Cellulose Nanofibers from Olive Tree Pruning as Food Packaging Additive of a Biodegradable Film

A biodegradable packaging film containing cellulose nanofibers from olive tree pruning, a by-product of olives production, was obtained using a solvent casting method. Nanocellulose was added to polyvinyl alcohol (PVA) to enhance the technological properties of the composite film as food packaging material. Nanocellulose was obtained from unbleached and bleached pulp through a mechanical and TEMPO pretreatment. Crystalline and chemical structure, surface microstructure, UV and gas barrier, optical, mechanical and antioxidant properties, as well as thermal stability were evaluated. Regarding optical properties, the UV barrier was increased from 6% for the pure PVA film to 50% and 24% for unbleached and bleached nanocellulose, respectively. The antioxidant capacity increased significantly in unbleached mechanical nanocellulose-films (5.3%) compared to pure PVA film (1.7%). In terms of mechanical properties, the tensile strength of the 5% unbleached mechanical nanocellulose films was significantly improved compared to the pure PVA film. Similarly, the 5% nanocellulose films had increased the thermal stability and improved barrier properties, reducing water vapor permeability by 38–59% and presenting an oxygen barrier comparable to aluminum layer and plastic films. Our results support the use of the developed films as a green alternative material for food packaging.


Introduction
Food packaging plays an essential role in the quality and safety of foods throughout their shelf life, protecting them from physical, chemical and biological hazards [1]. In recent decades, due to changes in consumer lifestyle, the demand for safe, high-quality, fresh, minimally processed and ready-to-eat foods, which are mainly packaged in single-use plastic packaging, has strongly increased, resulting in a negative environmental impact [2,3].
Plastics are the most widely used material in food packaging, with more than 30% of worldwide production destined for this application. Global demand and the responsive production of plastic materials for food packaging has increased considerably over the last six decades and is expected to continue for the next 20 years [4,5]. Petroleum-based polymeric plastic materials such as polyvinylchloride (PVC), polyethylene terephthalate (PET), polypropylene (PP), polyethylene (PE), polyamide (PA), polystyrene (PS) and ethylene vinyl alcohol (EVOH), are widely used in food packaging due to their good mechanical and barrier properties, low cost and high availability [6]. However, the high negative impact on the environment caused by low degradability of such materials has led to increased global concern; indeed, over ten megatons (Mt) of plastic waste ends up in the oceans [1,7].
Innovations in the food packaging industry have focused on the development of new biodegradable packaging materials, for which demand has increased considerably in past Unbleached (U) and bleached pulp (B) were used to obtain CNFs using two different pretreatments: mechanical (M) and TEMPO-mediated oxidation (T), both followed by a high-pressure homogenisation treatment [15].

Preparation of PVA/(Ligno)nanocellulose Films
The PVA-(L)CNF films were prepared by the solvent casting method. The PVA solution (3 wt%) was dissolved in distilled water at 90 • C for 4 h by mechanical stirring. Three concentrations of the CNF suspension (2.5, 5 and 7.5% (w/w)) were added to the PVA solution, which were mixed under continuous stirring at room temperature for 4 h. The bionanocomposite films were prepared by casting the suspensions into 14 cm diameter Petri plates and drying at room temperature until a dry weight of 0.35 g per film was reached. Finally, the dried films were peeled from the casting surface and preconditioned (23 • C with 50% RH for 24 h) prior to characterization.

Optical Properties
The light transmittance of the film in the UV-VIS regions (200-800 nm) was determined with a Perkin Elmer UV/VIS Lambda 25 spectrophotometer (Waltham, MA, USA), using the following equations to assess the UV barrier (Equation (1)) and transparency properties (Equation (2)) of the films by means of the transmittance values of the films at wavelengths of 280 nm and 660 nm, which are consistent with the UV-VIS absorption spectra [25].
where %T 660 is the transmittance percentage at 660 nm and x is the thickness of the film (mm).

Fourier Transform Infrared (FTIR) Spectroscopy
The spectra of the films were obtained using a FTIR-ATR Perkin-Elmer Spectrum Two (Waltham, MA, USA) with a resolution of 4 cm −1 in the range of 500 to 4000 cm −1 . FTIR spectra were compared to evaluate the effects of nanocellulose incorporated in the PVA films based on the intensity and shift of vibrational bands. A total of 40 scans were collected for each sample.

Thermogravimetric Analysis (TGA)
The thermal stability of the pure PVA and PVA-(L)CNF films was assessed by thermogravimetric analysis (TGA) using a Toledo Thermogravimetric analyzer (TGA/DSC 1) (Mettler Toledo, L'Hospitalet de Llobregat, Barcelona, Spain). The measurements were performed by heating the film samples (10.0 ± 1.0 mg) from room temperature to 600 • C at a heating rate of 10 • C/min under a nitrogen atmosphere with a nitrogen gas flow of 50 mL/min. The temperature at which the degradation rate reached its maximum (T max ), was evaluated by analyzing the TGA equivalent derivate (DTG).

Scanning Electron Microscopy (SEM)
The surface and cross-sectional morphology of the pure PVA and PVA-(L)CNF films were analyzed using a JEOL JSM-7800F Scanning Electron Microscopy (SEM). All film samples were coated using conductive gold sputter at 10 kV.

X-ray Diffraction (XRD) Analysis
The X-ray diffraction patterns of the composites were obtained using a Bruker D8 Discover with a monochromatic source CuKα1 between angular range of 5-50 • at a 1.56 • /min scan speed in order to evaluate the changes in crystallinity after addition of (L)CNFs to PVA films.

Antioxidant Activity
The ABTS scavenging activity assay of samples was determined as described by Espinosa et al. [15]. A radical solution was prepared (7 mM ABTS and 2.45 mM persulphate of potassium) and left in the dark overnight for 12-16 h before use. The radical solution was diluted with ethanol to an absorbance of 0.70 ± 0.02 at 734 nm. An aliquot of this solution (4 mL) was mixed with 1 cm 2 of the films and, after 6 min, the absorbance of the samples was measured with a spectrophotometer at 734 nm. The percentage of reduction of ABTS 734 (antioxidant power (AOP)) was calculated with the Equation (3).
where A ABTS6 is the absorbance at 734 nm of the ABTS radical solution after 6 min, A ABTSfilm6 is the absorbance at 734 nm of the sample after 6 min and A ABTS0 is the initial absorbance at 734 nm of the ABTS radical solution (0 min). ABTS scavenging activity was expressed as %AOP per gram of the film. All assays were performed in triplicate.

Mechanical Properties
An LF Plus Lloyd Instrument (AMETEK Measurement & Calibration Technologies Division, Largo, FL, USA) provided with a 1 kN load cell was used to calculate the tensile strength, elongation at break, traction and Young's modulus of the pure PVA and PVA-(L)CNF films according to the ASTM D638 standard test method [26]. Film samples were cut into 100 × 15 mm strips and then preconditioned at 25 • C for 48 h and 50% RH before the test. The initial distance was set at 65 mm and the stretching rate at 10 mm/min. The thickness of the films at five random positions was measured using a Digital Micrometre IP65 0-1", (Mitutoyo, Neuss, Germany) with a sensitivity of 0.001 mm. At least five samples at each CNF concentration were tested.

Barrier Properties: Water Vapor Permeability and Oxygen Transmission Rate (WVP and OTR)
The Water Vapor Permeability (WVP) of the films was measured according to the ASTM E96/E96M-10 standard method [27]. Square film samples of 4 cm 2 were attached with aluminum adhesive tape to containers, whose lids were punched with a circular cork borer of 10 mm diameter and containing CaCl 2 as desiccant material. Then, containers were placed in a controlled chamber at 25 • C and 50 ± 2% RH. In order to evaluate mass gain, the weight of the containers was measured at different time intervals for 24 h and used to calculate the water vapor transmission rate (WVTR) using Equation (4): where G is the weight change (g) at time t (h) and A is the test area (m 2 ) The WVTR was used to calculate de WVP coefficient according to Equation (5).
where Th is the thickness of the film (m), P is the saturation vapor pressure at the test temperature (Pa), RH out is the relative humidity outside the container (50%) and RH in is the relative humidity inside the container where the CaCl 2 is placed (0%). The Oxygen Transmission Rate (OTR) of the films was measured using a Mocon OX-TRAN ® 2/22 OTR analyser (Mocon, MN, USA) at 23 • C and 50% RH. The test was run according to the ASTM D3985-17 standard method by measuring the steady-state transmission rate of oxygen gas through the bionanocomposite using an oxygen-sensitive coulometric sensor operating at constant performance to monitor the amount of oxygen transmitted. For the measurement, oxygen gas was applied to one side of the barrier material and nitrogen gas was applied to the other side. For this purpose, the test film was placed between the two sides of the test cell. As the oxygen gas penetrated the film, the nitrogen carrier gas was transported into the coulometric detector [28].

Statistical Analysis
In order to evaluate the influence of the bleaching process, pretreatment and CNF concentration in the film's formulation, one-way analysis of variance (ANOVA) and Tukey's post hoc test were carried out using the IBM ® SPSS ® Statistics software Version 25 (IBM Corporation, New York, NY, United States), with a significance level of p < 0.05. All data are reported as mean ± standard deviation.

Structure Characterization
The micrographs obtained by SEM showed the microstructure and dispersion of the nanocellulose in the PVA film.
As can be seen in Figure 1a, the scanning electron micrographs of the PVA film's surface containing 7.5% of nanocellulose showed a distribution pattern rougher than that of the pure PVA film. It was observed that the higher the nanocellulose concentration, the higher the surface roughness of the PVA-(L)CNF films, due to the increased agglomeration of the nanofibers. In relation to the lignin content, it was observed that films with unbleached nanocellulose (MU and TU), due to their lignin content, showed greater roughness than films obtained with bleached nanocelluloses (MB and TB) [29]. According to the type of pretreatment, i.e., mechanical or TEMPO, differences were observed depending on the presence or not of lignin in the nanocellulose fibers. While the mechanical treatment resulted in films with smoother surfaces when lignin was present (MU vs. TU), the TEMPO pretreatment yielded smoother film surfaces in the case of bleached fibers (TB vs. MB). In both cases, this surface texture was probably the result of the lower diameter of the nanofibers, since in mechanical pretreatment the presence of lignin resulted in a higher fibrillation yield, whereas in TEMPO pretreatment the absence of lignin led to more effective oxidation [15,23].
The distribution and integration of the nanocellulose in the PVA matrix were evaluated by means of a cross section analysis. As can be observed in Figure 1b, the nanocellulose was homogeneously distributed with no (L)CNF precipitation zones observed during the dissolving-casting process, showing a well dispersed distribution pattern with no layer separation. Figure 2 shows the FTIR spectra of pure PVA and 7.5% PVA-(L)CNF films, revealing that the spectral pattern of pure PVA was very similar to the spectra of the films reinforced with 7.5% CNF. This fact would indicate that the addition of CNF had no influence on the molecular structure of the PVA, with the chemical structure remaining stable, with no changes.  Figure 2 shows the FTIR spectra of pure PVA and 7.5% PVA-(L)CNF films, revealing that the spectral pattern of pure PVA was very similar to the spectra of the films reinforced with 7.5% CNF. This fact would indicate that the addition of CNF had no influence on the molecular structure of the PVA, with the chemical structure remaining stable, with no changes.

Chemical Structure
All spectra showed peaks at 3250 cm −1 attributed to the typical O-H stretching vibration from intermolecular and intramolecular hydrogen bonds. The peak corresponding to the C-H stretching vibrations of methyl or methylene groups was observed at around 2930 boxyl groups, and the C=C of the lignin aromatic rings, respectively. The peak at 1615 cm −1 was more intense in the films reinforced by TEMPO CNF (TU and TB) as a consequence of carboxyl groups enhancement by the action of TEMPO oxidation, while the peak at 1514 cm −1 was less intense in the films reinforced with TU, TB and MB CNF, owing to the oxidation of lignin in TEMPO pretreatment and removal in the bleaching step [15,31]. 5% of mechanical unbleached nanocellulose (7.5% MU), PVA containing 7.5% of mechanical bleached nanocellulose (7.5% MB), PVA containing 7.5% of TEMPO unbleached nanocellulose (7.5% TU) and PVA containing 7.5% of TEMPO bleached nanocellulose (7.5% TB)).

Crystalline Structure
To evaluate the crystalline and amorphous regions of pure PVA and 7.5% PVA-(L)CNF films, X-ray diffraction patterns were studied ( Figure 3). The pure PVA film showed the typical strong semicrystalline structure of PVA, showing an X-ray diffractogram with a typically strong crystalline signal at 19.7° due to hydrogen bonds between the hydroxyl groups of the PVA chains, and a broad signal at 19.5° corresponding to the amorphous section [14].
The 7.5% PVA-(L)CNF films showed signals with higher intensity between 19.4° and 19.7°, similar to that of pure PVA, indicating that the incorporation of (L)CNFs did not modify the semicrystalline structure of PVA. It was also confirmed that the addition of CNF did not show significant changes in the position of the peaks. In olive tree pruning (L)CNFs, the known diffraction signal at 16.1° and 22.5°, corresponding to the 110 and 200 reflection planes of cellulose I structure [16,23], practically disappeared in the composite films of this study. ATR-FTIR spectra of pure PVA and PVA-(L)CNF films (PVA containing 7.5% of mechanical unbleached nanocellulose (7.5% MU), PVA containing 7.5% of mechanical bleached nanocellulose (7.5% MB), PVA containing 7.5% of TEMPO unbleached nanocellulose (7.5% TU) and PVA containing 7.5% of TEMPO bleached nanocellulose (7.5% TB)).
All spectra showed peaks at 3250 cm −1 attributed to the typical O-H stretching vibration from intermolecular and intramolecular hydrogen bonds. The peak corresponding to the C-H stretching vibrations of methyl or methylene groups was observed at around 2930 cm −1 . The vibration peak detected at 1425 cm −1 has been related to the bending mode of CH 2 bonds. The peaks at 1330 and 1084 cm −1 were associated with C-O stretching, the vibration around 920 cm −1 characterized the CH 2 groups, while the absorption band at 840 cm −1 is related to the stretching vibration of the C-C groups [13,30].
The slight differences found between the spectra of pure PVA and the PVA-(L)CNF films were due to the different chemical composition of the nanocelluloses and were observed in the 1615 and 1514 cm −1 peaks associated with C=O stretching vibration in carboxyl groups, and the C=C of the lignin aromatic rings, respectively. The peak at 1615 cm −1 was more intense in the films reinforced by TEMPO CNF (TU and TB) as a consequence of carboxyl groups enhancement by the action of TEMPO oxidation, while the peak at 1514 cm −1 was less intense in the films reinforced with TU, TB and MB CNF, owing to the oxidation of lignin in TEMPO pretreatment and removal in the bleaching step [15,31].

Crystalline Structure
To evaluate the crystalline and amorphous regions of pure PVA and 7.5% PVA-(L)CNF films, X-ray diffraction patterns were studied ( Figure 3). The pure PVA film showed the typical strong semicrystalline structure of PVA, showing an X-ray diffractogram with a typically strong crystalline signal at 19.7 • due to hydrogen bonds between the hydroxyl groups of the PVA chains, and a broad signal at 19.5 • corresponding to the amorphous section [14].

Optical Properties
UV light in the range of 200 to 280 nm is one cause of lipid oxidation of foodstuffs, which is a problem for the food industry. Therefore, the development of films with barrier properties against UV light would help to prevent, or at least slow down, UV light lipid oxidation [32].
In this work, the optical properties of pure PVA and PVA-(L)CNF films were determined by measuring the transmittance of light in the UV and visible ranges in order to analyze the transparency and UV barrier of the films. As can be seen in Figure 4a,b, the effect of CNF type, CNF concentration and their interaction on the UV barrier was statistically significant (p < 0.05), as was transparency, except for their interaction (p > 0.05).
To a large extent, the optical transmittance of the films depends on the dispersion of the nanocellulose in the PVA matrix. As shown in Figure 4a, the films reinforced with CNF were less transparent than the pure PVA film, with a minimum transparency value of 34.5% in 7.5% MU PVA-CNF film versus 96.1% for pure PVA (p < 0.05). Regarding the type of CNF, it can be observed that the addition of unbleached CNF obtained by TEMPO pretreatment (TU) in the PVA matrix led to films with lower transparency than the bleached ones (TB) (p < 0.05). This reduction in transparency was likely due to the slight brown color of the unbleached TEMPO PVA-(L)CNF films, which contains opaque lignin [29]. However, those obtained by mechanical pretreatment (MU and MB) did not show significantly different transparency values (p > 0.05). In relation to the concentration of CNF, the transparency values were higher at lower concentration (2.5%), but without significant differences at the remaining concentration (5% and 7.5%) (p > 0.05).
In contrast, as can be observed in Figure 4b, the pure PVA film exhibited a lower UVbarrier (5.7%) than the PVA-(L)CNF films (p < 0.05). Overall, the incorporation of higher concentrations of CNF in the PVA matrix increased the capacity of UV absorption, with the 7.5% TU PVA-CNF film exhibiting a maximum value of 48.8% (p < 0.05). The UVbarrier values of the unbleached PVA-CNF films (MU and TU) were higher than the bleached ones (MB and TB), showing their greater UV absorption capacity (p < 0.05). The enhanced UV light blocking of the unbleached PVA-CNF films can be explained by the strong UV-absorption behavior of the lignin, mainly by the chromophore groups present in the lignin [33]. The 7.5% PVA-(L)CNF films showed signals with higher intensity between 19.4 • and 19.7 • , similar to that of pure PVA, indicating that the incorporation of (L)CNFs did not modify the semicrystalline structure of PVA. It was also confirmed that the addition of CNF did not show significant changes in the position of the peaks. In olive tree pruning (L)CNFs, the known diffraction signal at 16.1 • and 22.5 • , corresponding to the 110 and 200 reflection planes of cellulose I structure [16,23], practically disappeared in the composite films of this study.

Optical Properties
UV light in the range of 200 to 280 nm is one cause of lipid oxidation of foodstuffs, which is a problem for the food industry. Therefore, the development of films with barrier properties against UV light would help to prevent, or at least slow down, UV light lipid oxidation [32].
In this work, the optical properties of pure PVA and PVA-(L)CNF films were determined by measuring the transmittance of light in the UV and visible ranges in order to analyze the transparency and UV barrier of the films. As can be seen in Figure 4a,b, the effect of CNF type, CNF concentration and their interaction on the UV barrier was statistically significant (p < 0.05), as was transparency, except for their interaction (p > 0.05).

Antioxidant Activity
The antioxidant activity of the films is an important functional feature as it can prevent the oxidative spoilage of foods through packaging [34]. Figure 5 shows the antioxidant activity of pure PVA and PVA-(L)CNF films measured by the ABTS assay and ex- To a large extent, the optical transmittance of the films depends on the dispersion of the nanocellulose in the PVA matrix. As shown in Figure 4a, the films reinforced with CNF were less transparent than the pure PVA film, with a minimum transparency value of 34.5% in 7.5% MU PVA-CNF film versus 96.1% for pure PVA (p < 0.05). Regarding the type of CNF, it can be observed that the addition of unbleached CNF obtained by TEMPO pretreatment (TU) in the PVA matrix led to films with lower transparency than the bleached ones (TB) (p < 0.05). This reduction in transparency was likely due to the slight brown color of the unbleached TEMPO PVA-(L)CNF films, which contains opaque lignin [29]. However, those obtained by mechanical pretreatment (MU and MB) did not show significantly different transparency values (p > 0.05). In relation to the concentration of CNF, the transparency values were higher at lower concentration (2.5%), but without significant differences at the remaining concentration (5% and 7.5%) (p > 0.05).
In contrast, as can be observed in Figure 4b, the pure PVA film exhibited a lower UV-barrier (5.7%) than the PVA-(L)CNF films (p < 0.05). Overall, the incorporation of higher concentrations of CNF in the PVA matrix increased the capacity of UV absorption, with the 7.5% TU PVA-CNF film exhibiting a maximum value of 48.8% (p < 0.05). The UV-barrier values of the unbleached PVA-CNF films (MU and TU) were higher than the bleached ones (MB and TB), showing their greater UV absorption capacity (p < 0.05). The enhanced UV light blocking of the unbleached PVA-CNF films can be explained by the strong UV-absorption behavior of the lignin, mainly by the chromophore groups present in the lignin [33].

Antioxidant Activity
The antioxidant activity of the films is an important functional feature as it can prevent the oxidative spoilage of foods through packaging [34]. Figure 5 shows the antioxidant activity of pure PVA and PVA-(L)CNF films measured by the ABTS assay and expressed as percentage of antioxidant power (AOP). The effect of CNF type and concentration was statistically significant (p < 0.05), whereas the interaction of both was not (p > 0.05). The AOP values of the unbleached CNF-reinforced films obtained by mechanical pretreatment (2.5, 5 and 7.5% MU) and unbleached 7.5% CNF TEMPO (7.5% TU) were significantly higher (p < 0.05) compared to pure PVA film, the latter showing the lowest value (1.7% AOP/mg film), and the 7.5% MU PVA-CNF film the highest value (5.29% AOP/mg film). These results suggest that the higher residual lignin content present in unbleached mechanical nanocellulose results in its higher antioxidative activity, mainly attributed to its phenolic hydroxyl groups, which would reduce the ABTS radical through electron transfer. In addition, the mechanical pretreatment would not affect the oxidation status of the composite material in comparison with the TEMPO pretreatment, favoring the antioxidant capacity of the former [35]. Although TEMPO pretreatment is carried out to perform a selective oxidation of the OH groups of cellulose, the sodium hypochlorite used as an initiating agent for catalytic oxidation can produce secondary and undesired oxidation of the lignin present in the fiber, which would cause a solubilization of the lignin, and therefore a bleaching of the fiber, reducing the lignin content in the fiber and thus the antioxidant capacity of the material. This would explain why the TU PVA-CNF films, despite having residual lignin content, exhibited a similar range to the MB PVA-CNF films in terms of antioxidant activity. Espinosa et al. [15] described that PVA films reinforced with wheat straw (ligno)nanocellulose showed up to twice the AOP compared to pure PVA film, values similar to those described in this work, while other authors reported how the incorporation of TEMPO-bleached CNF from eucalyptus pulp into a mucilage-based composite could significantly improve the antioxidant characteristics of the film almost three-fold compared to control [36].

Mechanical Properties
Since the use of PVA for food packaging applications is limited due to its poor mechanical properties, the effect of incorporating different (L)CNFs as reinforcement of PVA films was studied. As illustrated in Figure 6, in general terms, the mechanical properties (Young´s modulus, tensile strength, traction and elongation) were improved by the incorporation of the different (L)CNFs in the PVA matrix, but not significantly. A significant enhancement was observed in the tensile strength of the 5% MB PVA-CNF film, which can be explained by the compact structure of PVA, the stiffness of the nanocellulose chain, the homogeneous distribution of the CNF in the PVA matrix and the strong interaction and hydrogen bonding between OH groups of the nanocellulose and the PVA [14].
The type of nanocellulose significantly affected the Young's modulus and elongation at break values (p < 0.05), as opposed to tensile strength and traction (p > 0.05). The incorporation of different concentrations of CNF was significant except for elongation at break, while the interaction of nanocellulose type and percentage was not significant (p > 0.05), except for tensile strength.
Regarding Young's modulus (Figure 6a), the mean value of pure PVA film was 3578 MPa. Higher mean values were observed in films with 5% CNF TB and MB (4263 and 4229 MPa, respectively), although the differences with PVA were not significant (p > 0.05). Figure 6b shows the tensile strength value of the films assayed, with the 5% MB PVA-CNF film as the only bionanocomposite exhibiting a significant increase over the pure PVA film (69.8 MPa versus 52.5 MPa). With reference to traction and elongation at break, none of the films showed significantly improved values, although slightly higher mean values were observed in the 5% MB PVA-CNF film for traction (35.4 N), and in the 7.5% TB PVA-CNF for elongation at break value (143%) (Figure 6c,d).
In general, the slightly higher values of mechanical properties in MU and MB PVA-CNF films could be attributed to the high aspect ratio of these (L)CNFs. In terms of concentration, a decreasing trend in the mechanical properties of the films was observed when the (L)CNF loading was higher than 5%. This behaviour at concentrations over 5% can be explained by the higher agglomeration rate of nanocellulose in PVA-(L)CNF films acting as weak points of the films due to the breakdown of the interaction between CNF and the PVA matrix [37].

Mechanical Properties
Since the use of PVA for food packaging applications is limited due to its poor mechanical properties, the effect of incorporating different (L)CNFs as reinforcement of PVA films was studied. As illustrated in Figure 6, in general terms, the mechanical properties (Young s modulus, tensile strength, traction and elongation) were improved by the incorporation of the different (L)CNFs in the PVA matrix, but not significantly. A significant enhancement was observed in the tensile strength of the 5% MB PVA-CNF film, which can be explained by the compact structure of PVA, the stiffness of the nanocellulose chain, the homogeneous distribution of the CNF in the PVA matrix and the strong interaction and hydrogen bonding between OH groups of the nanocellulose and the PVA [14].

Barrier Properties
Determination of barrier properties is crucial in the development of food packaging. The type of specific barrier needed depends on the type of food commodity. For example, The type of nanocellulose significantly affected the Young's modulus and elongation at break values (p < 0.05), as opposed to tensile strength and traction (p > 0.05). The incorporation of different concentrations of CNF was significant except for elongation at break, while the interaction of nanocellulose type and percentage was not significant (p > 0.05), except for tensile strength.
Regarding Young's modulus (Figure 6a), the mean value of pure PVA film was 3578 MPa. Higher mean values were observed in films with 5% CNF TB and MB (4263 and 4229 MPa, respectively), although the differences with PVA were not significant (p > 0.05). Figure 6b shows the tensile strength value of the films assayed, with the 5% MB PVA-CNF film as the only bionanocomposite exhibiting a significant increase over the pure PVA film (69.8 MPa versus 52.5 MPa). With reference to traction and elongation at break, none of the films showed significantly improved values, although slightly higher mean values were observed in the 5% MB PVA-CNF film for traction (35.4 N), and in the 7.5% TB PVA-CNF for elongation at break value (143%) (Figure 6c,d).
In general, the slightly higher values of mechanical properties in MU and MB PVA-CNF films could be attributed to the high aspect ratio of these (L)CNFs. In terms of concentration, a decreasing trend in the mechanical properties of the films was observed when the (L)CNF loading was higher than 5%. This behaviour at concentrations over 5% can be explained by the higher agglomeration rate of nanocellulose in PVA-(L)CNF films acting as weak points of the films due to the breakdown of the interaction between CNF and the PVA matrix [37].

Barrier Properties
Determination of barrier properties is crucial in the development of food packaging. The type of specific barrier needed depends on the type of food commodity. For example, in most foods, an effective barrier against moisture and oxygen will increase the quality and shelf-life of a food product [38].
The water vapor and oxygen barrier properties of pure PVA and 5% PVA-(L)CNF films were evaluated. As can be seen in Table 1, PVA-(L)CNF films showed a lower water vapor permeability (WVP) (p < 0.05) compared to pure PVA films. The type of nanocellulose incorporated as reinforcement influenced the reduction of WVP (p < 0.05). In relation to the bleaching treatment, it was observed that the WVP in the unbleached PVA-CNF films (MU and TU) was higher than in the bleached ones (MB and TB), while in terms of the influence of the pretreatment, the nanocellulose reinforced films obtained by TEMPO (TU and TB) showed the lowest WVP values compared to the mechanical nanocellulose reinforced films (MU and MB). A drastic reduction of WVP was evident in the 5% TB PVA-CNF film, which showed the lowest WVP value (2.82 × 10 −7 g/s·m·Pa) compared to the pure PVA film (6.97 × 10 −7 g/s·m·Pa) (p < 0.05). The reduction of WVP in PVA-(L)CNF films could be attributed to the network formed through hydrogen bonds between PVA and (L)CNFs reducing the free space in the PVA matrix and sealing the gaps to the passage of water vapor across the film [39]. However, the presence of lignin could hinder these bonds, creating more hydrophobic pores that would facilitate water vapor passage through the film, which would explain the poor aptitude as a barrier to water vapor of LCNF (MU and TU) compared to CNF (MB and TB) [40]. As commented previously, the oxygen permeability of food packaging is highly important for food preservation, extending its shelf-life and preserving its quality. The oxygen transmission rate (OTR) values showed that PVA films reinforced with 5% CNF were impermeable to oxygen, i.e., with a high oxygen barrier. While the pure PVA film had a value of 3.75 cc/m 2 ·24 h, the films reinforced with MU, TU and TB CNF showed a high oxygen barrier comparable to an aluminum layer, whereas the film reinforced with MB CNF presented a good oxygen barrier, comparable to plastic films such as PET AlO X and PET SiO X . The lower oxygen permeability of PVA-(L)CNF films may be due to the inherent flexibility of CNF, which are able to form a denser film, sealing most of the spaces between fibrils where oxygen molecules would penetrate more slowly and with more difficulty [40]. In addition, the presence of lignin would contribute to a higher compaction of the film structure, making it less porous and less permeable to oxygen, which would account for the higher oxygen barrier of the unbleached PVA-CNF films (MU and TU) compared to the bleached ones (MB and TB) [15]. It is important to highlight that (L)CNFreinforced films obtained by mechanical pretreatment (MU and MB), because are more susceptible to environmental conditions such as humidity and temperature, required a longer stabilization time than (L)CNF-reinforced films obtained by TEMPO pretreatment (TU and TB). This behavior occurs in other materials such as EVOH, which are greatly affected by the ambient humidity. To overcome this limitation, this material must be sandwiched by creating a barrier layer through lamination or coextrusion, thus keeping its barrier function [41,42].

Thermal Stability
The evaluation of thermal properties is necessary to estimate the practical applications of PVA-(L)CNF films for use in the agrifood industry. The thermal stability of pure PVA and 5% PVA-(L)CNF films was measured by calculating the maximum degradation temperature (T max ) through the derivative of an TGA curve (DTG). As can be seen in Table 2, PVA showed the lowest maximum degradation temperature (250.32 • C), while the film reinforced with 5% MU PVA-CNF exhibited the highest maximum degradation temperature (264.99 • C). It was observed that the absence of lignin (MB and TB) resulted in films with lower T max than the MU and TU films, while the type of nanocellulose pretreatment did not influence the T max value since very similar values were reported for the MU-TU and MB-TB films. The influence of lignin, present in the unbleached samples, on T max , may be explained by the covalent bonds with cellulose providing high thermal stability to the fibers [43].

Conclusions
PVA films reinforced with (L)CNFs from olive tree pruning waste were obtained by a solvent casting method. Two different pretreatments (mechanical and TEMPO) were used on bleached and unbleached cellulose pulps to isolate nanocellulose. The incorporation of (L)CNFs did not modify the physical and chemical structure of the PVA matrix due to the homogeneous dispersion of the (L)CNFs within the matrix. However, antioxidant activity increased significantly after the addition of unbleached mechanical CNFs due to the higher lignin content. Similarly, the UV barrier properties were enhanced significantly with the incorporation of all types of (L)CNFs, resulting in films with lower transparency but higher UV-light blocking capacity. In addition, the introduction of nanocellulose improved water vapor and oxygen barriers, decreasing their values, with a drastic fall of the OTR. The thermal stability of 5% PVA-(L)CNF films was also higher in comparison to pure PVA. The addition of (L)CNFs slightly increased the values of the mechanical parameters measured, with the addition of 5% mechanical bleached nanocellulose (MB) causing a significant improvement in terms of tensile strength. The films developed in this work could be potentially employed in the food packaging sector, providing additional technological benefits. The barrier properties against UV, water vapor and oxygen, together with the proven antioxidant activity necessary for preservation of several food commodities, including the prevention of lipids oxidation in foods, would confer on the developed bionanocomposites the optimum functionalities for food stabilization during shelf-life. Further studies should be considered to investigate the mechanisms of biodegradation as well as the possible migration of film components into food.

Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.