is limited systematic research on proteinaceous matrices such as pork tendons, and their puffing behavior and mechanisms remain unclear. This study aims to evaluate how puffing temperature and time affect pork tendon quality (expansion ratio, bulk density, microstructure) and flavor (VOCs), and to identify optimal processing parameters that improve edible portability and flavor.

Background work indicates puffing techniques (extrusion, hot air, microwave) can alter structure and flavor via Maillard reactions and caramelization in various foods (e.g., ginseng, rice, potatoes). Prior studies documented browning with increased temperature/time in puffed ginseng and improvements in puffing characteristics of cereals and tubers. However, reports on puffed pork tendon flavor are lacking, and studies applying puffing to protein-rich matrices are limited. The present work addresses this gap by combining GC-IMS and GC-MS to comprehensively profile VOCs and link them to structural and physicochemical changes.

• Raw materials: Commercial dried pork tendons were washed, surface fat/muscle removed, oven-dried at 90 °C for 30 min, cut to ~5 cm pieces, then puffed at set temperatures (190, 210, 230, 250 °C) for defined times (reported ranges 1–7 min). Puffed samples were sealed for analyses. Experiments were performed in triplicate.
• Puffing performance: Expansion ratio (ER) determined as V2/V1×100%, where V1 is the volume of dried tendons and V2 the volume of puffed tendons measured via sand displacement. Bulk density (BD) measured as mass/volume (m/V2).
• Microstructure: Cryo-SEM. Samples were cryo-fractured, sublimed at −65 °C for 30 min, gold-coated, and imaged at −140 °C, 10.0 kV, 50× (selected samples also at 100×).
• Free amino acids (FAAs): Extracted with water and 5% sulfosalicylic acid, clarified by centrifugation, evaporated, reconstituted in sodium citrate buffer, filtered (0.45 µm), quantified using an amino acid analyzer.
• VOCs by GC-IMS: Headspace incubation (60 °C, 15 min, 500 rpm), 500 µL splitless injection via 85 °C syringe; MXT-WAX column at 60 °C, N2 carrier; IMS with tritium ionization, positive mode, 98 mm drift tube at 45 °C, drift gas 150 mL/min. VOCs identified via NIST/IMS databases; 2D spectra and fingerprints generated (VOCal software).
• VOCs by HS-SPME-GC-MS: 2 g sample in 20 mL vial, 50 °C incubation 15 min; DVB/CAR/PDMS fiber extraction 30 min; desorption at 250 °C for 5 min; DB-Wax column with temperature program (40 °C hold 5 min; 5 °C/min to 220 °C; 20 °C/min to 250 °C; hold 2.5 min); helium 1 mL/min; EI 70 eV; full scan (10 spec/s).
• Data analysis: PCA (MetaboAnalyst 5.0) on GC-IMS VOCs; OPLS-DA for differentiating VOCs with criteria FC>1, VIP>1, P<0.05; ROAV used to define key aroma contributors (ROAV≥1). Hierarchical clustering heatmaps for VOCs. Correlation analyses among ER, BD, FAAs, and VOCs with ROAV≥0.1. Statistics via one-way ANOVA (SPSS 18.0), significance at P<0.05.

• Puffing performance and microstructure: With increasing temperature and time, expansion ratio (ER) increased and bulk density (BD) decreased. Pores were scarce at 230–250 °C for 5 min; distinct, smooth, deeper pores formed at 230–250 °C for 6 min (square-like at 230 °C; elongated at 250 °C). At 7 min, pore collapse and structural damage appeared, indicating over-processing. Visual appearance shifted from pink to golden/yellow with whitening upon puffing; signs of baking/quality deterioration appeared at 250 °C for 5+ min. Overall, 230 °C for 6 min produced desirable porous structure and appearance.
• Free amino acids (FAAs): Seventeen FAAs detected (2 umami, 4 sweet, 11 bitter). Total FAAs (TEAA) showed decrease then increase with rising temperature at the same time, notably lower at 230 °C samples (230A/230B), suggesting participation in Maillard reactions/VOC formation. Bitter AAs predominated (notably Cys), with temperature-time effects modulating Cys content (e.g., at 6 min, higher temperature reduced Cys). Sweet AAs (Gly, Ala) accounted for ~50% of sweet AA pool, mitigating bitterness. Glu exceeded Asp, indicating stronger contribution to umami.
• GC-IMS VOCs: 87 VOC signals detected; 68 confidently characterized (26 aldehydes, 13 alcohols, 17 ketones, 7 esters, 3 acids, 2 furans). Differential and fingerprint analyses showed distinct VOC patterns across conditions; samples 210A and 250B clustered together, differing from 210B/230A/230B/250A.
• PCA (GC-IMS): PC1 and PC2 explained 74.9% and 14.5% of variance (cumulative 89.4%). Puffing time had a greater effect on flavor separation than temperature at a given temperature. Overlap indicated flavor similarity between 210A and 250B; 210B and 230A also overlapped.
• GC-MS VOCs: 376 VOCs detected; 109 with relative content ≥0.1 analyzed; 48 with ≥0.5 used for clustering. Sample-specific enrichments observed (e.g., n-octane and 1-octen-3-ol higher in 210A; 2-octenal, hexanal, 2-nonenal higher in 250B). Prolonged high-temperature puffing led to more oxidation-related VOCs.
• Key aroma contributors (ROAV): 22 VOCs had ROAV≥0.1; 16 had ROAV≥1. Across samples, 2-trans-4-trans-decadienal was the primary contributor in 210A, 210B, 230B, 250A, 250B; nonanal dominated in 230A. Nine VOCs with ROAV≥1 were common across all samples: nonanal, hexanal, 1-octen-3-ol, 2-ethylfuran, n-dodecane, 2-trans-4-trans-decadienal, trans,trans-2,4-nonadienal, (Z)-2-decenal, and trans-2-decenal. At 230 °C, contributions of key VOCs (except 2-trans-4-trans-decadienal) decreased with longer time, supporting 6 min as preferable.
• GC-IMS vs GC-MS: Thirteen VOCs overlapped between techniques; GC-MS detected many hydrocarbons not seen by GC-IMS due to ionization selectivity. Both methods consistently highlighted nonanal, 1-octen-3-ol, 2-amylfuran, and 2-ethylfuran as impactful, with low odor thresholds and high levels.
• Correlations: BD was highly negatively correlated with ER (P<0.01) and negatively with temperature (P<0.05). ER correlated positively with temperature (P<0.01) and time (P<0.05). Strong positive correlations among certain key VOCs (e.g., trans,trans-2,4-nonadienal with ethyl caproate; 1-octen-3-ol with 2-amylfuran).
• Optimal condition: Integrating ER, BD, microstructure, and VOC profiles, puffing at 230 °C for 6 min provided the best structure and flavor.

The study demonstrates that puffing conditions critically shape the physical structure and flavor chemistry of pork tendons. Higher temperatures and moderate times enhance expansion and reduce density, yielding a desirable porous microstructure that improves rehydration and texture. Concurrently, Maillard reaction and lipid oxidation-derived VOCs define the flavor profile; specific aldehydes (e.g., nonanal, hexanal, 2-trans-4-trans-decadienal), alcohols (1-octen-3-ol), and furans (2-ethylfuran, 2-amylfuran) emerged as key contributors. Multivariate analyses (PCA, clustering, OPLS-DA) confirmed distinct flavor fingerprints across conditions, with puffing time exerting strong within-temperature effects. Correlation analyses linked structural metrics (ER, BD) with thermal parameters and aroma-active compounds, supporting that 230 °C for 6 min optimally balances porosity and favorable VOC composition while avoiding over-processing (pore collapse, excessive oxidation) seen at longer times/higher temperatures. These findings address the research goal by identifying parameter settings that enhance both quality and flavor, thereby improving the portability and palatability of dried pork tendons.

By systematically varying puffing temperature and time and combining structural, compositional, and chemometric analyses, the study identifies 230 °C for 6 min as the optimal puffing condition for dried pork tendons. This setting maximizes expansion, lowers bulk density, forms a stable porous microstructure, and yields a favorable flavor profile dominated by key VOCs such as nonanal, 1-octen-3-ol, 2-ethylfuran, and 2-trans-4-trans-decadienal. The integrated GC-IMS and GC-MS approach provides complementary coverage of VOCs and clarifies the contributions of aroma-active compounds (ROAV). The work offers a practical processing strategy to improve edible portability and quality of protein-based matrices like pork tendons and provides theoretical references for further deep processing. Potential future work could extend to sensory validation, storage stability of puffed products, and exploration of alternative puffing techniques or other protein-rich matrices.