Development of an active packaging system containing zinc oxide nanoparticles for the extension of chicken fillet shelf life

Abstract The casting method was employed to prepare gelatin‐based nanocomposite films containing different concentrations of cellulose nanofiber (CNF) as a reinforcement filler (2.5%, 5%, and 7.5% w/w of gelatin) as well as zinc oxide nanoparticles (ZnO NPs) as an antimicrobial agent (1%, 3%, 5%, and 7% w/w of gelatin). The results showed that the incorporation of 5% CNFs (optimum concentration) significantly boosted the films' stiffness (YM; by 47%) and strength (TS; by 72%) but decreased its flexibility (EAB; by 28%), water vapor permeability, and moisture absorption. The best G/CNF film antibacterial activity was provided by the 5% concentration of ZnO NPs according to the disk diffusion assay; Gram‐positive bacteria were inhibited significantly more than Gram‐negative bacteria. The antimicrobial efficacy of the G/CNF/ZnO NPs film as a food packaging material was assessed via counts of Staphylococcus aureus and Pseudomonas fluorescens inoculated on chicken fillets (as a food model) in the treatment (G/5% CNF/5% ZnO) and control groups (plastic bag). The antibacterial film led to a significant reduction in the bacterial load of the chicken fillets (p < .05), especially against the Gram‐positive strain. This study illustrated that G/CNF/ZnO NPs films can be utilized as active packaging to prolong the shelf life of different perishable foods such as meat.

food products with prolonged shelf life have made the food and packaging industries pay more attention to the preparation of edible, biodegradable packaging films from natural macromolecules and biopolymers such as proteins, polysaccharides, and lipids or their combinations (Mohammadi, Kamkar, & Misaghi, 2018;Pirouzifard, Yorghanlu, & Pirsa, 2019).
Gelatin is one of the most common biopolymer proteins, obtained from thermal hydrolysis or physicochemical destruction of collagen derived from animal skin and bone (Núñez-Flores et al., 2012). The low cost, efficient lipid barrier properties, film-forming capacity, and biodegradability of gelatin make it suitable for use in edible film production. However, its weak mechanical properties, inadequate bonding capacity, and high permeability to water are its major drawbacks (Bodini, Sobral, Fávaro-Trindade, & Carvalho, 2013). Thus, suitable fillers can be added to gelatin films to form composites with amended properties. Nanocomposites illustrate a promising alternative for achieving improved physicomechanical, thermal, and water resistance properties (Arora & Padua, 2010).
Cellulose nanofibers (CNFs) are aggregations of primary cellulose fibrils with lengths of 2-20 nm in diameter isolated through mechanical operations, such as high-pressure homogenization, grinding, and purifying (Chen et al., 2011). As highly applicable organic nanofillers for the reinforcement of nanocomposite films, CNFs are biodegradable, abundant, accessible, nontoxic, lightweight and have high aspect ratios (Deepa et al., 2016). Studies on CNFs have shown that they can improve the mechanical performance of films and reduce their water solubility, swelling ratio, and water vapor permeability (Chaabouni & Boufi, 2017;Narita, Okahisa, & Yamada, 2019;Samadani, Behzad, & Enayati, 2019;Yu et al., 2017). As an effective reinforcement additive for biopolymer materials, CNF backbone chains have a unique intrinsic structure that facilitates interfacial interactions between the polymer matrix and the nanoparticles (Dai et al., 2017).
On the other hand, microbial contamination is a long-standing problem affecting foods (particularly meats) since foodborne bacteria and fungi are involved in food spoilage and poisoning, leading to economic losses, human health risks, quality reduction, and decreased product life (Clarke et al., 2016;Seydim & Sarikus, 2006). Among the various meats, chicken is more extensively consumed considering it is low-fat, nutritious and has a relatively low price (Azlin-Hasim et al., 2016). Notwithstanding all the mentioned advantages, this meat is highly susceptible to spoilage as its protein and moisture components, as well as suitable pH, allow the growth of both pathogenic and nonpathogenic microbes (Takma & Korel, 2019). The use of metal oxide nanoparticles such as TiO2, ZnO, and CuO as thermally stable antimicrobial agents for the production of active biopolymer packaging is one of the ways of preventing foodborne illness and thereby enhancing food safety (McMillin, 2017;Shankar, Teng, & Rhim, 2014).
In the present research, we aimed to fabricate G/CNF films via the casting method to achieve improved physicomechanical properties relative to neat gelatin. Then, we developed active antimicrobial films by adding ZnO NPs to the G/CNF film and evaluated the antimicrobial properties of the product using chicken fillet as a food model.

| Bacterial strains
The Biological and Genetic Resources Center supplied the Staphylococcus aureus (ATCC 33591) and Pseudomonas fluorescens (ATCC 13525) strains.

| Preparation of G/CNF and G/CNF/ZnO nanocomposite films
To prepare the films, we utilized the casting procedure. The gelatin film solution (as control) was made by mixing gelatin powder (3.5% w/w) in distilled water and heating the mixture with a water bath (JulaboMP-5) at 70°C for 1 hr until obtaining a clear, light yellow solution. To prepare the CNF suspension, different concentrations of CNF gel (2.5%, 5%, and 7.5% based on gelatin) were dispersed in distilled water with continuous stirring for 2 hr. To facilitate complete dispersion, an ultrasonic bath treatment (Belfor) was applied for half an hour. Then, the CNF suspensions were dropped into the gelatin solution. Then, as the plasticizer, 40% glycerol (based on gelatin) was added to prepare G/CNF nanocomposite films under stirring. ZnO NP solutions (1%, 3%, 5%, and 7% w/w based on gelatin) were prepared by the same production procedure of the CNF suspensions, before being added drop by drop to the G/CNF 5% film solution containing 1.4 g of glycerol under continuous stirring over 2 hr. The film solutions were then spread within Petri dishes (diameter = 8 cm) and left to dry at ambient temperature for 48 hr.
Subsequently, the dry film samples were separated and kept in polyethylene bags.

| Determination of mechanical properties
To evaluate the films' mechanical characteristics, including Young's modulus (YM), percentage of elongation at break (EAB), and tensile strength (TS), the standard methods of the ASTM (Standard, 2010) were followed using the Sanaf Universal Testing Machine (Tehran, Iran). The film specimens, after being conditioned for 24 hr at 55 ± 3% RH and 25°C, were cut into dumbbell shapes and loaded into the device (50 mm initial grip separation; 10 mm/min crosshead speed; 25 N load; room temperature).

| Film thickness
A digital micrometer (Guanglu; 0.01 mm precision) was used to evaluate the thickness of each film sample at five different points that were randomly selected. The mean value was reported.

| Light transmittance and opacity
To investigate the film's optical barrier characteristics, a spectrophotometer (Unico, UV-2100) was employed with a wavelength range of 400-800 nm. Rectangular film samples (1 × 5 cm) were used, with an empty cell being the point of reference. Equation 1 was used to calculate the opacity of the films (Salari, Khiabani, Mokarram, Ghanbarzadeh, & Kafil, 2018).
Here, Abs 600 is the absorbance at 600 nm and X denotes the thickness of the film in mm.
Higher opacity is indicative of lower light transmission.

| Moisture content
To determine the films' moisture content (MC), the gravimetric procedure was employed. Film strips (20 mm × 20 mm) were dried within a laboratory oven at 105 ± 1°C until reaching steady weights.
After triplicate experimentation, the mean weight values were inserted into Equation 2 to obtain the MC.
Here, M 0 and M 1 represent the film weight (g) initially and postdrying, respectively (Salari et al., 2018

| Moisture absorption (MA)
Evaluation of the films' capacity to absorb moisture was performed according to the method of Almasi, Ghanbarzadeh, Dehghannya, Entezami, and Asl (2015) with slight modifications. Film strips (20 × 20 mm 2 ) were subjected to conditioning with CaCl 2 (0% RH) at room temperature. The samples were then weighed before being placed in a desiccator containing saturated sodium chloride solution (75% RH). The weighing was repeated multiple times until an equilibrium state was achieved. The MA was obtained using Equation 3.
In the above equation, the initial sample weight (at 0% RH) and the weight at time t (at 75% RH) are represented by W 0 and W t , respectively.

| Water solubility (WS)
To evaluate the solubility of the film samples with respect to water, the procedure described by Hosseini, Rezaei, Zandi, and Farahmandghavi (2015) was slightly modified and employed. Three pieces (2 × 2 cm 2 ) of the films were dried for 6 hr in a laboratory oven (105°C). The initial weight (W i ) of the films was then recorded (±0.0001 g). Subsequently, the films were immersed in distilled water (50 ml) then gently shaken at 100 rpm overnight at ambient temperature. Preweighed filter paper was then used for sample filtration. The insoluble section and filter paper were subjected to oven-drying (105°C, 6 hr) before the final weight was recorded (W f ) and WS% was obtained by use of Equation 4.

| Determination of water vapor permeability (WVP)
To determine the films' WVP, the standard ASTM E96-05 (ASTM, 2005) procedure was followed. Cylindrical vials containing 5 g of anhydrous calcium sulfate (CaSO 4 , RH = 0%) were used as the permeation cells. The conditioned round-shaped film pieces were used to seal the vials, before the initial weight was recorded. The sealed vials were then placed in a desiccator containing distilled water (RH = 100%) and set at room temperature. The vials were repetitively weighed at 1 hr intervals over 8 hr. After plotting the weight as a function of time (weight vs. time) and determining the slope by linear regression (r 2 > .99), Equations 5-7 were applied to determine the WVP (g.mm/Pa.h.mm 2 ), water vapor transmission rate (WVTR; g/h.mm 2 ), and partial water vapor pressure difference across the film (ΔP; Pa).
In the above equations, the thickness (mm) of the films is denoted by L; P is the vapor pressure of water at the saturation point (3,169 Pa) and room temperature, with H 1 and H 2 denoting the relative humidity within the desiccator and vial, respectively.

| Scanning electron microscopy (SEM)
A TESCAN MIRA 3 XMU scanning electron microscope (SEM) was used to investigate the influence of the CNFs and ZnO NPs on the morphological characteristics of the gelatin-based films. For this purpose, the film samples were sputter-coated with gold; the acceleration voltage used during scanning was 10 kV.

| Fourier transform infrared (FT-IR) spectroscopy
To analyze the chemical structure and interactions between the components, FT-IR spectroscopy was applied in the range of 4,000-500 cm −1 ; and 100 scans were performed with a resolution of 1 cm −1 . The FT-IR spectra of the gelatin, G/CNF, and G/CNF/ ZnO films in KBr pellets were measured on a Bruker Tensor-27 Spectrometer at room temperature and reported based on the transmission.

| Evaluation of the antibacterial activity of G/ CNF/ZnO films
The activity of the nanocomposites against bacteria was evaluated by the disk diffusion test. Pseudomonas fluorescens and Staphylococcus aureus were cultured in nutrient broth for 18 hr.
The bacterial suspensions were then collected and set to 0.5 McFarland standard turbidity (1.5 × 10 8 CFU/ml). Next, through 1:100 dilutions, bacterial densities of 1.5 × 10 6 CFU/ml were achieved. Next, the nanocomposite films were cut into pieces with 6 mm diameter under sterile conditions and placed on the surface of Mueller-Hinton agar plates; 100 μl aliquots of the prepared suspensions were inoculated ahead of incubation, which was done overnight at 37 and 25°C for S. aureus and P. fluorescens, respectively. The zone of inhibition surrounding each disk was then evaluated with a digital micrometer, and the film with optimal antimicrobial properties was chosen for covering chicken fillet samples.

| Chicken fillet samples preparation and treatment
The samples of fresh chicken fillets were purchased locally. After immediately being taken to the laboratory, the samples were washed and cut aseptically into squared pieces weighing 10 g. After sterilizing the pieces using ethanol (95% v/v) and UV light, they were inoculated with 10 4 CFU/g of P. fluorescens and S. aureus suspensions.
The samples were divided into the treatment (packaged with the optimized G/5% CNF/5% ZnO film) and control (packaged with sterile transparent plastic polyethylene) groups. Storage occurred at the temperature of 4 ± 1°C over 12 days; microbial characteristics were evaluated initially and then at 3-day intervals.

| Microbial evaluation of bacteria inoculated in chicken fillets
To conduct the microbial analysis, chicken fillet samples weighing 10 g were blended for 5 min with 0.1% sterile peptone water (90 ml

| Statistical analysis
The SPSS 21.0 software (IBM) was utilized for statistical analysis, with values being presented as mean ± standard deviation (SD). All tests were performed in triplicates. The independent t test, analysis of variance (ANOVA), and Tukey's post hoc test were conducted; significance was regarded at p < .05.

| Thickness
The thickness of a film is influenced by the incorporation of fillers into its matrix. Table 1 reveals that the thickness of the neat gelatin film was 0.102 ± 0.005 mm; this value increased to 0.11 ± 0.006 mm after incorporation of up to 7.5% CNFs. Similar results were reported in whey protein isolate-based films when adding CNFs (Alizadeh- Sani, Khezerlou, & Ehsani, 2018).

| Light transmittance and opacity
Film transparency has significance as it directly affects the appearance of coated products, the rate of lipid oxidation, and the quality of the packaged food product. The light (400-800 nm) transmission characteristics of the gelatin-based films were evaluated via UV-vis spectrophotometry. According to the results (Table 2), the highest light transmittance was observed in the pure gelatin film. The G/ CNF composite films became significantly less (p < .05) transparent as the nanocellulose content increased. In the G/CNF composite films, light transmission at 800 nm decreased from 76.02 ± 0.30% to 64.03 ± 0.61% when the CNF component was increased from 0 to 7.5 wt%. The film opacity significantly increased to 1.603 ± 0.057 with 7.5% CNF incorporation (p ˂ .05), indicating decreased transparency. It has been reported that the addition of nanocellulose to proteins causes light transmittance to be lost mostly as a result of refraction/reflection occurring at the interface of the two species, resulting in increased film opacity (Liu, Tang, & Liu, 2015). The findings concur with those of Alizadeh- , who found that whey protein film transparency decreased with the addition of CNF.
Also, a reduction in film transparency with the incorporation of nanocellulose was observed in whey protein isolate-nanocellulose bionanocomposite films (Qazanfarzadeh & Kadivar, 2016).

| Moisture content (MC)
The MC is the amount of water "bound or confined" within a sample.
According to Table 1, the MC of the films stayed constant with the incorporation of 2.5% and 5% CNFs, whereas the CNF concentration of 7.5% induced a rise from 7.94 ± 0.44% net gelatin to 8.13 ± 0.54% (p < .05). This can be attributed to the decreased water binding ability that ensues cross-linking within the nanocomposite films and the trapping of free water molecules via the created network by the main components of film (Yu et al., 2018). The presence of a large number of free OH groups in the gelatin matrix resulted in a relatively high MC value in neat gelatin film .

| Moisture absorption (MA)
The pure gelatin film absorbed the highest amount of moisture after 24 hr (8.30 ± 0.06%), probably because of its hydrophilic amino acids.

| Water vapor permeability (WVP)
For food packaging, water vapor permeability is affected by two factors: the solubility and diffusion of water molecules (Pirsa, Karimi Sani, & Khodayvandi, 2018

| Mechanical properties
Extensibility and mechanical strength are typically required for the maintenance of the physical integrity and barrier properties of film materials in the face of external forces in food packaging applications (Souza et al., 2012). The influence of CNFs on the films' mechanical characteristics is presented in Figure 1.  Wang, Liu, et al., 2017). A similar mechanism was reported for sodium caseinate in a previous study (Ranjbaryan et al., 2019). According to the literature, CNFs improved the mechanical properties of different matrices such as banana starch (Tibolla et al., 2019), carboxymethyl cellulose (Zabihollahi et al., 2020), and triacetate cellulose (Wu, Danh, & Nakagaito, 2020 (Salari et al., 2018;Sarwar et al., 2018).
According to the results of the physicomechanical tests performed in the above sections, the concentration of 5% nanocellulose was selected as the optimal concentration for the next steps.

| Antibacterial activity of G/CNF/ZnO nanocomposite films
Active ingredients such as antioxidants or antimicrobials in packaging films can preserve food quality and safety (Takma & Korel, 2019).
To provide antimicrobial protection for chicken meat, ZnO NPs were incorporated into G/CNF edible packaging films as an active agent.  Table 3. No clear inhibition zones were found in the G/CNF film (control). As mentioned, CNFs were used to expand the physicomechanical properties of the gelatin film. Antibacterial activity against both species was seen in all active nanocomposites containing ZnO NPs; the sizes of the inhibition zones increased as the ZnO NP concentration increased from 1% to 5% of the polymer. The greatest zone of inhibition against both bacteria was seen in nanocomposites containing 5% ZnO NPs (10.44 ± 0.44 mm and 9.75 ± 0.11 mm for S. aureus and P. fluorescens, respectively). It has been reported that the antimicrobial activity of ZnO NPs against microorganisms is related to the production of different reactive oxygen species (ROS), including superoxides, hydroxyl radicals, and H 2 O 2 by Zn 2+ . These ROS destroy the cell membrane of bacteria then make reactions with the cytoplasmic constituents until killing the microorganism (Mizielińska et al., 2018).
Also, interactions occurring between Zn 2+ cations and nucleic acids and other biomolecules that have negative charges (e.g., phosphate, disulfide, and sulfhydryl groups pertaining to enzymes) result in the degradation of bacterial walls, membranes, and proteins, ultimately inducing lysis and death in the bacterial cells (Zhang et al., 2010). The outstanding antimicrobial properties of ZnO NPs and the related mechanisms of action have been documented by other researchers (Espitia et al., 2012;Jahed, Khaledabad, Bari, & Almasi, 2017).
Nonetheless, when increasing the concentration of ZnO NPs from 5% to 7% of the polymer, no further increase was observed in the inhibition zone, probably due to a fall in the diffusion ability of these nanoparticles (Ngo, Dang, Tran, & Rachtanapun, 2018). Interestingly, featured a surface that was smooth, compact, and homogeneous without porosity, proving an ordered film structure (Wang, Liu, et al., 2017). As seen, the G/CNF films had homogeneous surfaces with little roughness, lacking air bubbles and cracks; this indicates appropriate compound mixing. The appropriate aggregation and uniform, highly compact structure of these samples are due to the homogeneous distribution of the CNFs in the gelatin matrix, the electrostatic stabilization induced by the CNFs' superficial anionic carboxyl molecules, and the stable bonds formed between the hydrophilic compounds during drying (Alizadeh- Wang, Liu, et al., 2017). The same results have been reported for the influence of CNFs on other biopolymer films such as sodium caseinate (Ranjbaryan et al., 2019) and starch (Fazeli, Keley, & Biazar, 2018).
The nanocomposite films showed rough and granular surface structures with randomly distributed ZnO NPs. However, significant aggregation was not observed, which indicates that the ZnO NPs were homogeneously distributed through the whole G/CNF matrix. In the SEM images, white dots indicate the occurrence of these ZnO NPs at the polymer surface.

| Fourier transform infrared (FT-IR) spectroscopy
The FT-IR analysis was conducted to identify the functional groups and the interactions between components of composite film ( Figure 4). The peaks situated at the wavenumbers of ~855, ~1,540, ~1,700, ~2,140, ~2,850, and ~3,443 cm −1 were approximately found in all films with similar patterns mid slight changes in some peaks.
Pure gelatin film displayed a major band at 1,714 cm −1 , which is the characteristic peak of amide-I and represents C=O stretching/hydrogen bonding coupled with a COO group (Arfat, Benjakul, Prodpran, Sumpavapol, & Songtipya, 2014;Shankar et al., 2015). The  John, Kanth, & Umapathy, 2015). As it could be seen, some of the peaks are shifted to higher or lower wavenumbers when CNF and ZnO NPs were used. For example, the peak of amide-I slightly shifted to wavenumbers ~1,700 cm −1 and 1,696 cm −1 after incorporation of CNF and ZnO NPs, respectively. Like spectral changes in the amide-I region, a slight change at wavenumbers peaks of amide-II (shift of 1,520 cm −1 peak to 1,536 cm −1 and 1,542 cm −1 peak) was observed when CNF and ZnO NPs were added, respectively (Mohanapriya, Mumjitha, Purnasai, & Raj, 2016;Shankar et al., 2015). These results implied that the CNF and ZnO NPs were engrafted into the amide bonds of gelatin molecular chains via a conjugation process (Hosseini et al., 2015). By the incorporation of CNF and ZnO NPs, the broadness of the peak between 3,000 and 3,600 cm −1 decreased, signifying a decrease in hydrophilicity of gelatin films (Kumar et al., 2019).
The N-H stretching of the amide-A band has transferred from 3,400 cm −1 for gelatin film to 3,443 cm −1 for G/CNF/ZnO NPs film.
Such changes indicated the increased interaction between N-H groups of the protein chain and nanomaterials, mostly hydrogen bonding in the films (Zubair & Ullah, 2020). Gelatin films display a new peak around 600-900 cm −1 which represents the uptake of ZnO by the gelatin (Umamaheswari et al., 2015). Also, two new absorption bands at 926 and 855 cm −1 arise from the C-O-C stretching at the β-(1-4)-glycosidic linkages which related to cellulose nanofiber structure (Samadani et al., 2019). Incorporation of CNFs and ZnO NPs impressed the intermolecular interaction and molecular association in the gelatin film matrix due to the enhanced multiple covalent bonding and hydrogen bonding (Arfat et al., 2017).

| Assessment of inoculated foodborne pathogenic bacteria
The growth rates of S. aureus and P. fluorescens inoculated in chicken fillets packaged within sterile transparent polyethylene (control) and gelatin-based films incorporated with CNF/ZnO NPs (treatment) are shown in Figure 5a,b. As a pathogenic agent that is foodborne, S. aureus causes various diseases. Food poisoning by S. aureus is most commonly related to poultry, egg, red meat, and seafood products (Yuan & Yuk, 2018). According to the results, 3 days after inoculation, the number of S. aureus reached 3.58 ± 0.39 and 1.84 ± 0.21 log CFU/g in the control and treatment samples, respectively. This

F I G U R E 5
The counts of inoculated S. aureus (a) and P. fluorescens (b) in chicken fillet samples packaged within sterile transparent polyethylene (control) and the gelatin-based films incorporated with CNF/ZnO NPs (treatment) during storage at 4°C. Data are presented as mean ± standard deviation (n = 3). Each point represents the mean ± SD. CNFs, cellulose nanofibers; G, gelatin; ZnO NPs, zinc oxide nanoparticles number increased significantly (p < .05) across both samples from day 3 onwards during the storage time, though the control had a faster rate of rising in this parameter. At the end of the storage period, the number of bacteria was 6.58 ± 0.62 and 4.99 ± 0.42 log CFU/g in the control and treatment samples, respectively.
Regarding P. fluorescens, the results also showed that on day 3, a significantly higher count (4.32 ± 0.63 log CFU/g) was found in the control relative to the treatment sample coated with the G/ CNF/ZnO NP film (3.25 ± 0.23 log CFU/g). On day 12, the control sample had a P. fluorescens count of 7.20 ± 0.27 log CFU/g, while the treatment had 6.42 ± 0.38 log CFU/g. In both samples, the

| CON CLUS ION
In the present research, we introduced an eco-friendly, gelatinbased, active nanocomposite film containing CNFs and ZnO NPs.
Gelatin-based films are suitable carriers for reinforcement fillers (e.g., CNFs) and antimicrobial substances (e.g., ZnO NPs). Based on physicomechanical (WVP, WS, MA, MC, TS, YM, and EAB) and antibacterial characterization, the optimal concentrations of CNFs (5%) and ZnO NPs (5%) for addition to the gelatin matrix were ascertained. The TS and YM of the G/CNF film were significantly higher relative to the pure gelatin film. The G/CNF film had significantly improved resistance against water vapor (i.e., less WVP) relative to the gelatin film. Considering the results of the disk diffusion test and bacterial count, the ZnO NPs showed antibacterial effects against both Gram-positive and Gram-negative species, particularly the former. In conclusion, antimicrobial active packaging films containing ZnO NPs can be used as a beneficial solution for preserving the quality and safety of fresh meat products.