Design and characterization of a pH-responsive intelligent active film for monitoring and preserving fish freshness

Abstract

An active intelligent packaging film (COL/CS/ZnO/ATH) was developed based on a collagen (COL), chitosan (CS), and nano-ZnO composite matrix, incorporating anthocyanins (ATH) extracted from red cabbage as a natural pH-sensitive indicator. FT-IR analysis and molecular docking revealed that the film's structure is governed by hydrogen-bonding interactions among ATH, collagen, and chitosan. The COL/CS/ZnO/ATH film exhibited excellent barrier properties, and high thermal stability, along with notable antioxidant and antibacterial activities. Compared with COL/CS film, the tensile strength, DPPH radical free, and ABTS radical free scavenging rate of the COL/CS/ZnO/ATH film had increased by 11.18 MPa, 71.09%, and 73.54%, respectively. Additionally, the film exhibited sensitive responses to both pH and ammonia variations. Application of the film to fish preservation demonstrated its efficacy, extending the shelf life by approximately 2 days. In addition, it allowed freshness to be monitored via observable color shifts. For quantitative assessment of this relationship, a smartphone app was used to extract RGB values, which were then correlated with both TVB-N and TVC. A linear regression model was established to predict changes in fish freshness. This multifunctional intelligent packaging system offers a simple and efficient approach for real-time food quality monitoring.

1. Introduction

The spoilage of protein-rich foods such as seafood and meat during transportation and storage poses a significant challenge, primarily due to microbial contamination and adverse environmental conditions. This spoilage process leads to the release of harmful volatile compounds, particularly amines and thiols, which not only degrade food quality but also pose risks to consumer health (Duan, Li, Wang, et al., 2023). While conventional packaging provides basic protection, it cannot reflect the real-time status of food, thereby limiting its effectiveness in ensuring freshness. In contrast, real-time monitoring of food quality allows consumers to evaluate the current status of food and helps reduce the potential risk of foodborne illnesses (Yousefi, Su, Imani, et al., 2019). Given that the spoilage of protein-based foods often releases volatile alkaline compounds, leading to increased pH levels, considerable interest has been directed toward the advancement of pH-sensitive packaging materials (Wang, Zhang, & Zhang, 2022). Synthesized pH-responsive dyes may raise concerns regarding food safety and environmental pollution. Hence, pH-responsive indicators sourced from natural origins might offer advantages over chemically synthesized indicators (Ma, Wang, & Fan, 2026; Zhang, Liu, Xie, et al., 2024).

Anthocyanins (ATH) are water-soluble pigments, a class of bioactive phenolic compounds widely distributed in various plant organs, including roots, flowers, and fruits. Their distinct and reversible color changes over a broad pH range make them highly suitable for developing smart indicator films for intelligent food packaging (Chen, Wang, & Fan, 2026). However, despite their broad application, several studies have reported that anthocyanin-based films often present an initial blue, which undergoes only slight or indistinct changes during storage (Boonmahitthisud, Wongjampee, & Tanpichai, 2025; Hao, Kang, Guo, et al., 2023; Postolović, Radojević, & Stanić, 2026; Tang, Lu, Li, et al., 2025). This insufficient color variation hampers their effectiveness in accurately reflecting the gradual deterioration of food, particularly during the sub-fresh stage (Hashim, Elrasheid Tahir, Liu, et al., 2022; Wang, Xie, Tang, et al., 2023; Zheng, Tang, & Li, 2023). Studies suggest that shifting the initial color of the film closer to red can significantly improve the visibility and sensitivity of freshness indicators, particularly in response to sub-fresh or borderline spoilage stages (Li, Wang, Feng, et al., 2024). Therefore, modulation of the film's initial color may significantly affect the sensitivity and the accuracy of the resulting monitoring outcomes (Tang, He, & Fan, 2023; Li et al., 2024).

In recent years, polysaccharides and proteins and their derivatives have become ideal materials for the preparation of intelligent packaging films due to their environmental friendliness, non-toxicity, and biodegradability (Zheng, Liu, Yu, et al., 2022). Collagen, the predominant fibrous protein in mammals, confers excellent mechanical properties to collagen-based films due to its distinctive triple-helix structure (Tang, Zheng, Ran, et al., 2023). Collagen films prepared from a 10 mg/mL solution can achieve a mechanical strength of up to 120 MPa (Sionkowska, Michalska-Sionkowska, & Walczak, 2020), which is significantly higher than that of konjac glucomannan films (24.58 MPa, 10 mg/mL) (You, Wang, Zhou, et al., 2022). Good mechanical strength contributes to maintaining film integrity during storage, transportation, and retail handling. In addition to their outstanding mechanical strength, collagen films are also insoluble in water, which is advantageous for packaging applications. Among various sources, fish-derived collagen is considered safer than that obtained from traditional terrestrial animals. Swim bladders, a by-product of fish processing, are particularly rich in collagen and represent a high-quality raw material for collagen extraction. However, the absence of intrinsic antimicrobial activity and pH sensitivity in collagen severely limits its application in intelligent and active packaging. The linear cationic polysaccharide chitosan offers excellent film-forming ability, together with significant antibacterial and antioxidant properties, making it suitable for food packaging applications (Li et al., 2024). Incorporating collagen into chitosan matrices has been demonstrated to be an effective strategy for enhancing the mechanical performance and bioactivity of composite films. Simultaneously improving mechanical strength, along with antioxidant and antibacterial activities, is essential for extending shelf life and preserving the quality of meat products. Beyond polysaccharide-protein interactions, nanoparticles are commonly incorporated into polymer matrices to improve the functional performance of composite films. Nano-ZnO stands out for its remarkable broad-spectrum antimicrobial properties, leading to its extensive application in food processing (Shahid ul, Jaiswal, Butola, et al., 2023; Sun, Jiang, Wu, et al., 2020).

Therefore, this study developed pH-responsive active intelligent packaging films by combining swim bladder collagen (COL), chitosan (CS), and nano-ZnO with ATH extracted from red cabbage as a pH indicator. The structural, physicochemical, and functional properties of the film were systematically characterized, and its applicability for fish freshness monitoring and preservation was evaluated. By optimizing the initial film color and establishing quantitative correlations between color parameters and freshness indicators, including TVB-N and TVC, the film enabled continuous visual monitoring of fish quality from fresh to sub-fresh and spoiled stages. This study provides a practical strategy for constructing multifunctional biodegradable packaging films that integrate preservation, freshness indication, and early warning of quality deterioration.

2. Materials and methods

2.1. Materials

The grass carp and red cabbage were purchased from a local market in Ningbo, China. Collagen was obtained from grass carp swim bladders using the method described in our previous work (Dong & Dai, 2022). Acetic acid, chitosan (degree of deacetylation is 80–95%), and nano-ZnO (average particle size approximately 30 nm, purity ≥99%) were purchased from China National Pharmaceutical Group Chemical Reagent Co., Ltd. LB Broth and LB Agar were purchased from Qingdao Hope Bio-technology Co., Ltd.

2.2. Extraction of ATH and UV–Vis spectroscopy

Fresh red cabbage was chopped, and oven-dried at 50 °C. The dried cabbage was ground into a powder and sieved. ATH were extracted from the powder following the procedure of Li et al. (2024). The total anthocyanin content was determined using the pH differential method according to the method described by Ren, He, Lv, et al. (2023). After calculation, the total anthocyanin content of the extracted anthocyanins was 59.8 mg/g. The ATH extracts, dissolved in PBS buffers (pH 3–12), were analyzed by UV–visible spectroscopy over a wavelength range of 400–800 nm.

2.3. Preparation of COL/CS/ZnO/ATH films

“Chitosan (2% w/v) and collagen (2% w/v) were separately dissolved in 1% (v/v) acetic acid under magnetic stirring until completely solubilized. The resulting solutions were then mixed in a 1:1 (v/v) ratio to obtain a 1% (w/v) COL/CS blend solution. Subsequently, 0.5% (w/v) of ZnO and 10% (v/v) of glycerol were added and thoroughly stirred to ensure homogeneous dispersion. Add 8% ATH to obtain the COL/CS/ZnO/ATH composite film solution. Following pH adjustment to 2.0 with 0.5 mol/L HCl, 30 mL of the solution was cast into a 90 mm Petri dish and allowed to dry under ambient conditions (25 ± 1 °C, 50 ± 5% relative humidity). The obtained films were labeled as COL/CS, COL/CS/ZnO, and COL/CS/ZnO/ATH according to their compositions. The dried films were conditioned in a glass desiccator at 50 ± 2% relative humidity (RH) and 25 ± 1 °C before further characterization and testing.

2.4. Characterization of the COL/CS/ZnO/ATH films

2.4.1. Physical appearance, thickness and light transmittance properties

By capturing film images under natural light and using a colorimeter to extract color parameters of the films. Film thickness was measured at ten random locations per sample using a micrometer, from which the average was calculated and reported as the final thickness.

As outlined by Liu, Liu, Chen, et al. (2017), the optical characteristics of the films were assessed by recording their absorbance spectrum and transmittance over the wavelength range of 200 to 600 nm. The opacity of the film was calculated as the ratio of its absorbance at 600 nm to its thickness.

2.4.2. Mechanical properties

The mechanical performance of the films was assessed in terms of tensile strength (TS) and elongation at break (EAB) with a universal testing machine (UTM2102, New Sansi, Shenzhen, China). Before testing, the film samples were cut into rectangular strips with dimensions of 25 mm × 80 mm. Each sample was fixed between the grips of the testing machine, and the tensile speed was set at 50 mm/min. The average of six repeated measurements per sample is recorded.

$$\mathrm{TS}=\mathrm{F}/\mathrm{S}$$

$$\mathrm{EB}=∆\mathrm{L}/{\mathrm{L}}_0\times 100$$

where F is the maximum tensile force (N) at fracture, S is the cross-sectional area (mm²), ΔL is the elongation at break (mm), L₀ is the initial gauge length (mm), and L is the final length (mm).

2.4.3. Moisture content (MC), water solubility (WS) and water vapor permeability (WVP)

The moisture content (MC) and water solubility (WS) of the films were determined according to the method described by Niu, Xu, Han, et al. (2025). For MC determination, the film samples were dried at 105 °C to constant weight, and MC was calculated based on the weight loss before and after drying. For WS determination, the pre-dried film samples were immersed in distilled water for 24 h, and the undissolved residues were collected and dried again to constant weight. WS was calculated from the weight loss of the dried samples after immersion and re-drying.

The WVP of the films was measured according to Ren et al. (2023) with slight modifications. Briefly, film specimens (5 cm × 5 cm) were tightly sealed over the opening of cylindrical glass containers filled with 10 mL of distilled water to create a 100% relative humidity (RH) environment inside. The containers were then placed in a desiccator containing silica gel to maintain a lower RH outside the container, thereby establishing a humidity gradient across the film. The assemblies were weighed at 24 h intervals until the weight change became negligible, indicating equilibrium. WVP was then calculated using the standard equation:

$$\mathrm{WVP}=\mathrm{w}\times \mathrm{H}/\left(\mathrm{A}\times \mathrm{t}\times ∆\mathrm{P}\right)$$

where w, H, A, t, and ΔP represent the membrane mass (g), film thickness (m), permeation area (m²), permeation time (s), and vapor pressure difference (Pa), respectively.

2.4.4. Cross-sectional morphology characterization

The films were flash-frozen in liquid nitrogen, immediately fractured with tweezers to obtain a clean cross-section, sputter-coated with a thin gold layer, and then analyzed for cross-sectional microstructure using a SEM (Carl Zeiss, Germany) at an accelerating voltage of 15 kV under vacuum.

2.4.5. Thermal stability

The thermal stability of the films was investigated via a thermogravimetric analyzer (TG209F1, Netzsch, Germany), with heating applied from 30 to 800 °C at a rate of 10 °C/min in a N₂ atmosphere.

2.4.6. FTIR

FTIR spectroscopic analysis of the films was performed using a Nicolet iS10 spectrometer (Thermo Scientific, USA) configured with an ATR accessory. The spectrometer was configured to collect spectra from 400 to 4000 cm⁻¹ at a resolution of 4 cm⁻¹.

2.4.7. Responsiveness to ammonia

At room temperature, a film (1 cm × 1 cm) was placed and sealed over a glass dish containing an ammonia solution (5%, 20 mL) (Ren et al., 2023). Film coloration during ammonia exposure was documented photographically.

2.5. Antioxidant and antimicrobial activity

For the evaluation of antioxidant activity, a film sample (0.1 g) was immersed in 10 mL of distilled water for 1 h and then centrifuged at 3000 rpm for 10 min. The resulting supernatant was collected for subsequent analysis, with distilled water used as the blank control. The antioxidant activity was assessed by DPPH and ABTS radical scavenging assays according to the methods of Devita, Nurilmala, Lioe, et al. (2021), and Alemán, Pérez-Santín, Bordenave-Juchereau, et al. (2011). The antibacterial activity was assessed by the agar disk-diffusion method against Escherichia coli, Listeria monocytogenes, and Staphylococcus aureus according to Ren et al. (2023). Antibacterial efficacy was determined by measuring the diameter of the inhibition zones surrounding the filter paper discs. Filter paper discs without film-forming solution were used as the blank control.

2.6. Molecular docking

The following structures were retrieved for the docking study: type I collagen (PDB ID: 5K31) from the Protein Data Bank, chitosan (PubChem CID: 71853), and anthocyanin (PubChem CID: 443652) from the PubChem database (Guo, Huang, Huang, et al., 2022). Energy minimization was performed using the MMFF94 force field. Prior to docking, bound ions, water molecules, and ligands were removed from the protein structure. Molecular docking simulations of collagen–chitosan, collagen–anthocyanin, and chitosan–anthocyanin complexes were conducted using AutoDock Vina 1.5.7. The ligand–receptor interaction modes were analyzed using Discovery Studio 2019 Client, and visual interaction maps illustrating hydrogen bonds, hydrophobic interactions, and other molecular forces were generated.

2.7. Application in fish preservation

Fresh grass carp were stunned by percussion and then slaughtered. The fish were thoroughly rinsed with clean water, and the head, tail, scales, skin, and viscera were removed. The white dorsal muscle was collected and cut into fish pieces with a thickness of approximately 2 cm and a weight of approximately 30 g each. The samples were then randomly divided into five groups. The first group served as the control group without any treatment, whereas the second, third, fourth, and fifth groups were sealed and packaged with polyethylene (PE) film, COS/CS film, COL/CS/ZnO film, and COL/CS/ZnO/ATH film, respectively. All samples were stored at 4 ± 1 °C for 9 days, during which the physiological indicators of the fish fillets were determined.

Fish hardness was evaluated via Texture Profile Analysis (TPA) performed on a texture analyzer. The test parameters were set as follows: TMS 5 mm steel probe, 60% deformation, trigger force of 0.01 N. Water holding capacity (WHC) was determined according to Dong, Wang, Ma, et al. (2025), with slight modifications. TBARS analysis followed Zhang, Zhou, Zhang, et al. (2020). TVC and TVB-N content were quantified per the methods of Ran, Chen, Su, et al. (2022).

2.8. Freshness monitoring of fish samples

Fresh fish samples were packaged in sealed transparent Petri dishes (90 mm diameter) with a colorimetric film in the headspace, stored 9 days (4 °C). A standardized imaging protocol was followed to track film color changes, ensuring consistent camera positioning, lighting, and timing throughout storage.

2.9. Color image analysis based on smartphone applications

A quantitative approach was employed to assess fish freshness by analyzing the RGB of the colorimetric films. Throughout storage, digital images were captured using a smartphone camera. ColorColl application was used to extract R, G, and B values for subsequent analysis.

2.10. Establishment of a mathematical model for quantitative freshness monitoring of fish

A quantitative detection model for fish freshness was established based on the linear relationships between fish freshness indicators, TVC and, TVB-N, and the R, G, B color parameters extracted using the ColorColl application. A linear regression model was developed to enable prediction and evaluation of fish freshness.

The model performance was assessed via the coefficient of determination (R²) to quantify its goodness of fit and predictive accuracy. The prediction correlation coefficient (Rv²) was calculated using the following equation:

$${\mathrm{Rv}}^2=1-\frac{{\sum }_{\mathrm{i}=1}^{\mathrm{n}}{\left({\mathrm{Y}}_{\text{meas},\mathrm{i}}-{\mathrm{Y}}_{\text{pred},\mathrm{i}}\right)}^2}{{\sum }_{\mathrm{i}=1}^{\mathrm{n}}{\left({\mathrm{Y}}_{\text{meas},\mathrm{i}}-{\overline{\mathrm{Y}}}_{\text{meas}}\right)}^2}$$

2.11. Statistical analysis

Data are presented as mean ± SD and were analyzed using SPSS 21.0, with all graphs generated in OriginPro 2022. One-way ANOVA was conducted, and multiple comparison analysis was applied for significance testing among the groups.

3. Results and discussion

3.1. Color response of ATH

Distinct color changes and ultraviolet absorption spectral shifts of the extracted ATH were observed over a broad pH range of 2–13 (Fig. 1a). At pH 2–3, the solution was red and showed a single absorption peak near 530 nm, which is typical for the flavylium cation. When the pH was raised to 4–6, a purple color developed, corresponding to a redshift in the absorption peak to about 550 nm. As the pH elevated from 7 to 10, the solution color further transitioned from purple to blue, accompanied by a corresponding migration of the absorption peak to around 600 nm (Fig. 1b). When the pH exceeded 11, the solution turned from green to yellow, along with a noticeable decrease in the intensity of the UV absorption peak. The color variation of ATH in different pH solutions primarily stems from alterations in its chemical structure under varying pH conditions. Under strong acidic conditions, ATH presents as the red flavylium cation., while under weak acidic conditions, it appears as the colorless carbinol pseudo base. In weak alkaline conditions, it manifests as the blue quinoline base, and under strong alkaline conditions, it degrades into the yellow chalcone (Fig. 1c). Consequently, ATH exhibits distinct color variations across different pH levels, making it advantageous for use in pH-sensitive color indicators.

Fig. 1.

Color changes of anthocyanins at different pH levels (a), UV–Vis spectra (b), and structural transformations (c); surface appearance (d-f) and cross-sectional microstructure (g-i) of the composite films.

3.2. Physical properties of COL/CS/ZnO/ATH films

The COL/CS film exhibited L*, a*, and b* values of 87.04, −2.20, and 11.50, respectively (Table 1). No statistically significant alterations in the L, a, and b* values were induced by the introduction of nano-ZnO (P > 0.05). However, following the addition of ATH and pH adjustment, the COL/CS/ZnO/ATH film exhibited a red appearance (Fig. 1f), with L*, a*, and b* values of 38.40, 43.94, and 2.10, respectively. The ΔE of the COL/CS film was 12.54, which slightly decreased to 9.29 in the COL/CS/ZnO film, but increased markedly to 71.44 in the COL/CS/ZnO/ATH film. The ΔE value represents the degree of color difference between two samples, with higher ΔE values indicating more pronounced color variations (Zhao, Liu, Min, et al., 2025). This trend can be primarily attributed to the incorporation of anthocyanins. Under acidic conditions (pH 2), ATH predominantly exists in the form of red flavylium cations, which directly induce color changes and consequently exert a significant influence on the ΔE values. This observation is consistent with the findings of Xu, Liu, Tang, et al. (2025), who incorporated black rice anthocyanins into chitosan films and reported a significant increase in the ΔE value from 6.28 to 67.27.

Table 1.

L*, a*, b*, ΔE and opacity of the film.

Films	L*	a*	b*	ΔE	Opacity
COL/CS	87.04 ± 2.11 a	−2.20 ± 0.04 a	11.50 ± 0.98 a	12.54 ± 0.32 b	0.01 ± 0.00 c
COL/CS/ZnO	88.74 ± 1.86 a	−2.27 ± 0.03 a	8.70 ± 0.58 b	9.29 ± 0.53 c	0.43 ± 0.02 b
COL/CS/ZnO/ATH	38.40 ± 4.67 b	43.94 ± 4.17 b	2.10 ± 0.55 c	71.44 ± 0.99 a	4.47 ± 0.17 a

Note: Different superscript lowercase letters in the same column indicate significant differences among samples (P < 0.05).

The COL/CS film exhibited the lowest opacity, followed by the COL/CS/ZnO film with an opacity value of 0.43, indicating that both films possessed good transparency. The incorporation of ATH into the COL/CS/ZnO matrix resulted in a film opacity of 4.47. However, as shown in (Fig. f), the film still demonstrated satisfactory transparency, as the text beneath the film remained clearly visible. Film thickness is a critical factor that directly influences mechanical and barrier properties. The thicknesses of COL/CS, COL/CS/ZnO, and COL/CS/ZnO/ATH films were 0.103, 0.133, and 0.179 mm, respectively (Fig. 2a). The increased thickness of the film upon incorporating ZnO and ATH is primarily due to the elevated solid content in the casting solution.

Fig. 2.

Film properties: thickness (a), tensile strength (TS) and elongation at break (EB) (b), water vapor permeability (WVP) (c), moisture content (MC) and water solubility (WS) (d), UV–visible light transmittance (e), FTIR spectra (f), thermogravimetric analysis (TGA) (g), DTG curves (h).

The cross-sectional microstructure of the films was observed using SEM. The images (Fig. 1g–i) showed that all three films exhibited smooth, uniform textures free of visible cracks, voids, or agglomerations. This indicates good compatibility between collagen and chitosan, as well as satisfactory dispersion of ZnO and ATH within the film matrix. Such structural characteristics may result from the strong interactions among collagen, chitosan, and ATH, leading to a compact internal arrangement of the film components. This hypothesis will be further verified in subsequent studies using FTIR and molecular docking analysis. This observation is consistent with the cross-sectional microstructural results reported by Zheng et al. (2023) for CC/ZnO/ME films, as well as with those reported by Chen, Meng, Zhang, et al. (2025) for CS/CMC/BA/TP film.

3.3. Mechanical property

Fig. 2b shows that the TS and EB of the COL/CS film were measured at 73.14 MPa and 28.65%, respectively. The incorporation of ZnO notably enhanced the TS of the COL/CS/ZnO composite film to 95.34 MPa. These enhanced properties principally arise from strengthened intermolecular interactions in the composite matrix. Nano-ZnO served as a reinforcing filler within the film matrix. Under the acidic conditions of the film-forming solution, a portion of the ZnO dissociated into Zn²⁺ ions, leading to a co-existence of both ZnO particles and Zn²⁺ in the final film. The coordination interaction between Zn²⁺ and chitosan (CS) likely contributed to the formation of a more stable structure. In addition, ZnO functioned as a filler, and the synergistic interaction between Zn²⁺ and CS further enhanced the TS of the COL/CS/ZnO. After the incorporation of ATH, the TS of the COL/CS/ZnO/ATH decreased to 84.32 MPa. Nevertheless, this value was notably superior to those observed in other pH-responsive films derived from polysaccharides and water-soluble proteins (Lu, Zhou, Yu, et al., 2022; Ran et al., 2022; Sun et al., 2020). The EB of the COL/CS/ZnO film was 24.32%, and it increased to 27.45% after the addition of ATH, indicating an improvement in the film's flexibility. Our observation is in agreement with the work of Zheng et al. (2023). The improvement may be due to the enhanced stability of the CS–ATH combination under acidic conditions, which could strengthen the interactions among the components in the COL/CS/ZnO/ATH film. These interactions likely improved interfacial adhesion, thereby enhancing the EB of the composite film.

3.4. WVP

The WVP of the COL/CS film was approximately 3.31 × 10⁻¹² g/m·s·Pa. The WVP of the COL/CS/ZnO film decreased significantly to 2.52 × 10⁻¹² g/m·s·Pa (Fig. 2c). This diminution in WVP is primarily due to the void-filling effect of ZnO nanoparticles within the film matrix, coupled with the intermolecular interactions between COL/CS and Zn²⁺ ions, which collectively impede the transmission of water vapor. The introduction of anthocyanins led to a slight increase in the WVP of the COL/CS/ZnO/ATH film to 2.97 × 10⁻¹² g/m·s·Pa. This final value lies between that of the COL/CS and COL/CS/ZnO films, which can be attributed to the inherent hydrophilicity of anthocyanins. The enhanced hydrophilicity of the COL/CS/ZnO/ATH film facilitates the interaction between volatile amines and water molecules, promoting the formation of hydroxide ions (OH⁻) and generating a more alkaline microenvironment at the film surface. Thus, the film exhibited a marked improvement in its pH-responsive color sensitivity, consistent with the findings of Ma, Du, and Wang (2017).

3.5. MC and WH

The MC and WS of the COL/CS, COL/CS/ZnO, and COL/CS/ZnO/ATH films are shown in Fig. 2d. The incorporation of nano-ZnO had little effect on MC but reduced WS, which may be attributed to the reinforcing effect of ZnO and its interactions with hydroxyl, amino, and carbonyl groups in the COL/CS matrix. These interactions could enhance the compactness of the film network and restrict polymer chain dissolution in water. In contrast, the addition of ATH increased both MC and WS, probably due to the abundant hydrophilic phenolic hydroxyl groups in anthocyanins, which promoted water adsorption. Moreover, the water-soluble nature of ATH may have contributed to greater mass loss during the solubility test.

3.6. UV–vis transmittance properties

Food packaging films were required to possess the ability to block ultraviolet (UV) light, thereby reducing the loss of nutrients and spoilage of food due to light exposure. The COL/CS, COL/CS/ZnO, and COL/CS/ZnO/ATH films effectively blocked ultraviolet (UV) light within the range of 200–350 nm (approaching 0% transmittance), with a noticeable decrease in transmittance observed in the range of 350–400 nm. A notable reduction in transmittance was observed for all three films within the visible spectrum, particularly across the 550–622 nm range (Fig. 2e). ZnO, being an N-type semiconductor, can effectively absorb ultraviolet light through electron excitation processes between the conduction band and valence band, thereby serving as an efficient ultraviolet filter (Rochima, Octaviani, Azhary, et al., 2024).The addition of ATH further enhanced the light-blocking capability of COL/CS/ZnO/ATH films, with a significant decrease in transmittance observed within the range of 350–400 nm. This can be associated with the chromophores present in the aromatic rings of ATH, such as C C and C O, which are structures capable of effectively absorbing ultraviolet-visible light (Oun & Rhim, 2020; Qin, Yun, Xu, et al., 2021).

3.7. FTIR

The FTIR spectrum of the COL/CS film exhibited characteristic absorption bands at 3272.92 cm⁻¹ (O—H and N—H stretching), 2928.82 cm⁻¹ (C—H stretching), 1652.58 cm⁻¹ (C O stretching), 1550.51 cm⁻¹ (N—H bending), along with a peak at 1029.30 cm⁻¹ (C—O—C stretching) (Fig. 2f). After the incorporation of ZnO, the peak at 3272.92 cm⁻¹ shifted to 3295.35 cm⁻¹, and the peak at 1550.51 cm⁻¹ shifted to 1556.23 cm⁻¹. With the addition of ATH, the peak at 3272.92 cm⁻¹ shifted to 3285.91 cm⁻¹, and the peak at 2928.82 cm⁻¹ shifted to 2933.72 cm⁻¹, indicating the formation of hydrogen bonding and electrostatic interactions among ATH, chitosan, and collagen. Additionally, a slight decrease and shift in the peak intensity around 1633 cm⁻¹ were observed after ATH incorporation, which may be attributed to the interaction between aromatic rings (C=C) and glycosidic C O groups in ATH and the amide (—NH) groups of chitosan. These FTIR results confirm the successful incorporation of ATH into the film matrix via physical interactions, specifically hydrogen bonding and electrostatic forces, with chitosan and collagen, with no evidence of new covalent bond formation (Li et al., 2024).

3.8. Thermal stability

Thermogravimetric analysis (TGA) demonstrated that all films displayed consistent three-stage weight loss behavior over the temperature range of 30–600 °C (Fig. 2g). The TGA curves revealed three distinct degradation stages. The initial stage (30–120 °C) corresponded to the evaporation of loosely bound water. The subsequent stage (121–290 °C) was associated with the volatilization of glycerol, the onset of chitosan degradation, and the decomposition of ATH. Finally, the third stage (291–600 °C) represented the extensive thermal degradation of the polymer matrix, primarily involving chitosan and collagen. It is well-established that incorporating metal oxide nanoparticles into polymer matrices can improve the thermal stability of composite films (Chen, Luo, Cen, et al., 2022). The addition of ZnO to the COL/CS matrix delayed the onset of thermal degradation and improved the thermal stability of the pure COL/CS film. Specifically, the residual weight rose from 16.11% for the COL/CS film to 25.74% for the COL/CS/ZnO film. The subsequent incorporation of ATH further improved thermal stability, yielding a residual weight of 29.49% for the COL/CS/ZnO/ATH film. This enhancement can be explained by the formation of additional hydrogen bonds between ATH and the film matrix, which is consistent with the FTIR findings.

The DTG curves (Fig. 2h) further revealed that the main degradation peak occurred at approximately 200–280 °C. Compared with COL/CS, the COL/CS/ZnO and COL/CS/ZnO/ATH films showed lower maximum degradation rates, suggesting that ZnO and ATH reduced the rapid thermal decomposition of the film matrix. This may be attributed to the reinforcing effect of ZnO and the hydrogen-bonding interactions among ATH, ZnO, and the COL/CS matrix, which enhanced the compactness and thermal stability of the composite films.

3.9. pH and ammonia sensitivity analysis

Evaluation of the COL/CS/ZnO/ATH film's pH sensitivity revealed its functional utility. The film was observed to undergo a visible color shift (Fig. 3a), appearing light pink at pH 2 and evolving into a purplish-gray hue over a pH range of 3 to 7. Under alkaline conditions, the film exhibited a blue-green color due to the coexistence of quinonoid (blue) and chalcone (yellow) forms of anthocyanins. During the spoilage of meats, volatile nitrogen-containing compounds, especially amines, are commonly produced. To mimic the release of volatile nitrogenous metabolites from deteriorating protein-rich meat, we utilized ammonia gas as a model compound to assess the film's responsiveness. As shown in Fig. 3b, upon exposure to ammonia, the color of the COL/CS/ZnO/ATH film underwent a progressive color transition, evolving from red to purple and eventually to blue over time.

Fig. 3.

pH responsiveness (a) and ammonia sensitivity (b) of the COL/CS/ZnO/ATH film; DPPH and ABTS radical scavenging activities (c); antibacterial effects against Escherichia coli, Staphylococcus aureus, and Listeria monocytogenes (d and e).

This color change resulted from the reaction of ammonia with water, which generated NH₄⁺ and OH⁻ ions and thereby shifted the film's microenvironment to an alkaline state. This pH change triggers structural transformations of anthocyanins embedded in the film, resulting in noticeable color variation. These results demonstrate that the COL/CS/ZnO/ATH film exhibits pronounced color differences and a wide color gamut in the presence of alkaline volatile compounds, suggesting its promising potential as a visual indicator for monitoring the freshness or spoilage level of meats.

3.10. Antioxidant

During food storage, oxidative reactions may lead to undesirable changes in sensory and physicochemical properties. Therefore, packaging films with free radical scavenging capability play a crucial role in delaying food oxidation (Hashim, Tahir, Lui, et al., 2023). The DPPH radical scavenging activities of COL/CS, COL/CS/ZnO, and COL/CS/ZnO/ATH films were 10.37%, 11.88%, and 81.46%, respectively, while their ABTS radical scavenging rates were 12.45%, 13.00%, and 85.99%, respectively (Fig. 3c). The repeating units of chitosan contain hydroxyl and amino groups, and the free amino groups can react with free radicals to form stable macromolecular complexes, thereby scavenging free radicals (Zhao et al., 2025). The incorporation of ZnO increased the antioxidant activity of the films, which was mainly attributed to the electron density transfer promoted by ZnO nanoparticles (Shi, Ju, Wang, et al., 2025). The antioxidant activity observed in the COL/CS and COL/CS/ZnO films was primarily attributed to the radical scavenging ability of chitosan, collagen and ZnO in the film matrix. The potent radical scavenging activity of anthocyanins stems from the multiple phenolic hydroxyl groups present in their molecular structure (Wang, Yong, Gao, et al., 2019). The incorporation of anthocyanin extracts significantly enhanced the antioxidant capacity of the COL/CS/ZnO/ATH film compared to the COL/CS and COL/CS/ZnO films (P < 0.05). The high antioxidant activity of the COL/CS/ZnO/ATH film can be attributed to the strong antioxidant capacity of phenolic flavonoids present in anthocyanins, whose mechanisms include acting as reducing agents, hydrogen donors, singlet oxygen quenchers, and metal chelators (Zhao et al., 2025). The enhancement of antioxidant activity in bio-based polymer films through the incorporation of anthocyanin extracts has been widely documented in previous studies (Ren et al., 2023; Sun et al., 2020).

3.11. Antibacterial activities

Strong antibacterial activity of packaging films against foodborne pathogens is crucial for ensuring food safety and extending shelf life (Xu, Fang, Shi, et al., 2025). All three films demonstrated antibacterial efficacy, inhibiting the growth of both Gram-negative and Gram-positive bacteria as observed in Fig. 3d–e. The antibacterial activity of the COL/CS film originates from the inherent antimicrobial nature of chitosan. The inhibition zone diameters of the COL/CS against E. coli, S. aureus, and L. monocytogenes were 15.25 mm, 17.30 mm, and 17.25 mm, respectively. The intrinsic positive charge of chitosan enables it to interact with negatively charged bacterial cell membranes, thereby disrupting bacterial cell structures. This disruption allows chitosan to penetrate the cells and interfere with bacterial metabolism and gene expression (Zhao et al., 2025). The incorporation of nano-ZnO imparted superior antibacterial properties to the composite film, notably against Gram-negative bacteria. The antibacterial mechanism of the film is primarily ascribed to the inclusion of nano-ZnO (Sun et al., 2020). Zn²⁺ ions released from the matrix disrupt bacterial cell membranes, consequently inhibiting microbial growth (Lu et al., 2022). Rochima et al. (2024) reported that composite films containing nano-ZnO also exhibited good antibacterial activity. Furthermore, the presence of ATH contributed to an additional enhancement in antibacterial performance. The inhibition zone diameters of the COL/CS/ZnO/ATH film against E. coli, S. aureus, and L. monocytogenes reached 20.50 mm, 20.75 mm, and 22.50 mm, respectively. ATH contains abundant phenolic hydroxyl groups, which can inactivate proteins through hydrogen bonding, ultimately leading to bacterial death. In addition, ATH can significantly reduce bacterial activity by suppressing the tricarboxylic acid cycle and impairing energy transfer processes (Zhao et al., 2025). Given that ATH can inhibit bacterial growth by interacting with cellular proteins (Ren et al., 2023). These results demonstrate that the COL/CS/ZnO/ATH film possesses promising potential as an antimicrobial packaging material for extending food shelf life. The results of this study are consistent with those reported by Praseptiangga, Sesari, Rochima, et al. (2024), who found that the SRIC/FG/SiO₂–ZnO composite film exhibited significantly stronger inhibitory activity against Staphylococcus aureus than against Escherichia coli. This difference can be attributed to the structural characteristics of Gram-positive bacteria, whose outer cell envelope is more hydrophobic and porous, facilitating the adhesion and penetration of antibacterial agents and ultimately leading to cell death (Praseptiangga et al., 2024).

3.12. Molecular docking

Molecular docking was used in this study only as a supplementary theoretical tool to explore possible intermolecular interactions among collagen, chitosan, and anthocyanins. Fig. 4 illustrate the molecular docking results of collagen–chitosan, collagen–anthocyanin, and chitosan–anthocyanin complexes, respectively. The binding energies were calculated as −6.2, −10.9, and −6.0 kcal/mol, indicating strong binding stability between the ligands and receptors. In the figures, the active amino acid residues are represented by blue ribbons, with the names of successfully docked residues labeled; hydrogen bonds are depicted as green dashed lines. Among them, the collagen–anthocyanin complex exhibited the most negative binding energy, suggesting the strongest predicted binding affinity. This result was supported by the formation of seven hydrogen bonds with an average bond length of 3.05 Å, which can be attributed to the abundant phenolic hydroxyl groups in anthocyanins interacting with the amide, carbonyl, amino, and hydroxyl groups of collagen. For the collagen–chitosan complex, several hydrogen bonds were formed between the amino and hydroxyl groups of chitosan and the polar functional groups of collagen, indicating good molecular compatibility between the two biopolymers. The docking results reveal that multiple hydrogen bonds are formed between chitosan and collagen amino acid residues. In the chitosan–anthocyanin complex, five hydrogen bonds were identified between anthocyanins and chitosan. These interactions were mainly associated with the hydroxyl and amino groups of chitosan and the phenolic hydroxyl groups of anthocyanins. The docking results suggest that the dense and stable network structure of the composite film is mainly derived from synergistic hydrogen-bonding interactions among collagen, chitosan, and anthocyanins. This molecular-level interpretation is consistent with the FTIR results, in which the shifts of characteristic —OH, —NH, C O and —CH absorption bands indicated the formation of intermolecular hydrogen bonds.

Fig. 4.

Structural schematic of the COL/CS/ZnO/ATH film (a); molecular docking interactions between collagen and chitosan (b), collagen and anthocyanins (c), and chitosan and anthocyanins (d).

3.13. Preservation effect of films on fish

As shown in Fig. 5a, all fish samples exhibited a progressive decrease in hardness over time, which can be primarily attributed to autolysis and protein degradation induced by endogenous enzymes and psychrophilic bacteria. These processes ultimately lead to the breakdown of muscle fibers and connective tissue (Li, Liu, Li, et al., 2023). The control group exhibited a significant drop from 2.02 N to 1.13 N by day 5 (P < 0.05), while the COL/CS/ZnO/ATH film group showed the slowest decline, retaining 1.69 N, followed by the COL/CS/ZnO group (1.63 N). By day 9, the control group dropped to 0.48 N, whereas COL/CS, COL/CS/ZnO, and COL/CS/ZnO/ATH groups maintained 0.81 N, 0.96 N, and 1.16 N, respectively. These films effectively delayed hardness loss, likely by limiting oxygen exposure and microbial activity. The ATH component further enhanced microbial and enzymatic inhibition, improving texture retention.

Fig. 5.

Changes in hardness (a), water-holding capacity (WHC) (b), thiobarbituric acid reactive substances (TBARS) (c), total viable count (TVC) (d), and total volatile basic nitrogen (TVB-N) (e) during fish storage; correlation matrix of fish quality indicators (f); color changes of COL/CS/ZnO/ATH film during fish storage (g); correlation between TVB-N values and RGB values (h).

WHC also declined in all groups during storage (Fig. 5b), mainly due to protein denaturation. Initial WHC was 87.48%, which decreased to 70.14% in the control group and 71.34% in the PE group by day 9. In contrast, COL/CS/ZnO/ATH groups retained 76.92%. The incorporation of both ZnO and ATH led to a remarkable enhancement in WHC (P < 0.05), likely by reducing oxidative damage and moisture loss.

The lipid oxidation levels of fish meat were significantly elevated in all groups. By day 9, the MDA content in the control and PE groups increased to 1.25 and 1.19 mg MDA/kg, respectively, both exceeding the safety threshold of 1 mg/kg (Fig. 5c). In contrast, the values for the COL/CS, COL/CS/ZnO, and COL/CS/ZnO/ATH groups were 0.99, 0.86, and 0.81 mg/kg, respectively. These results indicate that COL/CS/ZnO/ATH effectively inhibit lipid oxidation in fish meat.

In contrast to the film-packaged groups, the control group showed a TVC exceeding 6 log CFU/g on day 5, indicating a loss of freshness. Subsequently, it reached 9.79 log CFU/g by the end of the storage period (day 9). In contrast, the TVC values in the COL/CS, COL/CS/ZnO, and COL/CS/ZnO/ATH film-treated groups were 8.30, 7.06, and 6.55 log CFU/g, respectively (Fig. 5d). These results indicate that COL/CS/ZnO/ATH film packaging can effectively control microbial proliferation. The synergistic antimicrobial effects of chitosan, ZnO, and ATH effectively delayed microbial growth. In addition, the oxygen barrier properties of the composite film further inhibited the growth of aerobic microorganisms (Chen et al., 2025).

A TVB-N value of 15 mg/100 g was adopted as the spoilage threshold. On day 5, the TVB-N contents in the control and PE groups reached 17.84 and 16.13 mg/100 g, respectively, both exceeding the safety limit, indicating that these samples had reached the end of their acceptable shelf life. In contrast, the COL/CS/ZnO/ATH group showed a significantly lower TVB-N value of 9.12 mg/100 g on day 5. By day 7, the TVB-N content in the COL/CS/ZnO/ATH group was 13.69 mg/100 g, which remained below the spoilage threshold and indicated a sub-fresh but still acceptable freshness level. Based on the TVB-N criterion, the shelf life of the control and PE-packaged samples was approximately 5 days, whereas the COL/CS/ZnO/ATH film-treated samples maintained acceptable freshness for at least 7 days. Therefore, the application of the COL/CS/ZnO/ATH composite film extended the shelf life of grass carp by approximately 2 days. This preservation effect may be attributed to the combined antioxidant and antibacterial properties of ATH and nano-ZnO, which inhibited microbial growth and slowed protein and lipid oxidation (Wang, Liu, Xiao, et al., 2025). These findings are consistent with the results reported by Jebel, Roufegarinejad, Alizadeh, et al. (2025), who found that trout fillets packaged with PVA/FG films enriched with anthocyanin extract from lychee peels showed significantly reduced protein hydroperoxide and TVB-N levels after 5 days of refrigerated storage. In conclusion, the COL/CS/ZnO/ATH composite film effectively maintained the physicochemical and microbial quality of grass carp during refrigerated storage, thereby extending its shelf life from approximately 5 days to at least 7 days.

3.14. Application of the films

The freshness of fish was monitored using the COL/CS/ZnO/ATH film to evaluate its effectiveness for real-time monitoring. The film color demonstrates a clear correlation with fish freshness (Fig. 5g). The COL/CS/ZnO/ATH film was red when the fish was fresh, turned light pink in the sub-fresh stage, changed to purple-red with bluish edges upon initial spoilage, and finally shifted to blue when the fish was spoiled. The color change was attributed to volatile amines generated during fish spoilage. During storage, microbial metabolism associated with fish deterioration produces various volatile compounds that are gradually released into the headspace of the storage container. Over time, the COL/CS/ZnO/ATH film absorbs these volatile gases, creating an alkaline microenvironment that induces deprotonation of the anthocyanin structure. This structural transformation leads to a visible color change in the film, thereby enabling color-based monitoring of fish freshness (Qin, Wang, Tang, et al., 2024). This result demonstrates the capability of the COL/CS/ZnO/ATH for monitoring of fish freshness.

The prevalent method for evaluating packaged food quality is currently based on the visual observation of color transitions in indicator films. Nevertheless, this approach is limited by the subjectivity of human color perception, which varies among individuals and complicates the accurate detection of subtle color shifts. Although laboratory-grade colorimeters can precisely measure color parameters such as L*, a*, and b*, their use is associated with certain inconveniences and limitations. The RGB color model, widely employed in digital image color measurement, offers a potential solution. During storage, proteins in fish are continuously degraded under the action of enzymes and bacteria, leading to an increase in TVB-N content, accompanied by corresponding changes in the RGB values of the film. Leveraging this model, the ColorColl application for smartphones acts as a simple digital image processing tool. By photographing the indicator film, it quickly extracts the R, G, and B values, thereby enhancing the accuracy and practicality of indicator films for monitoring food freshness. A relationship between the indicator label's color information (RGB) and the freshness indicator TVB-N was established. The freshness level of fish was evaluated based on the variation ranges of the R, G, and B values. Regions I–IV correspond to the fish freshness states in Fig. 5h. The changes in RGB values in Fig. 5h are significantly correlated with TVB-N content: as TVB-N gradually accumulates, the R value decreases progressively, while the G and B values increase. When TVB-N exceed 15 /100 g, it indicates that the fish has entered the spoilage stage. Despite a minor increase in TVB-N (5 mg/100 g), pronounced changes were observed in the RGB: the R value decreased from 147 to 103, while the G and B values increased from 84 to 106 and from 129 to 139, respectively. This phenomenon indicates that the RGB of the indicator label can accurately reflect changes in TVB-N content. To provide a quantitative reference for practical application, the RGB value ranges corresponding to different freshness stages are summarized in Table 2. By using the ColorColl application to extract the RGB color information from the indicator label, consumers can obtain a convenient, intuitive, and effective method to accurately assess the freshness of fish.

Table 2.

RGB value ranges of COL/CS/ZnO/ATH films corresponding to different fish freshness stages.

R	G	B	Freshness
170–200	20–65	80–110	Fresh
160–180	70–100	120–140	Sub-fresh
30–150	90–120	120–160	Spoiled

The correlations among fish quality indicators for the COL/CS/ZnO/ATH film treatment are displayed in a heatmap (Fig. 5f), where each cell encodes the correlation coefficient, with red/blue color and its intensity to represent its direction and magnitude, respectively. Water-holding capacity and hardness exhibited negative correlations (blue) with TBARS, TVC, and TVB-N, while TBARS, TVC, and TVB-N showed positive correlations (red). A strong negative correlation was found between WHC and TVB-N (r = −0.893). Conversely, TVC and TBARS showed a very strong positive correlation (r = 0.993). These results demonstrate strong correlations among the indicators, enabling a reliable assessment of fish meat freshness.

3.15. Mathematical model for quantitative monitoring of fish freshness

To enable quantitative monitoring of fish freshness, linear regression models were established based on 12 samples (n = 12) to evaluate the relationships between the color indices of the COL/CS/ZnO/ATH packaging film and fish quality indicators, including TVC and TVB-N. The high coefficients of determination obtained for TVB-N and TVC (R² = 0.985 and 0.984, respectively; Table 3) suggested strong correlations between the film color response and the corresponding freshness indices within the experimental range. To further validate the predictive capability of the models, linear correlations between the measured and predicted values were established. Good agreement was observed between the predicted and measured results, with Rv² of 0.977 for TVB-N and 0.983 for TVC. These results indicate that the color parameters of the COL/CS/ZnO/ATH film have potential as useful indicators for estimating fish freshness under the present experimental conditions.

Table 3.

Mathematical model for quantitative evaluation of fish freshness.

Fish quality indicators	R2	Rv2	multiple regression equations
TVB-N	0.985	0.977	TVB-N = − 0.154R − 0.001G + 0.115B + 23.428
TVC	0.984	0.983	TVC = − 0.037R − 0.005G + 0.038B + 7.331

4. Conclusion

A novel multifunctional film based on COL/CS/ZnO/ATH was fabricated, which integrates intelligent pH-responsive colorimetry and antimicrobial properties for real-time monitoring of fish spoilage. FT-IR and molecular docking revealed that a dense network structure formed in the film, driven by intermolecular hydrogen bonding and electrostatic interactions between chitosan and collagen. The stable embedding of anthocyanins within the film matrix was achieved via the formation of hydrogen bonds with chitosan and collagen. The microstructure of the film appeared uniform, indicating good compatibility among the film-forming components. The COL/CS/ZnO/ATH film exhibited an exceptional combination of mechanical strength, low WVP, potent antioxidant and antibacterial activities, and effective UV-blocking capability. These integrated properties thus establish the COL/CS/ZnO/ATH film as a dual-function intelligent/active packaging material that not only extends shelf life through antioxidant and antimicrobial activities but also enables real-time, visual freshness monitoring via its naked-eye-detectable sensitivity to pH and ammonia. This study was primarily validated using a single fish species under laboratory conditions, and its stability and applicability in different meat systems and real circulation environments remain to be further investigated. Future work will focus on evaluating the monitoring performance of the film in various meat products under realistic storage and transportation conditions, as well as enhancing its stability and practical feasibility through formulation and structural optimization, thereby promoting its application in the field of intelligent packaging.

CRediT authorship contribution statement

Ye Dong: Writing – original draft, Supervision, Methodology, Funding acquisition, Conceptualization. Lvxuan Zheng: Writing – original draft, Data curation. Ying Yang: Investigation, Data curation. Yuhao Hu: Writing – review & editing. Xina Yu: Writing – review & editing, Supervision. Jie Wang: Software, Methodology.

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.