Results

Antifungal activity of composite films against spoilage Fungi

The antifungal activity of various film formulations was evaluated against four spoilage fungi isolated from S. rugosoannulata: F. pseudoanthophilum, A. niger, R. azygosporus, and T. purpureogenus (Fig. 1A). The CS film exhibited moderate inhibition, with rates ranging from 18.98% against A. niger to 55.32% against R. azygosporus. Incorporation of TiO₂ into CS (CS/TiO₂) resulted in marginal, non-significant improvements against selected strains, whereas nitrogen-doped TiO₂ (CS/N-TiO₂) composites achieved significantly enhanced inhibition against A. niger, R. azygosporus, and T. purpureogenus (P < 0.05). Such enhancement stems from nitrogen doping-induced bandgap narrowing, extending photocatalytic activity into the visible-light region (Salzano et al., 2026) and enabling reactive oxygen species generation under typical storage illumination, a synergistic interplay between CS-mediated membrane compromise and ROS-driven oxidative damage underpinning the superior performance of CS/N-TiO₂ relative to binary counterparts.

Fig. 1: Inhibition rates of different film materials against the rot-inducing fungus from S. rugosoannulata (A and B) and the mycelial growth and colony morphology (C) of the rot-inducing fungus on different film materials.Different capital letters indicate significant differences among different spoilage fungi for the same material, while different lowercase letters indicate significant differences among different materials for the same spoilage fungus (P < 0.05).

Further enhancement was achieved by incorporating four antimicrobial agents, natamycin (NT), nisin, ε-polylysine (ε-PL), and tea polyphenols (TP), into the CS/N-TiO₂ matrix (Fig. 1B). All four composites achieved high inhibition against F. pseudoanthophilum and A. niger, with no significant differences among NT, nisin, and ε-PL groups. Against R. azygosporus and T. purpureogenus, however, the CS/N-TiO₂/NT composite demonstrated significantly superior inhibition compared to nisin- or ε-PL-containing counterparts (P < 0.05), whereas TP incorporation consistently yielded the lowest activity across all strains. Notably, the inhibition rates of nisin-, ε-PL-, and TP-containing composites against A. niger and T. purpureogenus were lower than that of CS/N-TiO₂ alone (Fig. 1B and Fig. 1A). This observation may be attributed to electrostatic repulsion between the cationic polypeptides (nisin and ε-PL) and the positively charged chitosan matrix, which could cause these agents to localize on the film surface and undergo premature loss, while the strong antioxidant properties of tea polyphenols may scavenge reactive oxygen species generated by N-TiO₂, thereby interfering with the antifungal mechanism.

The superior antifungal performance of the CS/N-TiO₂/NT composite was further corroborated by morphological observations (Fig. 1C). Relative to CS films, the ternary composite markedly restricted mycelial development, manifesting as reduced colony diameters and sparse, suppressed hyphal growth across all four fungal species.

Physicochemical characterisation of CS/N-TiO₂/NT composite film

Structural characterisation

As shown in Fig. 2A, pristine TiO₂ exhibited characteristic diffraction peaks at 2θ values of 25.3°, 27.4°, 37.8°, 48.09°, 55.18°, 62.96°, and 68.81°, consistent with the anatase crystalline phase. The CS film displayed four diffraction peaks at 8.4°, 11.5°, 18.3°, and 22.9°, where the former two correspond to the crystalline structure of CS and the latter two to its amorphous and crystalline components, respectively (Madian & Mohamed, 2020). Upon formation of the CS/N-TiO₂ composite, the characteristic peaks of both components were retained with reduced intensities, indicating successful incorporation of N-TiO₂ into the CS matrix without disruption of the original crystal structures. The CS/N-TiO₂/NT composite exhibited a similar diffraction pattern to that of CS/N-TiO₂, suggesting molecular dispersion of NT within the composite matrix rather than formation of a separate crystalline phase.

Fig. 2: XRD pattern (A), FT-IR spectra (B), and flat (C, D) as well as cross-sectional (G, H) SEM images. Among them, C, D and G are images of the CS film, while E, F and H are images of the CS/N-TiO2/NT film.

FT-IR spectra (Fig. 2B) provided evidence of intermolecular interactions among the composite components. Compared with the CS film, the CS/N-TiO₂/NT composite exhibited a red-shift of the hydroxyl stretching vibration from 3431 cm−1 to 3411 cm−1 and a shift of the amide-I band from 1632 cm−1 to 1616 cm−1, indicative of hydrogen bonding formation between CS and N-TiO₂. The characteristic Ti—O stretching band narrowed and shifted from 668 cm−1 to 651 cm−1, further confirming the interaction between TiO₂ nanoparticles and the CS polymer backbone. A new absorption peak appeared at 1802 cm−1 in the CS/N-TiO₂/NT composite, attributable to the C Created by potrace 1.16, written by Peter Selinger 2001-2019 O stretching vibration characteristic of NT (Chakravartula et al., 2020), confirming successful NT incorporation.

Surface and cross-sectional morphologies were examined by SEM (Figs. 2C–2H). The CS film (Fig. 2C, D, G) exhibited a smooth, compact surface with a densely packed cross-sectional structure, reflecting the homogeneous nature of the pure CS matrix. In contrast, the CS/N-TiO₂/NT composite (Fig. 2E, F, H) displayed a noticeably rougher surface morphology with uniformly distributed small protrusions corresponding to N-TiO₂ and NT aggregates. Notably, no cracks or pores were observed on the surface or cross-section, indicating the formation of an integrated, defect-free composite structure.

Mechanical and barrier properties

As shown in Fig. 3A, the CS/N-TiO₂/NT composite exhibited a significant increase in thickness compared to the CS film (P < 0.05), with an average increment of 35.47%. The tensile strength (TS) of the composite increased by 132.20% relative to the CS film (Fig. 3C), a substantial enhancement attributable to the reinforcing effect of N-TiO₂ nanoparticles within the CS matrix. The formation of hydrogen bonds between N-TiO₂ and CS, confirmed by FT-IR analysis, establishes a stable three-dimensional polymer network that facilitates effective distribution of mechanical stress (Fig. 2 B). In contrast, the elongation at break (EAB) increased by only 10.69% (Fig. 3B), a change that did not reach statistical significance. This may be because the incorporation of N-TiO₂ enhances the film's strength while simultaneously restricting the mobility of chitosan chains, thereby limiting ductility. In addition, the incorporation of NT (natamycin) is known to reduce EAB (Sun et al., 2020). Notably, As illustrated in Fig. 3D, the water vapor permeability (WVP) of the CS/N-TiO₂/NT composite was significantly reduced (P < 0.05), a decrease attributed to two primary mechanisms: hydrogen bonding between N-TiO₂ and hydroxyl groups reduces the availability of free -OH groups in CS, thereby suppressing water molecule diffusion; additionally, the incorporation of N-TiO₂ nanoparticles results in a more compact film structure, prolonging the diffusion path for water vapor (Qian et al., 2011). Oxygen permeability (OP) exhibited a substantial reduction of 69.29% (Fig. 3E), indicating the composite's efficacy in restricting oxygen transmission, a favorable characteristic for suppressing oxidative browning and respiratory activity in stored mushrooms. In contrast, carbon dioxide permeability (CDP) showed no significant difference from the CS film (P ≥ 0.05) (Fig. 3F). This selective permeability pattern may facilitate the formation of a modified atmosphere around the packaged produce**.**

Fig. 3: Thickness (A), elongation at break (B), tensile strength (C), water vapor permeability (D), oxygen permeability (E), and carbon dioxide permeability (F) of CS and CS/N-TiO₂/NT films.

Preservation efficacy of CS/N-TiO₂/NT coating for S. rugosoannulata storage under variable temperatures

Effects on quality attributes

A progressive decline in L* values was observed across all storage conditions, indicating continuous browning of mushroom caps (Fig. 4A). At 25 °C, the control group exhibited rapid deterioration, with L* values dropping below 80 after 3 days, whereas the coated group maintained L* values above 80 throughout the same period (P < 0.05). The protective effect of the coating on color retention was more pronounced at 25 °C and 10 °C than at 4 °C, suggesting that the coating partially compensates for temperature-induced color deterioration. Weight loss increased progressively during storage, with higher temperatures accelerating moisture loss (Fig. 4B). The coated groups exhibited significantly lower weight loss rates than the control groups at both 4 °C and 10 °C (P < 0.05), with the most pronounced effect observed at 4 °C. This reduction reflects the effective moisture barrier properties of the composite film, as demonstrated by the reduced WVP values (Fig. 3 D). Decay rate, defined as the percentage of mushrooms showing visible spoilage, was substantially reduced by the coating treatment (Fig. 4C). At 4 °C on day 15, the control group exhibited a decay rate of 26.0%, while the coated group showed only 16.5%, representing a 36.54% reduction. This substantial decrease underscores the efficacy of the composite film in suppressing microbial spoilage during extended low-temperature storage.

Fig. 4: Changes in the quality parameters of S. rugosoannulata during storage. L*(A), weight loss rate (B), decay rate (C), respiratory rate (D), total colony count (E) and MDA (F). 4 °C CK, 10 °C CK and 25 °C CK: uncovered S. rugosoannulata stored at 4 °C, 10 °C and 25 °C, respectively. 4 °C Coating, 10 °C Coating and 25 °C Coating: coating-treated S. rugosoannulata with CS/N-TiO2/NT stored at 4 °C, 10 °C and 25 °C, respectively.

Effects on physiological and biochemical parameters

The composite film suppressed respiratory activity, delaying the respiratory peak from 9 days to 12 days at 4 °C and from 6 days to 9 days at 10 °C (Fig. 4D). This suppression is consistent with the reduced OP of the composite film (Fig. 3E), which limits oxygen availability for aerobic respiration. Total colony counts were significantly reduced by the coating treatment, particularly at 10 °C (Fig. 4E). On days 3, 6, and 9 of storage at 10 °C, the coated groups exhibited reductions of 20.93%, 8.35%, and 6.56%, respectively, compared to the control groups. The pronounced antimicrobial effect at 10 °C reflects the temperature-dependent activity of the composite components, with moderate temperatures potentially facilitating the release of antimicrobial agents or enhancing their interaction with microbial cells. MDA content, an indicator of lipid peroxidation, increased progressively during storage (Fig. 4F). At 4 °C, no significant difference was observed between control and coated groups, whereas at 10 °C and 25 °C, the coated groups exhibited significantly lower MDA levels (P < 0.05), indicating that the composite film effectively mitigates oxidative damage under moderate and high temperature conditions. SOD activity showed an initial increase followed by a decline at 4 °C and 10 °C, whereas a continuous decrease was observed at 25 °C (Fig. 5A). The coated group maintained significantly higher SOD activity than the control group across all three temperatures, with the most pronounced effect at 10 °C, where the control group exhibited a significant decline after 6 days while the coated group maintained elevated activity until 9 days. CAT activity exhibited a decreasing trend throughout storage, with the coated group consistently showing higher activity than the control group, particularly at 25 °C (Fig. 5B). POD activity was activated by the coating treatment across all storage temperatures without altering the overall trend (Fig. 5C). The most substantial enhancement was observed at 10 °C, suggesting a temperature-dependent activation mechanism. At 25 °C, POD activity peaked at day 6 before declining, consistent with the onset of accelerated senescence. PPO activity increased continuously throughout storage, with the coated group exhibiting significantly lower activity than the control group at 10 °C and 25 °C (P < 0.05) (Fig. 5D). This reduction is consistent with the limited oxygen availability within the coated samples (Fig. 3E), as PPO-mediated browning requires oxygen as a substrate (Qin et al., 2015).

Fig. 5: Changes in the enzyme activity of S. rugosoannulata during storage. SOD activity (A), CAT activity (B), POD activity (C) and PPO activity (D). 4 °C CK, 10 °C CK and 25 °C CK: uncovered S. rugosoannulata stored at 4 °C, 10 °C and 25 °C, respectively. 4 °C Coating, 10 °C Coating and 25 °C Coating: coating-treated S. rugosoannulata with CS/N-TiO2/NT stored at 4 °C, 10 °C and 25 °C, respectively.

Correlation analysis of quality and physiological parameters

To elucidate the relationship between quality deterioration and physiological responses, Pearson correlation analysis was conducted among key quality indicators (L* value, weight loss rate, decay rate) and physiological parameters (respiration rate, total colony count, MDA content, CAT, POD, PPO, and SOD activities). The results are presented separately for each storage temperature and coating condition in Fig. 6.

Fig. 6: The correlations between key quality indicators (L⁎, weight loss rate, and decay rate) and physiological parameters (respiration rate, total number of colonies, MDA content, CAT activity, POD activity, PPO activity, and SOD activity).

At 25 °C, no pronounced changes were observed between CK group and coating group (Fig. 6A, B). This finding suggests that, under high-temperature stress, the accelerated progression of senescence decouples quality attributes from the underlying metabolic and oxidative regulatory networks, with the coating treatment insufficient to restore such associations.

At 10 °C, distinct correlation patterns emerged between uncoated and coated samples (Fig. 6C, Fig. 6D). For uncoated mushrooms, quality parameters exhibited significant correlations with MDA content and the activities of CAT, POD, PPO, and SOD (P < 0.05), indicating a close association between oxidative stress, antioxidant enzymes, and quality maintenance at this moderate temperature. Notably, total colony count showed no significant correlation with quality in the uncoated group. Following coating application, the correlations between quality and POD, PPO, and SOD activities were substantially diminished, whereas total colony count became significantly correlated with quality (P < 0.05). This shift implies that, at 10 °C, the antimicrobial activity of the coating assumes a dominant role in quality preservation, effectively substituting for the oxidative regulatory mechanisms that governed quality in uncoated samples.

At 4 °C, the uncoated group (Fig. 6E, Fig. 6F) exhibited the most extensive correlation network, with quality parameters significantly correlated with respiration rate, total colony count, MDA content, and all four enzyme activities (P < 0.05). Such widespread correlations reflect sustained metabolic and oxidative activities during low-temperature storage. Coating treatment (Fig. 6F) markedly reduced these correlations, particularly for respiration rate, total colony count, and SOD activity, which no longer showed significant associations with quality. This selective decoupling indicates that the composite film mitigates quality deterioration through the inhibition of microbial growth, the suppression of respiratory activity, and the stabilization of SOD-mediated antioxidant defense.

Storage life prediction using cox proportional hazards model

To further evaluate the effect of the composite film, the storage life prediction of S. rugosoannulata under different storage temperatures was conducted. After the feature importance evaluation using the random forest, the key core factors were determined as L*, PPO activity, temperature, MDA content, and coating treatment. The hazard ratios (HRs) were calculated through Model (6).where h(t): The spoilage risk rate at time t; h₀(t): The baseline hazard function (for the untreated group at 4 °C); T_₁₀ °C, T_₂₅ °C: Dummy variables (with 4 °C as the reference, set to 1 for 10 °C and 25 °C respectively, and 0 otherwise); Coating: Coating treatment (1 indicates yes, 0 indicates no).

(6)ht=h0t∙expβ1PPO+β2L∗+β3T10°C+β4T25°C+β5Coating

The event was defined as follows: When the spoilage rate was ≥30%, it was considered a failure (e.g., if the spoilage rate of the CK group was 31% on the 6th day, it was marked as the occurrence of an event).

With decay rate ≥ 30% defined as the endpoint event, storage life was determined using the Cox proportional hazards model (7).

(7)Shelf Life=Ln0.3λ·expβ1PPO+β2L∗+β3T10°C+β4T25°C+β5Coating

The baseline hazard rate (λ) was calibrated using the untreated control group at 4 °C, with an empirically measured storage life of approximately 12 days (the decay rate of 30% observed between 13 and 15 days).

The predicted storage life indicated that, among the three storage temperatures, the 4 °C coating treatment demonstrated optimal effectiveness, with the potential to extend storage life by 33.33% to 38.46%. This percentage range was calculated based on the ratio of predicted storage life between coated and uncoated samples at 4 °C.