Poly(lactic acid)/ZIF-67 colorimetric headspace indicators for monitoring chicken breast freshness

Abstract

Poly(lactic acid) (PLA)/zeolitic imidazolate framework-67 (ZIF-67) composite films were developed as intelligent indicators for monitoring chicken freshness through volatile amine detection. ZIF-67 particles, with an average particle size of 413 ± 104 nm and a BET surface area of 2163 m² g⁻¹, were incorporated into PLA at concentrations of 0.3–0.9 wt%. Film properties depended on ZIF-67 loading, with improved mechanical performance and reduced water vapor and apparent oxygen permeability at the intermediate concentration. The films exhibited rapid colorimetric responses to ammonia, dimethylamine, and trimethylamine, which are volatile basic nitrogen compounds formed during protein degradation. A perceptible color change (ΔE > 5) occurred within minutes of exposure. When applied as a headspace indicator to chicken stored at 23 °C, the film containing 0.6 wt% ZIF-67 showed a clear color transition corresponding to the increase in total volatile basic nitrogen, which exceeded the spoilage threshold of 30 mg N/100 g at approximately 36 h. These findings support PLA/ZIF0.6 as a non-contact headspace indicator label attached to the lid, not as a food-contact packaging film for monitoring poultry freshness.

Poly(lactic acid) (PLA)/zeolitic imidazolate framework-67 (ZIF-67) composite films were developed as intelligent indicators for monitoring chicken freshness through volatile amine detection.

1. Introduction

The extensive use of non-biodegradable plastic packaging has raised significant environmental concerns. Plant-based biodegradable materials and sustainable alternatives have been proposed to address the global crisis of plastic pollution. Biopolymers, such as starch, pectin, pullulan, zein/chitosan, and gelatin, generally exhibit weaker mechanical and barrier properties than conventional synthetic packaging polymers.^1–4 PLA is a bio-based polyester produced from renewable feedstocks and has been widely investigated for food packaging applications because of its potential biodegradability under suitable conditions. However, it is brittle and has low flexibility, which restricts its application. Additionally, submicron particles such as metal–organic frameworks (MOFs) can reinforce the PLA matrix, improving its thermal and barrier properties. These submicron materials can also introduce added properties to packaging films, including UV protection, antimicrobial properties, and colorimetric indicators.^5,6

The growing demand for affordable, protein-rich foods has emphasized the need for reliable and rapid freshness monitoring, particularly for chicken meat. It is crucial to ensure product quality and safety during storage and distribution to protect consumer health in the poultry industry. Chicken meat is highly perishable because of its rich protein composition. This spoilage is primarily driven by microbial growth, leading to the enzymatic breakdown of proteins into free amino acids, which subsequently undergo deamination and decarboxylation, forming volatile nitrogenous compounds collectively referred to as TVB-N, including ammonia, dimethylamine (DMA), and trimethylamine (TMA).⁷

TVB-N analysis is widely used to assess meat freshness; however, its time-consuming and laboratory-dependent nature limits its suitability for real-time, on-site monitoring. Consequently, intelligent packaging systems that provide visible freshness information have attracted increasing attention. ZIF-67 is a cobalt-based MOF characterized by a porous structure, high specific surface area, straightforward synthesis, and a visible response to ammonia exposure.⁵ The interaction of ammonia with the Co-based coordination environment can alter the optical properties of the framework, producing a color transition from deep purple to pale purple and eventually to brownish-yellow.⁶ These characteristics make ZIF-67 a promising sensing component for detecting volatile basic compounds associated with food spoilage. Nevertheless, its incorporation into PLA at low concentrations as a dual-function filler that simultaneously modifies film performance and provides a colorimetric response to spoilage-related volatile amines has not been sufficiently investigated, particularly for monitoring chicken freshness.

Accordingly, the novelty of this study lies in the integration of low amounts of ZIF-67 into a PLA matrix and the evaluation of how this incorporation affects the structural, mechanical, and barrier properties of the films while enabling a visible response to spoilage-related volatile amines. ZIF-67 submicron particles were synthesized and characterized by XRD, FTIR, SEM, TEM, and BET analyses to confirm their crystalline structure, chemical features, morphology, surface area, and porosity. PLA/ZIF-67 composite films containing different ZIF-67 loadings were then prepared and evaluated in terms of their structural, morphological, mechanical, barrier, thermal degradation, and colorimetric properties. Their time-dependent responses to ammonia, DMA, and TMA were comparatively examined. Finally, the selected formulation was applied as a non-contact indicator for monitoring chicken breast freshness under accelerated storage conditions, and its color response was related to the increase in TVB-N content.

2. Materials and methods

2.1. Materials

Polylactic acid (PLA, M_W ∼ 80 000) was obtained from Macklin (Shanghai, China). 2-Methylimidazole, cobalt nitrate hexahydrate (Co(NO₃)₂·6H₂O), and polyethylene glycol (PEG400) were supplied by Sigma-Aldrich (Germany). Methanol and ethanol (99% purity) were obtained from Isolab Co. (Turkey). Dichloromethane (DCM), DMA, TMA, and ammonia solution (25%) were obtained from Sigma-Aldrich (Germany).

2.2. ZIF-67 synthesis

Co(NO₃)₂·6H₂O (2.91 g) and 2-methylimidazole (3.3 g) were separately solubilized in methanol (80 mL). The cobalt solution was immediately mixed with the 2-methylimidazole solution and stirred overnight at room temperature (300 rpm). The resulting purple suspension was pelleted by centrifugation at 9500 rpm, and the collected pellet was rinsed three times using methanol (50 mL each). The final purple precipitate was dried in a vacuum oven at 80 °C for 24 h.

2.3. Film preparation

PLA (10 g) was dissolved in 100 mL of DCM. PEG400 was added at 20 wt% relative to the PLA mass (2 g per 10 g PLA), while ZIF-67 was incorporated at 0.3, 0.6, or 0.9 wt% relative to the PLA mass. The mixtures were stirred at room temperature for 18 h to obtain homogeneous film-forming solutions. Subsequently, 10 g of the resulting solution was cast onto circular glass dishes (6 cm in diameter) and dried at room temperature for 1 h. The obtained PLA/ZIF-67 films were then conditioned at 25 °C and 50% relative humidity for 48 h. The films containing 0.3, 0.6, and 0.9 wt% ZIF-67 (based on PLA) were designated as PLA/ZIF0.3, PLA/ZIF0.6, and PLA/ZIF0.9, respectively.

2.4. Structural and functional evaluation of ZIF-67 and PLA/ZIF-67 films

2.4.1. SEM and TEM

The surface structure of ZIF-67, which was dispersed in methanol, as well as the PLA/ZIF-67 films, was observed via a field emission scanning electron microscope (FESEM, Thermo Scientific, Chicago). Additionally, microscopic images of ZIF-67 were obtained using a FEI/Tecnai G2 Spirit Biotwin high-contrast TEM (CTEM) at 120 kV. The diameter of the ZIF-67 submicron particles was measured using ImageJ software.

2.4.2. BET, XRD, TGA, and FTIR

N₂ adsorption–desorption isotherms of ZIF-67 were obtained at 77.35 K using an Autosorb 6100 FKM XR high-vacuum physisorption analyzer (Anton Paar QuantaTec, Boynton Beach, FL, USA). Prior to analysis, the sample was degassed under vacuum at 120 °C for 24 h. The specific surface area was determined using the multipoint Brunauer–Emmett–Teller (BET) method. The pore size distribution was calculated using the slit-pore density functional theory (SF-DFT) model, while the micropore volume and external surface area were determined using the t-plot method. The crystalline structures of ZIF-67 and the PLA-based composite films were characterized using an Ultima IV X-ray diffractometer (Rigaku, Tokyo, Japan) operated with Cu radiation at 40 kV and 30 mA. The diffraction patterns were recorded over a 2θ range of 5–50° in continuous scanning mode, with a step size of 0.02° and a scanning rate of 2° min⁻¹. Thermal degradation behavior was evaluated using a Discovery SDT 650 simultaneous thermal analyzer (TA Instruments, New Castle, DE, USA). Approximately 5.0 ± 0.3 mg of each sample was placed in a sample pan and heated from room temperature to 700 °C at a heating rate of 20 °C min⁻¹ under a continuous nitrogen atmosphere. In addition, the interactions between ZIF-67 particles and the PLA matrix were investigated using a Fourier transform infrared spectrometer (Shimadzu, Kyoto, Japan) in transmission mode. The spectra were recorded over the range of 4000–500 cm⁻¹ at a resolution of 4 cm⁻¹ using 64 scans per spectrum.

2.4.3. Mechanical properties

Tensile testing of the films was performed using a Zwick/Roell Z250 Universal Testing Machine. Dumbbell-shaped specimens were prepared using a cutting die and mounted with an initial grip separation of 20 mm. Before testing, the thickness of each specimen was measured at five points along the narrow gauge section, and the mean thickness was used to calculate the initial cross-sectional area. A preload of 0.1 N was applied, and the films were tested at a crosshead speed of 10 mm min⁻¹ until failure. Tensile strength (TS) was calculated by dividing the maximum force by the initial cross-sectional area, determined as the product of the specimen width and mean thickness. Elongation at break (EAB) was calculated as the percentage increase in specimen length at rupture relative to the initial gauge length.

2.4.4. Evaluation of water vapor and oxygen barrier characteristics

Water vapor and apparent oxygen permeabilities (WVP and OP) of the films were evaluated using standard procedures previously reported by McHugh et al. (1993) and Zhang et al. (2019) under controlled temperature and relative humidity conditions.^8,9

Before testing, the thickness of each film was measured at five randomly selected points using a digital film thickness gauge (TQC Sheen, model SP1570, Netherlands), and the mean thickness was used in subsequent calculations. For the WVP analysis, circular film samples were sealed over permeability cups containing distilled water to create a humidity gradient. These cups were afterwards placed in a desiccator containing silica gel for drying. Water vapor transmission rate was calculated by dividing the slope of the linear mass–loss curve by the effective exposed film area. WVP was then obtained by multiplying the transmission rate by the mean film thickness and dividing the result by the water vapor partial-pressure difference across the film.

Apparent oxygen permeability was measured by the deoxidizer absorption method of Zhang et al. (2019). The films, which were previously cut and conditioned at 75% RH, were sealed onto cups containing 3 g of an oxygen absorber (deoxidizer) and stored at 75% relative humidity.¹⁰ For the apparent OP measurement, the mass gain of the cups was determined after 48 h (eqn (1)).

Apparent OP = (Δm × d)/(A × t × P) 1

where Δm is the measured mass gain of the cup (kg), d is the mean film thickness (m), A is the exposed area (m²), t is the exposure time (s), and P is the oxygen partial-pressure difference across the film (Pa). Values are expressed as kg m⁻¹ s⁻¹ Pa⁻¹. No water-vapor blank (for example under nitrogen) was applied. The values are therefore reported as apparent oxygen permeability and are used to compare formulations measured under identical gravimetric conditions; they are not equivalent to coulometric oxygen permeability.

2.4.5. Colorimetric analysis

2.4.5.1. Response to ammonia, DMA, and TMA

The ammonia-induced colorimetric response of the PLA/ZIF composite films was evaluated by exposing 2 × 2 cm² film specimens to ammonia vapor at 23 °C. The films were attached to cellophane with the sensing surface facing downward and positioned above 20 mL of 0.8 M ammonia solution. Images were captured every 2 min for 20 min using an LG G8 smartphone camera with the flash activated under fixed imaging conditions. The entire cropped image was processed without defining a separate region of interest. The CIELAB color coordinates (L*, a*, and b*) were extracted using a custom Python script run in the Thonny integrated development environment. The total color difference at each exposure time was calculated relative to the initial film color according to ΔE = [(ΔL)² + (Δa*)² + (Δb*)²]^1/2, and the resulting ΔE values were plotted as a function of exposure time.¹¹ After selecting the formulation with the most favorable physicochemical properties and colorimetric response, its comparative responses to DMA and TMA vapors were evaluated under the same conditions using 20 mL of 0.407 M DMA or 0.117 M TMA. These liquid-phase concentrations were chosen from literature Henry solubilities so that the estimated equilibrium partial pressures matched that of 0.8 M ammonia, rather than matching the liquid-phase molarity.

2.4.6. Application of the film label for monitoring chicken breast freshness

Fresh chicken breast was purchased from a local supermarket in Ankara, Türkiye. Portions of 10 g were placed in Petri dishes. PLA/ZIF0.6 indicator labels measuring 1.5 × 1.5 cm² were attached to the inner surfaces of the Petri dish lids and positioned in the headspace above the chicken samples without direct food contact. The labels were secured with adhesive tape, and the dishes were sealed with Parafilm. The samples were stored at 23 °C for up to 60 h. Images of the indicator labels were captured at 12 h intervals, and ΔE values were determined using the Python-based image analysis procedure described in Section 2.4.5.1. Chicken freshness was assessed at the same intervals by determining the TVB-N content using the micro-Conway method described in our previous study.¹¹

2.5. Statistical analysis

All measurements were performed in triplicate, and the data are presented as mean ± standard deviation. Differences among the formulations were evaluated by one-way ANOVA followed by Tukey's multiple-comparison test. Statistical significance was accepted at p < 0.05.

3. Results and discussion

3.1. Analysis of ZIF-67

SEM and TEM images of the synthesized ZIF-67 submicron crystals are presented in Fig. 1a–c. The SEM micrographs (Fig. 1a and b) reveal well-defined polyhedral particles with a predominantly rhombic dodecahedral morphology, which is characteristic of ZIF-67 crystallizing in a sodalite type framework. This morphology is widely reported and originates from the intrinsic crystal symmetry and coordination environment of Co²⁺ with 2-methylimidazole ligands.¹² The particles exhibit smooth surfaces and sharp edges, indicative of high crystallinity. The average particle size, calculated from measurements of 100 individual particles, was 413 ± 104 nm. The TEM image (Fig. 1c) further confirms the uniform polyhedral structure of the particles. The homogeneous electron contrast across the crystals suggests a uniform internal structure, consistent with the formation of well-defined ZIF-67 frameworks.¹³

Fig. 1. (a) SEM images of ZIF-67 (×20 000 with scale bar of 1 µm) and particle size distribution based on measurements of 100 particles (inset), (b) SEM images (×100 000, scale bar: 500 nm), (c) TEM image of ZIF-67 particles (scale bar: 200 nm), and (d) XRD pattern of ZIF-67.

The XRD pattern of ZIF-67 is presented in Fig. 1d. Distinct diffraction peaks observed at 2θ values of 7.3, 10.4, 12.7, 14.7, 16.4, 18.0, 22.1, 24.5, and 26.6° can be indexed to the (011), (002), (112), (022), (013), (222), (114), (233), and (134) crystallographic planes, respectively, which are characteristic of ZIF-67. In addition, weaker reflections detected at higher angles, specifically at 29.6, 30.56, 31.46, and 32.36°, can be attributed to higher-order planes such as (044), (334), and (235), in agreement with the simulated diffraction pattern reported for ZIF-67.12 The presence of sharp, well-defined diffraction peaks indicates that the synthesized material is highly crystalline and phase pure. These findings are consistent with the SEM and TEM observations, which revealed well-defined polyhedral particles with uniform morphology.

The N₂ adsorption–desorption isotherm, pore size distribution, FTIR spectrum, and TGA/DTG curves are presented in Fig. 2a–d to characterize the surface, chemical, and thermal properties of ZIF-67. As shown in Fig. 2a, ZIF-67 exhibits a type I adsorption isotherm, characteristic of microporous materials, with a sharp increase in gas uptake at low relative pressure followed by a plateau region. This behavior is indicative of micropore filling and the absence of significant mesoporosity.¹⁴ The SF-DFT pore size distribution shown in Fig. 2b exhibited a dominant adsorption-derived micropore width centered at approximately 6.1 Å, confirming the microporous character of the synthesized ZIF-67. Reported larger pore sizes in the range of 1.6–2.0 nm are typically associated with internal cage dimensions or interparticle voids rather than the accessible micropore size derived from DFT analysis.^15,16 Additionally, ZIF-67 exhibited a specific surface area of 2163 m² g⁻¹, determined by multipoint BET analysis. The t-plot method yielded a micropore volume of 0.661 cm³ g⁻¹ and an external surface area of 53.93 m² g⁻¹, confirming the predominantly microporous character of the material, consistent with the type I isotherm and pore size distribution. The measured surface area and micropore volume are comparable to or exceed values reported in the literature, including 1743 m² g⁻¹,¹⁶ 1696 m² g⁻¹,¹⁷ and 1229.8 m² g⁻¹,¹⁵ likely reflecting differences in synthesis conditions and crystallinity.

Fig. 2. (a) N₂ adsorption isotherm, (b) pore size distribution, (c) FTIR spectrum, and (d) TGA/DTG curves of ZIF-67.

FTIR spectrum of ZIF-67 (Fig. 2c) provides insight into the chemical structure of the framework. The band observed at 1578 cm⁻¹ is attributed to C N/C C stretching vibrations of the imidazolate ring, indicating the presence of the aromatic heterocyclic structure.¹⁸ In the fingerprint region, several distinct bands are evident. The peak at 1420 cm⁻¹ corresponds to in-plane C–H bending vibrations of the imidazole ring, while the band at 1306 cm⁻¹ is assigned to C–N stretching vibrations. The strong absorption at 1142 cm⁻¹ is associated with C–N stretching modes of the imidazolate linker. Additionally, the band at 993 cm⁻¹ can be attributed to ring deformation and C–N bending vibrations.^19,20 Additionally, the peaks at 756 and 694 cm⁻¹ are assigned to out-of-plane C–H bending vibrations of the imidazole ring, which are typical for ZIF-type structures.²¹ The presence of all these characteristic imidazolate vibrations confirms the successful incorporation of 2-methylimidazole within the framework and further verifies the formation of the ZIF-67 structure.

Finally, the thermal stability of ZIF-67 was evaluated by TGA/DTG analysis (Fig. 2d). A minor weight loss observed below 150 °C is attributed to the removal of physically adsorbed moisture and surface-bound species. In the temperature range of approximately 150–300 °C, a slight and gradual mass decrease is observed, which can be associated with the desorption of residual solvents or loosely confined guest molecules within the pores. No distinct DTG peak in this region suggests the absence of significant structural degradation. The main decomposition step occurs above ∼500 °C, with a sharp DTG peak centered at approximately 550–560 °C, indicating the maximum rate of mass loss. This stage corresponds to the thermal decomposition of the imidazolate linkers and the consequent collapse of the ZIF-67 framework. Following this event, the remaining mass is likely associated with cobalt-containing residues and carbonaceous species formed during decomposition. The high thermal stability up to ∼500 °C is characteristic of well-formed ZIF-67 structures and is consistent with the crystalline framework confirmed by XRD and the coordination environment identified by FTIR, supporting the successful formation of the metal–organic framework.²²

3.2. Morphology of PLA/ZIF-67 films

SEM micrographs illustrating the surface and cross-sectional morphology of PLA/ZIF-67 films are presented in Fig. 3(a–h). The results indicate that ZIF-67 incorporation influences the microstructure of the PLA matrix in a concentration-dependent manner. Neat PLA and PLA/ZIF0.3 films exhibit smooth, continuous surfaces and relatively compact cross-sections (Fig. 3a, b, e and f), suggesting a homogeneous matrix without apparent structural defects. At intermediate loading (PLA/ZIF0.6), the film maintains a uniform and dense morphology with no obvious large-scale agglomeration (Fig. 3c and g). The absence of visible phase separation or microvoids at this scale indicates improved filler–matrix compatibility compared to higher loadings.

Fig. 3. Surface SEM images of PLA/ZIF composite films recorded at 1 000× (scale bar 50 µm): (a) PLA/ZIF0, (b) PLA/ZIF0.3, (c) PLA/ZIF0.6, and (d) PLA/ZIF0.9; (e–h) corresponding cross-sectional SEM images in the same order.

In contrast, the PLA/ZIF0.9 film shows clear signs of structural heterogeneity (Fig. 3d and h), including irregular domains and localized particle clusters, which are indicative of particle agglomeration at higher filler content. These agglomerates can act as stress concentration points and disrupt the continuity of the polymer matrix. Such morphological changes provide a structural basis for the variations observed in the mechanical and barrier properties. The relatively uniform structure of PLA/ZIF0.6 is consistent with its improved tensile performance and reduced permeability, whereas the heterogeneous structure of PLA/ZIF0.9 likely contributes to its inferior mechanical properties and increased permeability. A more detailed discussion of these structure–property relationships is provided in the subsequent sections. Similar behavior has been reported in PLA-based systems containing ZIF-type fillers, where excessive loading leads to aggregation and deterioration of performance.²³

3.3. The XRD and FTIR of PLA/ZIF-67 films

XRD patterns of PLA/ZIF films with varying ZIF-67 content are presented in Fig. 4a. The PLA/ZIF0 sample exhibits a broad diffraction halo in the range of 2θ ≈ 10–25°, with a weak shoulder around 16–18°, indicating a predominantly amorphous structure with only limited crystalline ordering. This behavior is characteristic of solvent-cast PLA systems, where rapid solvent evaporation restricts chain rearrangement and suppresses the formation of well-defined crystalline domains.²⁴

Fig. 4. (a) XRD diffraction patterns, and (b) FTIR spectra, (c) TS and (d) EAB of PLA/ZIF films. Different uppercase letters indicate significant differences among film formulations (p < 0.05).

Following the incorporation of ZIF-67, the overall diffraction profiles remain largely unchanged. All samples display similar broad halos within 2θ ≈ 10–25°, and no distinct diffraction peaks corresponding to crystalline ZIF-67 are detected. This indicates that the amorphous character of the PLA matrix dominates the diffraction response across the studied composition range. The absence of well-defined reflections associated with either PLA crystallites or ZIF-67 suggests that no significant long-range ordering develops upon filler incorporation. Considering the relatively low filler loadings (≤0.9 wt%), this observation is expected, as the scattering contribution of ZIF-67 is insufficient to overcome the strong amorphous background of PLA. Similar masking effects have been reported in PLA-based composite systems, where the broad amorphous halo of the polymer overlaps with and suppresses filler-related diffraction features.²⁵ Due to the amorphous-dominated nature of the patterns and the lack of clearly resolved crystalline peaks, the crystallinity index (CI) was estimated under fixed fitting constraints, where identical peak width parameters were applied to all samples to ensure consistency. Under these conditions, only minor variations in CI were observed. The calculated values were 17.29 ± 0.38 for PLA/ZIF0, 17.92 ± 0.47 for PLA/ZIF0.3, 16.73 ± 0.70 for PLA/ZIF0.6, and 16.40 ± 0.81 for PLA/ZIF0.9. Although a slight decrease in peak intensity and broadening of the diffraction halo can be visually observed with increasing ZIF-67 content, these changes do not translate into a meaningful difference in crystallinity index. This is attributed to the predominantly amorphous nature of the patterns where the overlap between amorphous and weak crystalline contributions limits the sensitivity of CI calculations.

In Fig. 4b, FTIR spectra of PLA/ZIF-67 films showed the characteristic bands of PLA, including weak –CH₃ stretching vibrations at 2998 and 2945 cm⁻¹, a strong ester carbonyl band at 1751 cm⁻¹, –CH₃ bending/deformation bands at 1452 and 1359 cm⁻¹, and intense C–O–C/C–O stretching bands at 1184 and 1084 cm⁻¹, together with fingerprint-region bands at 868 and 752 cm⁻¹.²⁶ No distinct new absorption bands assigned solely to ZIF-67 were detected in the composite films, most likely due to the low filler loading and overlap between PLA and imidazolate-related vibrations of ZIF-67. The absence of new absorption bands indicates that the chemical structure of the PLA matrix was preserved after ZIF-67 incorporation, with no evidence of new bond formation.²⁷

3.4. Mechanical property

The incorporation of ZIF-67 into PLA films leads to a non-monotonic mechanical response governed by dispersion state and interfacial interactions. As observed in Fig. 4c, TS increases from PLA/ZIF0 to PLA/ZIF0.3 and reaches a maximum at PLA/ZIF0.6, followed by a decrease at PLA/ZIF0.9. The FTIR spectra of the composites are dominated by PLA and show no clear evidence of new ZIF-67–PLA interfacial bonding (Fig. 4b). The optimum response at 0.6 wt% and the decline at 0.9 wt% are therefore interpreted as being consistent with improved filler dispersion at lower loading and possible aggregation at higher loading, based collectively on the SEM, tensile, and barrier results, rather than as evidence of strong interfacial chemical bonding.

EAB data further support this interpretation in Fig. 4d. While PLA/ZIF0 exhibits moderate ductility, the PLA/ZIF0.6 sample shows a clear increase in EAB, indicating enhanced toughness. Such simultaneous improvement in strength and ductility is typically attributed to well-dispersed fillers that promote a more uniform stress distribution and hinder crack propagation.²⁸ In MOF–polymer systems, the high surface area of the filler may improve stress transfer and increase energy dissipation during deformation.²⁹

In contrast, the PLA/ZIF0.9 formulation exhibits a reduction in TS and a dramatic decrease in EAB, indicating a transition from reinforcement to defect-dominated behavior. At higher loadings, ZIF-67 particles tend to agglomerate due to their high sulrface energy and porous structure, leading to poor interfacial adhesion and the formation of stress concentration sites. These aggregates disrupt the continuity of the PLA matrix and facilitate premature crack propagation, thereby reducing both strength and ductility. Similar deterioration at higher filler contents has been widely reported in composite systems when particle agglomeration reduces dispersion quality and creates structural defects within the polymer matrix.³⁰ In short, 0.6 wt% ZIF-67 is the loading at which tensile strength and elongation are highest. At higher loadings (0.9 wt%), aggregation becomes dominant, resulting in deterioration of the mechanical properties, which is consistent with aggregation suggested by SEM (Fig. 3d and h).

3.5. Thermal properties

The thermal degradation behavior of PLA/ZIF-67 composite films was evaluated by TGA and DTG analysis (Fig. 5a and b). The thermogravimetric behavior of PLA/ZIF-67 films showed a clear change in degradation pathway following ZIF-67 incorporation. PLA/ZIF0 exhibits a single dominant degradation step with a sharp mass loss between ∼320–380 °C, which is characteristic of PLA thermal depolymerization via random chain scission and unzipping reactions leading to lactide formation.^31,32 This is also reflected in the DTG curve, where a single intense peak is observed around ∼360–370 °C, indicating a one-step degradation mechanism.

Fig. 5. (a) TGA, (b) DTG, (c) WVP and (d) OP of PLA/ZIF films. Different uppercase letters indicate significant differences among film formulations (p < 0.05).

Upon addition of ZIF-67, the degradation profile changes significantly. All composite samples (PLA/ZIF0.3, PLA/ZIF0.6, PLA/ZIF0.9) show an earlier onset of degradation, with the main weight loss shifted toward lower temperatures (∼260–310 °C). This reduction in thermal stability is commonly reported in MOF containing polymer systems and is generally attributed to the catalytic effect of metal centers.³³ In the case of ZIF-67, the presence of Co-containing sites may contribute to the earlier onset of PLA degradation, as reflected by the shift of the main decomposition step to lower temperatures. Similar decreases in thermal stability have been reported for polymer systems containing metal-based fillers or MOFs, where interactions between the polymer matrix and the inorganic phase modify degradation pathways rather than purely acting as inert reinforcement.³⁴ The DTG curves support this interpretation, as the main degradation peaks for the composites are broader and shifted to lower temperatures compared to PLA/ZIF0.

Another important observation is the emergence of multistep degradation behavior in the ZIF-67-containing films. Unlike the PLA/ZIF0 control film, which exhibited a predominantly single-step degradation profile, the PLA/ZIF composites showed a shoulder or secondary degradation region at approximately 350–420 °C. Since PEG400 was present at the same concentration in all formulations, this difference is associated with the incorporation of ZIF-67. However, the secondary degradation region cannot be assigned exclusively to the direct decomposition of the ZIF-67 framework. Instead, it likely reflects a ZIF-67-induced modification of the degradation pathway of the plasticized PLA matrix, possibly related to the catalytic effect of Co-containing sites.¹⁸

The residual mass at high temperature (above ∼450 °C) increases with increasing ZIF-67 content, with PLA/ZIF0.9 showing the highest char yield. This is expected since MOFs leave behind inorganic residues (e.g., cobalt oxides or metallic cobalt) after decomposition. Increased char formation is often associated with improved barrier properties against heat and mass transfer during degradation, although in this system, it does not compensate for the catalytic reduction in onset temperature.³⁵ In summary, ZIF-67 incorporation shifted the main degradation of PLA to lower temperatures while increasing the residual mass and introducing a secondary high-temperature degradation stage.

3.6. Barrier behavior of PLA/ZIF-67 films

Embedding ZIF-67 into PLA films significantly affected their barrier performance against water vapor and oxygen (p < 0.05). WVP decreased from 3.61 × 10¹¹ g m⁻¹ s⁻¹ Pa⁻¹ for PLA/ZIF0 to 2.80 × 10⁻¹¹ g m⁻¹ s⁻¹ Pa⁻¹ for PLA/ZIF0.6 (Fig. 5c). Increasing the ZIF-67 loading to 0.9 wt% did not result in a further statistically significant reduction. A similar trend was observed for OP, which decreased from 1.15 × 10⁻¹⁵ for PLA/ZIF0 to 5.3 × 10⁻¹⁶ for PLA/ZIF0.6, followed by an increase at 0.9 wt% (Fig. 5d). The reduction in both WVP and OP at low to intermediate ZIF-67 loadings can be attributed to the introduction of dispersed inorganic domains within the PLA matrix, which increase the tortuosity of diffusion pathways and hinder the transport of gas and moisture molecules.^36,37 This effect is most pronounced at 0.6 wt%, where improved dispersion is expected based on the mechanical results. At higher loading (0.9 wt%), the slight increase in permeability suggests the onset of particle aggregation, which can disrupt matrix continuity and generate preferential diffusion pathways, thereby partially offsetting the barrier effect. Similar behavior has been widely reported in polymer composites when filler dispersion deteriorates at elevated loadings.^36,37

3.7. Sensitivity of PLA/ZIF-67 films to ammonia vapor

During the early stages of protein-rich food spoilage, protein degradation leads to the formation of total volatile basic nitrogen (TVB-N) compounds such as TMA, DMA, and ammonia, primarily due to microbial and enzymatic activity. Since ammonia is one of the volatile basic compounds associated with the spoilage of protein-rich foods, the colorimetric response of PLA/ZIF-67 films was evaluated under vapor generated from a 0.8 M ammonia solution (Fig. 6a; visual images in Fig. S1a). Before exposure, the ZIF-67-containing films exhibited a purple color. Exposure to ammonia vapor caused progressive fading and a subsequent transition toward brownish-yellow tones.

Fig. 6. Time-dependent ΔE values of PLA/ZIF films. (a) PLA/ZIF films exposed to ammonia vapor. (b) Comparative response of PLA/ZIF0.6 to ammonia, DMA, and TMA vapors generated from solutions prepared to give the same estimated headspace partial pressure as 0.8 M ammonia. (c) Response of PLA/ZIF0.6 to ammonia vapor before and after humid storage (80% RH, 48 h). Values are presented as mean ± SD.

The color transition is proposed to result from perturbation of the Co(ii)–imidazolate chromophore and/or its local optical environment upon uptake of volatile bases.³⁸ Pure ZIF-67 powder showed no visible color change, whereas the 0.3% and 0.6% ZIF-67 films responded, while the 0.9% film did not, indicating a strong dependence on the composite-film environment. FTIR analysis performed before and after ammonia exposure showed no new characteristic bands (data not shown), providing no direct evidence of new chemical bonding. Due to the low ZIF-67 loading, its XRD reflections could not be reliably resolved from the PLA matrix. The color change did not recover after ammonia exposure under the tested conditions; therefore, the film is considered a single-use, non-contact freshness indicator. With increasing exposure time, a gradual fading of the color was observed (Fig. S1a), indicating progressive interaction between ammonia and the active sites of the composite.

A ΔE value above 5 is generally considered perceptible to an untrained observer.¹¹ The PLA/ZIF0.3 film showed a gradual increase in ΔE and remained near or below the perceptibility threshold during the first 10 min, reaching approximately 8 after 20 min. PLA/ZIF0.6 exhibited a larger initial response (ΔE ≈ 4.9 at 2 min), rose to approximately 8.2 at 4 min, and then increased more gradually, reaching approximately 11.5 at the end of the exposure (Fig. 6a).

The PLA/ZIF0.9 film exhibited ΔE values below 5 throughout the test, suggesting reduced sensitivity at higher filler loading. This behavior is consistent with previous studies, where excessive MOF incorporation led to partial aggregation and reduced accessibility of active sites, limiting effective gas interaction and color response.⁶

Considering its rapid and perceptible response together with balanced physicochemical properties, PLA/ZIF0.6 was selected for the subsequent non-contact headspace test with chicken breast. Similar MOF-based intelligent films have demonstrated effective ammonia-responsive discoloration and applicability in monitoring seafood and meat spoilage, supporting the suitability of ZIF-67 as a functional sensing component in intelligent packaging systems.³⁹

3.8. Comparative response to ammonia, DMA, and TMA vapors

The comparative colorimetric response of PLA/ZIF0.6 to ammonia, DMA, and TMA vapors was evaluated under otherwise identical exposure conditions using aqueous solutions prepared to achieve the same estimated headspace partial pressure as 0.8 M ammonia, corresponding to 0.8 M ammonia, 0.407 M DMA, and 0.117 M TMA. The film exhibited a rapid and visible response to all three volatile bases (Fig. 6b).

The ΔE values of the PLA/ZIF0.6 sample during exposure to ammonia, DMA, and TMA were initially close to the perceptibility threshold (ΔE ≈ 5). For ammonia, ΔE increased more substantially, reaching approximately 11.5 after 20 min, compared with approximately 6.6 for DMA and 5.8 for TMA. Thus, under the estimated equivalent headspace conditions, all three amines produced a similar initial color response, whereas ammonia resulted in the largest color difference after 20 min of exposure.

The rapid response to these volatile bases may be associated with their interactions with the Co-based coordination environment of ZIF-67. Previous studies have related the color change of ZIF-67-based intelligent films to adsorption of basic nitrogenous compounds, changes in Co–N coordination, hydrogen-bonding interactions with the imidazolate framework, and partial modification of the framework structure.^6,38 Overall, the response to ammonia, DMA, and TMA demonstrates the ability of PLA/ZIF0.6 to detect multiple volatile basic compounds associated with the spoilage of protein-rich foods.

A proposed structural representation is shown in Scheme 1. ZIF-67 is physically dispersed within the PEG-plasticized PLA matrix, with no spectroscopic evidence of new covalent bonding between PLA and ZIF-67. The colorimetric response is attributed to the interaction of volatile amines with the tetrahedral Co(ii)–2-methylimidazolate framework, with the Co(ii) center proposed as the most plausible optical response site.

Scheme 1. Proposed assembly of the PLA/ZIF-67 indicator and the proposed site of amine interaction. (a) ZIF-67 particles are physically dispersed within PEG400-plasticized PLA, with FTIR showing no evidence of new covalent PLA–ZIF-67 bonding. (b) Ammonia is proposed to interact primarily with the tetrahedral Co(ii) center of ZIF-67, the main visible chromophore, potentially perturbing its coordination environment and producing the observed color change from purple to brownish-yellow. A secondary N–H⋯N interaction with the imidazolate nitrogen may also contribute. DMA and TMA are proposed to interact through a similar mechanism at the Co(ii) center.

3.9. Stability under high relative humidity

The effect of humid storage on the ammonia response of PLA/ZIF0.6 was evaluated after 48 h at 80% RH and compared with films conditioned at 50% RH (Fig. 6c). Both films showed a perceptible color change within 2 min (ΔE ≈ 5). However, the film conditioned at 50% RH exhibited a greater subsequent increase in ΔE, reaching approximately 11.5 at 20 min, whereas the ΔE of the film subjected to high moisture (80% RH) increased from approximately 3 to 4 after 4 min, reached approximately 5 after 8 min, and continued to increase to approximately 7 after 20 min.

Therefore, humid storage generally weakened the color response of the PLA/ZIF0.6 film. This behavior may be associated with water uptake by the hygroscopic PEG400-plasticized PLA matrix and ZIF-67 micropores, where pre-adsorbed water can compete with ammonia for adsorption sites and limit further interaction with Co(ii) centers. The initial ΔE values indicate that sensing sites remained after humid storage; the color response reduced rather than completely deactivated. Notably, this pretreatment differs from the chicken-headspace experiment, where moisture and volatile amines are generated simultaneously over time; nevertheless, the latter still produced a visible response under humid conditions.

3.10. Application of PLA/ZIF-67 films for the detection of chicken freshness

The changes in total volatile basic nitrogen (TVB-N) and color difference ΔE of chicken samples during accelerated storage at 23 °C over 60 h are presented in Fig. 7a. The TVB-N values increased continuously with storage time due to microbial and enzymatic degradation of proteins. During spoilage, volatile compounds such as ammonia, DMA, and TMA are generated, contributing to the increase in TVB-N and the development of off-odours.⁴⁰ The TVB-N content increased from 9.6 ± 0.8 to 112.8 ± 1.6 mg N/100 g over 60 h. Turkish Standard TS 2409:2014 specifies a critical TVB-N limit of 30 mg N/100 g for chicken carcasses. In the present study, this value was adopted as the reference spoilage threshold and was exceeded at approximately 36 h, indicating the transition from fresh to spoiled chicken.⁷

Fig. 7. Changes in TVB-N and ΔE of chicken samples during storage at 23 °C (a) TVB-N and ΔE as a function of time (b) visual color change of PLA/ZIF0.6 indicator film during storage. The dashed horizontal line represents the TVB-N acceptability limit (30 mg N/100 g). Error bars indicate standard deviation. (c) Correlation of ΔE with TVB-N (means ± SD, n = 3). Solid line: 0–60 h (ΔE = 6.98 + 0.163 × TVB-N, R² = 0.552). Dashed line: 0–48 h (R² = 0.689). The dotted vertical line marks the TVB-N acceptability limit (30 mg N/100 g).

In parallel, ΔE values increased progressively with storage time and TVB-N content. When the TVB-N value approached the spoilage threshold, ΔE exceeded 15, which is well above the human perceptibility limit (ΔE > 5), and further increased to 21.8 ± 0.6 at 60 h. The corresponding visual changes of the PLA/ZIF0.6 film, from purple to pale and subsequently brownish tones, are shown in Fig. 7b. This color response is associated with the interaction of volatile amines with the Co-based coordination environment in ZIF-67, leading to changes in its optical properties. The agreement between ΔE development and TVB-N accumulation indicates that PLA/ZIF0.6 can monitor the progression of chicken spoilage under the applied storage conditions. These findings support its use as a non-contact headspace indicator label that provides visible information on chicken freshness without directly contacting the food.

To assess whether the color response tracks chemical freshness, ΔE was plotted against TVB-N (Fig. 7c). Regression of the six time-point means (0–60 h) yielded ΔE = 6.98 + 0.163 × TVB-N (R² = 0.552), while fitting 0–48 h gave ΔE = 2.32 + 0.359 × TVB-N (R² = 0.689). The modest R² values likely reflect partial saturation of the optical response. ΔE exceeded 15 at 36 h, when TVB-N reached 38.2 ± 1.1 mg N/100 g, above the 30 mg N/100 g acceptability limit, whereas TVB-N increased further to 112.7 ± 1.6 mg N/100 g at 60 h with ΔE reaching only 21.8 ± 0.6. Thus, the relationship is correlative under accelerated storage at 23 °C rather than a transferable calibration for quantitative TVB-N prediction. Importantly, the label produced a perceptible color change (ΔE > 15) around the TVB-N acceptability limit. The chicken test serves as a proof-of-concept under accelerated conditions and does not establish performance under refrigerated storage or in commercial packaging.

4. Conclusion

PLA/ZIF-67 composite films were successfully prepared and evaluated as non contact headspace labels under accelerated chicken breast storage. ZIF-67 was confirmed as a crystalline, microporous, high-surface-area cobalt-based framework with a characteristic polyhedral morphology. Incorporating it into PLA affected film structure and performance in a loading-dependent manner. The PLA/ZIF0.6 formulation provided the best overall balance, showing improved tensile behavior, reduced oxygen and water vapor permeability, and a rapid colorimetric response to volatile basic nitrogen compounds. In contrast, the higher ZIF-67 loading caused particle aggregation, which weakened the mechanical and barrier performance.

The spectroscopic and diffraction results indicate that ZIF-67 was physically incorporated into the PLA matrix without evidence of new covalent or coordination bond formation. Thermal analysis showed that ZIF-67 shifted the main PLA degradation step to lower temperatures while increasing the residual mass at high temperature, indicating a modified degradation pathway. In the chicken freshness test, PLA/ZIF0.6 was used as a non-contact headspace indicator label and showed a visible transition from purple to pale brownish-yellow as TVB-N increased during accelerated storage at 23 °C. These findings support PLA/ZIF-67 films as non-contact colorimetric labels for poultry freshness under accelerated storage at 23 °C; refrigerated performance remains to be evaluated. The present results relate specifically to non-contact indicator use, as cobalt migration and residual solvent levels were not evaluated. These aspects should be investigated further before any direct food-contact application is considered.

Conflicts of interest

There are no conflicts to declare.

Data availability

Data are available upon request from the authors. Further, the data supporting this article have been included as part of the supplementary information (SI). Supplementary information: Fig. S1 visual color response of PLA/ZIF films upon exposure to volatile amines: (a) gaseous ammonia and (b) dimethylamine and trimethylamine for PLA/ZIF_0.6 films. Fig. S2 visual color change of PLA/ZIF_0.6 indicator film during storage of chicken samples at 23 °C under closed conditions. The film gradually changes from purple to pale and brownish tones with increasing storage time, indicating accumulation of volatile amines. See DOI: https://doi.org/10.1039/d6ra06836g.