Improving Hydrophobicity and Water Vapor Barrier Properties in Paper Using Cellulose Nanofiber-Stabilized Cocoa Butter and PLA Emulsions

Abstract

This study explores the use of cellulose nanofiber (CNF)-stabilized Pickering emulsions for paper coatings, focusing on their rheological properties and effects on hydrophilicity and water vapor transmission rate (WVTR). Two types of Pickering emulsions, oil-in-water (O/W), were stabilized with 1 wt% CNF extracted from fique by-products. The oily phases of the emulsions were composed of poly(lactic acid) (PLA) and cocoa butter (CB). The physical stability, viscosity, and viscoelasticity of the emulsions were characterized. The emulsions were applied to the surfaces of Bond and Kraft papers using the rod-coating method. The coating process involved first applying a layer of the PLA emulsion followed by a layer of the CB emulsion. The coated papers were then evaluated by FE-SEM, contact angle, adhesion work, and water vapor transmission rate (WVTR). The results indicated that the coatings effectively produced a slightly hydrophobic surface on the papers, with contact angles approaching 90°. Initially, Kraft paper exhibited a WVTR value of 29.20 ± 1.13 g/m²·h, which significantly decreased to 7.06 ± 2.80 g/m²·h after coating, representing a reduction of 75.82%. Similarly, natural Bond paper showed a WVTR value of 30.56 ± 0.34 g/m²·h, which decreased to 14.37 ± 5.91 g/m²·h after coating, indicating a reduction of 47.02%. These findings demonstrate the potential of CNF-stabilized Pickering emulsions for enhancing the performance of paper coatings in terms of hydrophobicity and moisture barrier properties. The approach of this study aligns with global sustainability goals in packaging materials combining the use of PLA and CB to develop a waterborne coating to enhance the moisture barrier properties, demonstrated by a substantial reduction in water vapor transmission rates, and an improved hydrophobicity of coated papers.

1. Introduction

Currently, various food and non-food products are packaged using paper and cardboard. These materials account for 36% of the global packaging and container market; in 2022, 265 million metric tons were produced exclusively for this type of application [1]. However, due to the hygroscopic nature of paper, it must be laminated with materials that enhance its barrier properties against water, oxygen, and grease [2]. Among the compounds used, there are different synthetic polymers, such as polyethylene, poly(ethylene terephthalate), poly(butylene terephthalate), among others, to improve its water or water vapor barrier properties [3]. Although lamination offers technical advantages, it compromises the recyclability of the resulting material and reduces its biodegradability [4].

In this way, further use of paper as a packaging material involves replacing the components used as coatings with alternatives that have a lower environmental impact. Among the possibilities are bio-based materials, which offer biodegradability, biocompatibility, and sustainability [5,6]. Many studies have attempted to improve the moisture resistance of paper and paperboard with these materials [5].

Some of the bio-based materials investigated in recent years for the development of these types of coatings are based on nanocomposites, which exhibit high barrier properties due to their large surface area and distribution within the coating [7]. Abidi et al. reported that the addition of 5 wt% cellulose nanocrystals decreased the water vapor permeability of pure poly(vinyl alcohol) films from 0.61 ± 0.04 g·mm/kPa·h·m² to 0.44 ± 0.01 g·mm/kPa·h·m² [8]. Among some nanomaterials that exhibit a high potential are chitosan [9], cellulose nanofibers [10], and hydrophobized cellulose nanofibers [11].

Specifically, cellulose is a natural polymer, a polysaccharide that can be obtained from various plant sources and agro-industrial by-products, such as fique bagasse, which is an abundant and low-cost agricultural by-product in Colombia. The fique plant produces a significant amount of bagasse, traditionally considered waste. In 2020, 35,675 tons of fique yielded 1427 tons of fiber and 34,248 tons of by-products. Fiber accounts for just 4% of the leaf mass, with 96% being fique juice and bagasse, rich in cellulosic components [12].

However, given the hydrophilic nature of cellulose, some natural solid crystalline lipids, such as beeswax and carnauba wax [13], and biopolymers, such as poly(lactic acid) (PLA) [14] and chitosan [15], have been implemented to reduce the water vapor permeability of the coating. G. Bayés et al. report the development of emulsions with beeswax and anionic cellulose nanofibers that improve the water barrier properties of coated papers [16]. While Z. Song et al. report the use of modified cellulose nanofibers and PLA to decrease the water vapor permeability (P) of papers coated with a combination of these materials [14].

Aliphatic polyesters, such as PLA, play a crucial role in developing biodegradable coatings due to their high biodegradability and biocompatibility [17]. In addition, melting temperatures between 160 and 180 °C make it suitable for melt processing and coating technologies. However, the brittleness of PLA limits its use in paper coatings [18]. Blending PLA with other materials like cellulose nanofibers (CNF) or cocoa butter (CB) can make it more suitable for commercial applications and enhance its barrier performance [19]. On the other hand, lipid-based films and coatings, due to their low polarity, have a hydrophobic character and are highly effective at reducing moisture loss [20]. Cocoa butter, a plant-based lipid, is widely used in food and cosmetic applications [21]; in this study, it has been explored in the development of a composite hydrophobic film. Moreover, both PLA and CB are recognized as safe for use in food contact applications [22], making them ideal candidates for sustainable packaging development.

The combination of PLA and CB to develop a coating is not sufficient to meet the goal of developing sustainable coating materials. Traditional solvent casting methods are being replaced by waterborne processes to avoid toxic solvents and environmental harm [19]. Water is a safer and non-toxic alternative; however, its use is limited by the solubility constraints of many polymers [19]. For this reason, Pickering oil-in-water emulsions of PLA and CB were prepared to develop a hydrophobic coating.

In this study, CNF derived from fique bagasse was used to develop aqueous coating suspensions with CB and PLA. The coatings were created through the layered application of two Pickering emulsions: the first consisting of CNF/PLA and the second of CNF/CB. The particle size, physical stability, physicochemical properties, and rheological characteristics of the emulsions were analyzed. The morphology of the papers was evaluated using FE-SEM, their water vapor resistance was tested according to the standard ASTM E96, and wettability was assessed through contact angle measurement.

2. Materials and Methods

2.1. Materials

Cellulose nanofibers (CNF) were obtained from fique bagasse using the modified KOH5 methodology proposed by Zuluaga et al. [23]. Specifically, the process began with an alkaline treatment using 5 wt% KOH for 14 h. The second step involved treating the sample with 1 wt% NaClO₂ for 2 h at 70 °C. A second alkaline treatment with 5 wt% KOH was then performed for 14 h, followed by washing. Lastly, the sample was treated with 1 wt.% HCl at 70 °C for 1 h to remove residual minerals. Between each step, samples were washed until a neutral pH was achieved. Finally, the sample was mechanically processed using high-shear milling (Masuko Sangyo, Supermasscolloider, Kawaguchi, Japan) to homogenize the nanofiber size, following the G30 protocol proposed by Velásquez-Cock et al. [24].

On the other hand, Dichloromethane CH₂ Cl₂ (DCM) (Merck, Darmstadt, Germany), and poly(lactic acid) 2003D (PLA Ingeo^TM NatureWorks^®) (Plymouth, MN, USA) M_w 180,477 g/mol [25] were acquired through a local supplier. Cocoa butter was donated by Compañía Nacional de Chocolates^® (Medellín, Colombia). Finally, the papers to be coated were commercial Bond paper and Kraft paper of 67 ± 2.3 g/m² and 90 ± 1.3 g/m², respectively, purchased from a local store. Type III distilled water was used in all experiments.

2.2. Preparation of Pickering Emulsions Stabilized with CNF

To coat the paper samples, Pickering emulsions stabilized with CNF were prepared. Two oil phases were used: PLA dissolved in DCM and melted CB along with a constant CNF concentration of 1 wt.% as the aqueous phase [26]. The oil-phase-to-aqueous-phase ratio was kept constant at 1:4.

The aqueous phase was prepared by diluting a concentrated suspension of CNF with type III distilled water. Meanwhile, the oil phase was prepared by solubilizing 50 g of PLA in 450 g of DCM at 21 °C [27,28] and then stored under refrigeration at 4 °C until use. Separately, cocoa butter was melted for 2 h at 110 °C to eliminate its crystalline memory [29]. The emulsions were prepared by adding the aqueous phase to the oil phase and homogenizing using an Ultra-turrax^® T-25 equipped with an S25N-18G dispersion unit (IKA-Werke, Staufen im Breisgau, Germany) at 24,000 rpm for 3 min. Then, they were taken to an ultrasonic bath (Elma Schmidbauer GmbH, Singen, Germany) for 3 min at 37 Hz, with 30% power [30].

The obtained emulsions, both CNF/PLA and CNF/CB, were stored in cylindrical glass vials for the characterizations described below.

2.3. Particle Size Distribution

Particle size distribution measurements of nanocellulose suspension and Pickering emulsions were performed using a Mastersizer 3000 (Malvern instruments Ltd., Malvern, UK) equipped with a Hydro MV wet dispersion unit and a blue light source at 470 nm. The samples were dispersed in water until an obscuration of about 8% was reached [31] and a constant agitation of 1230 rpm. Ten measurements per sample were performed when analyzing the emulsions acording to ISO 13320:2020 [32]. Refractive indices of 1.467 and 1.456 were used for PLA and cocoa butter, respectively [33]. A refractive index of 1.493 was used for the nanocellulose suspension [34].

To determine the polydispersity, Span was calculated using Equation (1), where D₉₀, D₅₀ and D₁₀ are the cumulative volumetric diameters at 10%, 50%, and 90%, respectively [35].

$$\mathrm{S}\mathrm{p}\mathrm{a}\mathrm{n}={\displaystyle \frac{{\mathrm{D}}_{90}-{\mathrm{D}}_{10}}{{\mathrm{D}}_{50}}}$$

(1)

2.4. Rheological Behavior

For the rheological analysis of both Pickering emulsions and the nanocellulose suspension, a stress-controlled rheometer Discovery HR-2^® rheometer (T.A. Instruments, New Castle, DE, USA) was used with a geometry of 40 mm parallel flat plates, with a separation between plates of 700 µm and a solvent trap. An amount of 0.9 mL of sample was applied, and all measurements were made at 20 °C. Viscosity tests were performed by logarithmic sweeps from 0.1 s⁻¹ to 100 s⁻¹ [36].

Viscoelastic amplitude measurements were performed at a frequency of 6.28 rad/s, with sweeps between 0.01% and 100% [36]. Finally, logarithmic frequency sweeps were performed with a stress fixed in the linear zone chosen from the amplitude sweep, while angular frequency was modified from 6.28 to 62.80 rad/s.

2.5. Coatings Application

Bond and Kraft paper were cut to dimensions of 5 cm × 5 cm and dried at 80 °C for 24 h to remove humidity before use. Coating application was carried out in two stages. In the first stage, 0.5 mL of previously stirred CNF/PLA emulsion was applied over the paper using a #10 Mayer bar, which allowed for a homogeneous distribution. The coated paper was vacuum dried (Vaciotem-T, JP Selecta, Barcelona Spain) at 40 °C for 24 h; after this process, it was pressed at 160 °C for 3 min at 20 MPa [37,38]. Since this step is well above the boiling temperature of DCM (around 40 °C), no residual solvent is expected to remain in the final coating.

Once pressed, a layer of the CNF/CB emulsion was applied using the same conditions as the first CNF/PLA emulsion. Both emulsions were prepared one to two days prior to the coating process. The CNF/CB emulsion was heated in a water bath at 60 °C immediately before use. Dried and pressed samples were stored in a desiccator until further analysis.

According to the literature, applying a lipid coating first could lead to potential failure in the mechanical properties of paper [18]. Therefore, the CNF/PLA coating was applied initially.

The final grammage of the papers was determined by Equation (2), and the weight of the coating was obtained by subtracting the weight of the paper before and after coating.

$$\mathrm{G}\mathrm{r}\mathrm{a}\mathrm{m}\mathrm{m}\mathrm{a}\mathrm{g}\mathrm{e}={\displaystyle \frac{\mathrm{c}\mathrm{o}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{n}\mathrm{g} \mathrm{w}\mathrm{e}\mathrm{i}\mathrm{g}\mathrm{h}\mathrm{t}}{\mathrm{p}\mathrm{a}\mathrm{p}\mathrm{e}\mathrm{r} \mathrm{a}\mathrm{r}\mathrm{e}\mathrm{a}}}$$

(2)

2.6. Field Emission Scanning Electron Microscopy (FESEM)

The papers were observed using a field emission scanning electron microscope, FE-SEM (Thermo Fisher Scientific Apreo 2-S, Waltham, MA, USA). Samples were stored in a desiccator and subsequently placed on stubs for further observation at magnifications of 500× and 2000×. The samples were observed before and after coating with CNF/PLA—CNF/CB.

2.7. Contact Angle Measurements

To evaluate changes in the surface hydrophilicity of the coated papers, contact angle measurements were performed. A flat surface of each sample was placed on a goniometer coupled to a Dataphysics OCA 15EC camera. The system was calibrated using ASTM D7490-08 [39]. After that, a drop of deionized water (10 µL) was deposited, and the contact angle was measured using the software. The test was performed in triplicate in different areas of the material. With the angles obtained, it was possible to calculate the work of adhesion (WA) between water and the surface of the papers using the Young–Dupré equation, which is defined in Equation (3) [40].

$$\mathrm{W}\mathrm{A}=\gamma LV(1+cos\theta )$$

(3)

where γLV is the surface tension of the liquid at the interface with air, which was taken from the literature [41], and θ is the contact angle.

2.8. Water Vapor Transmisión Rate (WVTR)

Permeability tests were performed according to ASTM E96M [42]. The test capsule was filled with desiccant (silica gel previously regenerated at 105 °C for 12 h). The paper was weighed, fixed to the dish with the coated side up to prevent water vapor from entering the desiccant, and then placed in the capsule. Subsequently, the capsule was weighed and positioned in a chamber with a relative humidity (RH) of 54%. The weight of the capsule was recorded every hour until 10 data points were obtained. Each data point represents the weight of the system at a given time. The data points were plotted on a weight change vs. time graph, and water vapor transmission rate (WVTR) values were calculated using Equation (4).

$$\mathrm{W}\mathrm{V}\mathrm{T}\mathrm{R}={\displaystyle \frac{G}{tA}}=(G/t)/A$$

(4)

where G is the change in weight in g, t is the time during which the change occurs in h, G/t is the slope of the linear regression in g/h, and A is the effective test area in m². Permeance is calculated as described in Equation (5).

$$\mathrm{P}\mathrm{e}\mathrm{r}\mathrm{m}\mathrm{e}\mathrm{a}\mathrm{n}\mathrm{c}\mathrm{e}(\mathrm{P})={\displaystyle \frac{WVT}{∆p}}={\displaystyle \frac{WVT}{S\left(R_1-R_2\right)}}$$

(5)

where ∆p is the vapor pressure difference (Pa); S is the vapor saturation pressure at the test temperature (Pa); R₁ is the relative humidity inside the test capsule, assumed as 0% according to ASTM E96M for the desiccant method [42]; and R₂ is the relative humidity in the chamber.

3. Results and Discussion

3.1. Pickering Emulsion Stability

Pickering emulsions containing CNF/PLA and CNF/CB in a 1:4 ratio, prepared by high shear homogenization, did not show phase separation for 15 d (see Figure S1 in Supplementary Materials). This result indicates that the emulsions were stable regardless of the oil phase used, as 1 wt% of CNF favors emulsion stability due to the effective coverage of particles at the interface of the dispersed and continuous phases, as have been reported by different authors [43,44].

Gómez-Hoyos et al. developed Pickering emulsions of cocoa butter in water; stabilized with cellulose nanofibers, these emulsions were stable for 15 d using a concentration of nanofibers of 1 wt%, and they found that values lower than this concentration did not show stability [45]. Additionally, authors such as Y. Han et al. mention that increasing from 0.1% to 0.9 wt%, the concentration of cellulose nanofibers (CNF) in Pickering emulsions with a fixed 3 wt% of cellulose nanocrystals (CNC) increases the stability of emulsion [44]. Also, C. Silva et al. report similar behavior, even in cationic cellulose nanofibers, and mention that the concentration of cellulose has a greater effect on stability than the degree of cationization when analyzing concentrations between 0.5 and 1 wt% [46]. Thus, the aforementioned authors agree that cellulose nanofibers form a three-dimensional network that prevents particle aggregation and the destabilization of emulsions at concentrations close to 1 wt%, which aligns with the findings of this study for both CNF/PLA and CNF/CB emulsions.

The emulsions underwent particle size distribution analysis on the day they were made and 15 d later. A 1 wt% CNF suspension was also analyzed. In

Table 1. Particle size distribution results.

Sample	D90 (μm)	D50 (μm)	D10 (μm)	Polydispersion Index (a.u.)
CNF suspension	60.80 ± 2.25 a	20.60 ± 0.60 a	5.53 ± 0.19 a	2.68 ± 0.08 a
CNF/PLA-1D	156.00 ± 7.72 b	64.10 ± 1.99 b	7.04 ± 0.15 b	2.32 ± 0.07 b
CNF/PLA-15D	137.00 ± 3.07 c	41.40 ± 0.74 c	9.63 ± 0.35 c	3.08 ± 0.04 c
CNF/CB-1D	26.30 ± 0.87 d	13.30 ± 0.40 d	4.43 ± 0.07 d	1.64 ± 0.03 d
CNF/CB-15D	58.70 ± 2.87 e	16.40 ± 0.22 e	4.06 ± 0.03 e	3.33 ± 0.12 e

Values with different superscript letters in the same column are significantly different (p < 0.05).

, D₉₀, D₅₀, and D₁₀ are presented as well as the polydispersion index of the emulsions.

During the evaluation period, the emulsions CNF/PLA showed a decrease in D₉₀ and D₅₀, indicating a reduction in the size of the largest and medium-sized droplets, respectively. This behavior suggests that during storage, the steric hindrance between nanofibers and PLA solution changed, which promoted the evaporation of the DCM, favoring the collapse and breakup of the larger droplets into smaller and heterogeneous ones in a manner like that presented in techniques such as solvent emulsion–evaporation [47,48].

In contrast, CNF/CB emulsions exhibited an increase in their polydispersity index after 15 d, associated with the partial coalescence of the droplets in the arrested coalescence, as previously reported by Gomez et al., in CNF/CB emulsions [49], causing a slight increase in the distribution of large and medium droplet diameters. Authors such as C. Whitby et al. have reported that in Pickering emulsions with arrested coalescence, an increase in volumetric diameters can occur due to the merging of droplets. They also note that polydispersity can increase due to droplet rupture, resulting from the presence of both merged and individual droplets of varying sizes [50,51]. Despite the slight increase in particle size distribution, the CNF/CB emulsion remains stable after 15 d of preparation.

Since both emulsions CNF/PLA and CNF/CB remained stable throughout the storage period, they can be applied to paper at any time without compromising quality. This is particularly important for industrial processes, where consistent emulsion stability ensures flexibility in production schedules and reduces the need for frequent preparation. In addition, characterizing the rheological properties of both emulsions is crucial for ensuring uniform application in the rod-coating process and optimal coating performance.

3.2. Rheological Behavior

Rheological analysis is widely recognized in the industry as a key method for assessing the relationship between structure and function in aqueous systems, such as emulsions. Understanding changes in rheological behavior is particularly important when incorporating nanocellulose, as it significantly influences the application methods for paper coatings. Therefore, when considering nanomaterials in coatings, it is essential to carefully evaluate factors like viscosity and viscoelasticity.

Viscosity curves as a function of shear rate for CNF and emulsions at day 1 and after 15 days are shown in Figure 1. Pseudoplastic (shear-thinning) behavior was observed for all samples, which is favorable for paper coating production. Specifically, in Figure 1a, the CNF suspension exhibits a reduction in viscosity with an increasing shear rate. Additionally, after 10 s⁻¹, a slight increase in viscosity is observed due to the reorganization of cellulose nanofibers and the formation of aggregates [52,53,54]. This behavior agrees with previous reports of the rheology of CNFs in the literature [55,56,57]. This occurs because the nanofibers tend to align in the direction of the applied shear, reducing network resistance and facilitating flow, which in turn decreases viscosity [58].

Flow curves were fitted using the Herschel–Bulkley model, as has been previously reported for a Pickering emulsion analysis [59,60]. This model considers that the suspension has a yield stress. If the stress applied to the sample exceeds the yield stress, it will behave as a power-law fluid and can be described by the consistency index k and the flow behavior index n. If the fluid behaves as a Newtonian fluid, it exhibits a flow behavior index n = 1, while n < 1 indicates shear-thinning behavior, and n > 1 indicates shear-thickening.

Table 2. Herschel–Bulkley model parameters for CNF suspension and emulsions at day 1 and day 15.

Sample Name	Yield Stress T0 (Pa)	k-Values (Pa·s)n	Values n	R-Values2
CNF	0.78 ± 0.08 a	0.12 ± 0.01 a	0.90 ± 0.01 a	0.95
CNF/PLA-1D	5.17 ± 0.74 b	0.52 ± 0.07 a	0.64 ± 0.01 b	0.96
CNF/PLA-15D	3.91 ± 0.32 b	0.45 ± 0.01 a	0.68 ± 0.03 b	0.97
CNF/CB-1D	118.67 ± 23.49 b	6.81 ± 0.70 b	0.55 ± 0.05 c	0.45
CNF/CB-15D	24.60 ± 2.02 e	3.35 ± 0.09 c	0.78 ± 0.04 d	0.97

Values with different superscript letters in the same column are significantly different (p < 0.05).

presents the parameters fitted to the Herschel–Bulkley model.

All the samples analyzed presented n < 1, which agrees with the previously mentioned characteristics of pseudoplastic fluids; this is related to the fact that during the shearing process, the network rupture rate of the samples is greater than the recovery rate [61]. This behavior has been commonly reported for O/W emulsions [62]. In such cases, the decrease in k can be related to the reduction in apparent viscosity, as observed in Figure 1.

Finally, it is observed that the yield stress values of the emulsions decrease after 15 days. Some authors report that the value of τ₀ is related to the sample’s ability to resist shear as well as sedimentation or creaming since it represents the minimum shear stress required for the sample to transition from elastic solid behavior to that of a viscous fluid [63].

The viscoelastic behavior is crucial for understanding the rheological behavior of the emulsions during coating, as it can influence the application process [64]. Based on the amplitude sweeps, oscillatory deformations of 0.1%, 0.02%, and 0.158% were selected for the frequency sweep of the CNF suspension, CNF/PLA, and CNF/CB, respectively. These frequency sweeps are presented in Figure 2.

For all the samples presented in Figure 2, the storage modulus (G′) was higher than the loss modulus (G″) across the entire frequency range analyzed, indicating a gel-like structure characterized by an interconnected network in both the CNF suspension and the Pickering emulsions. This observation aligns with previous studies of Pickering emulsions using cocoa butter [45] and other bio-based polymers such as PCL and PVA [65]. Previous studies on the rheological analysis of PLA/CNF emulsions at room temperature by Liu et al. [66] and Li et al. [67] reported similar results as those found in this paper, demonstrating a gel-like behavior in the resultant emulsions.

The viscoelastic properties of emulsions are associated with the formation of an interconnected network between the droplets of the dispersed phase and the particles at the interface, which provides rigidity and structural stability [62]. The higher modulus values of CNF/CB emulsions compared to CNF/PLA emulsions can be attributed to a more densely packed structure, partly due to a higher solid content (38.26 ± 13.05 wt% for CNF/CB, compared to 2.69 ± 0.05 wt% for CNF/PLA). Increases in the modulus of CNF/CB can be linked to several phenomena, including the hardening of the cocoa butter droplets associated with the partial crystallization of triglycerides at room temperature [68]. As mentioned by Marangoni et al., cooling cocoa butter produces a solid material with a crystalline structure. Additionally, the arrested coalescence described in previous studies of CNF/CB emulsions [49] may also play a role. It has been reported that some Pickering emulsions with higher proportions of oil phases tend to exhibit such packing, demonstrating a positive relationship between the moduli and the fractions of the oil phases [69,70,71].

The gel behavior observed in these emulsions provides insights into their stability and applicability. This behavior was maintained for 15 days, indicating a stable emulsion during storage. While the viscoelastic properties significantly impact the coating process [64], they also affect the pressure required by the blade in high-speed blade coating, which is commonly used in paper-related applications.

It is noteworthy that despite its relatively high viscosity and viscoelastic behavior at low shear rates, the systems rapidly reduce their viscosity, even reaching a Newtonian high-shear plateau in some samples (Figure 1c). This behavior is associated with a decreased formation of defects, such as stalagmites, compared to shear-thickening fluids [64]. This pseudoplastic behavior is relevant, as industrial coating application methods typically involve shear rates in the range of 10³ to 10⁶ s⁻¹, which aids in the dispersion of the cellulose nanofibers (CNFs) [57,72]. To replicate the behavior that these suspensions would exhibit during a coating operation, a bar-coating test was performed to apply a double layer over two different types of commercial paper: Kraft and Bond. This was followed by the thermal conditioning of the samples to enhance their homogeneity.

3.3. Coating Processing

Table 3. Characteristics of coated papers. Thickness, grammage, and coating weight.

Paper	Thickness (mm)	Grammage (g/m2)	Coating Weight (g/m2)
Natural Bond	0.090 ± 0.005 a	67.306 ± 2.325 c	NA
Kraft natural	0.148 ± 0.007 b	90.120 ± 1.272 d	NA
Coated Bond	0.099 ± 0.0008 a	87.657 ± 3.599 a	17.472 ± 2.451 a
Coated Kraft	0.155 ± 0.006 b	116.185 ± 4.002 b	25.944 ± 2.807 b

Values with different superscript letters in the same column are significantly different (p < 0.05). NA reffers to “Not applicable”.

presents comparative data on grammage and thickness between Bond and Kraft papers, both before and after coating. For the coated Bond paper, there was an increase in both thickness and grammage compared to natural Bond paper. Additionally, the coated Kraft paper exhibited a significant increase in both parameters compared to natural Kraft, suggesting a higher adsorption of the coating due to the greater porosity and hydrophilic nature of Kraft paper [14,73]. Authors such as N. Sundar et al. report increases in grammage and thickness values in Kraft papers coated with PLA solutions from 5 to 25 w/v% [74], and Inthamat et al. report a significant increase in the thickness of Kraft papers when coated with chitosan-astaxanthin and argue that the applied coatings filled the pores between the fibers of the Kraft paper [75].

In addition, it has been reported that weights of 14 g/m² can be obtained by applying up to 10 coats with a Mayer # 35 bar [76]. The higher the Mayer bar number is, the higher the coating thickness and weight [77] are and consequently the better the coverage of the paper and its subsequent permeability properties is [14]. In this work, higher grammages than those reported in the literature were obtained, not only by using a lower Mayer bar number (#10) but also by making lower numbers of layers, which is directly related to the reduction in the amount of material used to achieve uniform coatings and hence the importance of the formation of an emulsion since the CNF alone does not achieve coating [78]. The uniformity of the coating must be verified by an electron microscopy technique.

3.4. Scanning Electron Microscopy (FE-SEM)

Figure 3a,b show the surfaces of the uncoated papers. Figure 3c,d show the surfaces of Bond and Kraft papers after the application of two layers of coating, respectively. The surface morphology of the natural Bond paper in Figure 3a presents fibers with less heterogeneity and porosity than that of the natural Kraft paper in Figure 3b due to the bleaching treatment in the Bond paper [79].

In addition, as shown in Figure 3c,d, the application of CNF/PLA and CNF/CB coatings results in a coated paper surface that is uniform and significantly less rough compared to the uncoated paper (Figure 3a,b). The coated Bond paper has a smooth surface that is slightly less homogeneous than that of the coated Kraft paper, as some slightly exposed fibers are still visible, indicated by arrows in Figure 3a. In contrast, the porosity observed in the natural Kraft paper is completely reduced on the Kraft paper surface.

3.5. Contact Angle

Contact angle measurements are fundamental for assessing the hydrophobicity introduced by coatings on paper surfaces, as they provide a quantitative evaluation of the interaction between the paper surface and a water droplet.

Table 4. The contact angles (Theta) and work of adhesion (WA) of the evaluated papers.

Paper	Theta @ 5 s (°)	Theta @ 55 s (°)	WA @ 5 s (mN/m)	WA @ 55 s (mN/m)
Natural Bond	53.40 ± 8.37 a	26.87 ± 6.33 b	114.47 ± 10.64 a	135.82 ± 4.72 a
Natural Kraft	97.27 ± 1.30 a	50.77 ± 4.15 b	62.89 ± 1.99 a	117.40 ± 4.83 a
Coated Bond	86.93 ± 0.82 a	82.67 ± 4.03 a	75.84 ± 1.26 a	81.15 ± 6.12 a
Coated Kraft	88.47 ± 1.94 a	87.30 ±1.69 a	73.92 ± 2.99 a	75.38 ± 2.59 a

Values with different superscript letters in the same row are significantly different (p < 0.05).

presents the results of the contact angles and adhesion work for the papers.

In Table 4, it is observed that both natural Bond and Kraft paper are hydrophilic, with a contact angle of 26.87 ± 6.33 and 50.77 ± 4.15 at 55 s, respectively. In comparing adhesion work, it is evident that natural Bond paper, without coating, exhibits significantly higher adhesion work from the outset, which increases over time. This behavior reflects high wettability, indicating low intrinsic resistance to water. This is further supported by the contact angles presented in Figure 4a, where the lowest contact angle of 26.87 ± 6.33° is observed at 55 s of analysis [14,73]

After coating, both Bond and Kraft papers increase their contact angles, approaching those of hydrophobic materials (contact angle ≥ 90°), with measured contact angles of 82.67 ± 4.03° and 87.30 ± 1.69°, respectively. As observed in Figure 4c,d. These changes in behavior are associated with the presence of cocoa butter in the CNF/CB coating.

The most significant change introduced by the coatings on paper surfaces is the reduction in adhesion work. For Bond paper, adhesion work decreased from 135.82 ± 4.72 to 81.15 ± 6.12, and for Kraft paper, it decreased from 135.82 ± 4.72 to 75.38 ± 2.59 at 55 s. No statistically significant differences were identified in adhesion work between 5 and 55 s. This result suggests that the coating is not only effective in repelling water but also maintains its effectiveness relatively constantly over the evaluated time. The effectiveness of the coating in modifying the surface interaction with water in both Bond and Kraft papers can be attributed to the presence of cocoa butter (CB) and PLA, which reduce interaction with water molecules on the paper surface [14].

Finally, when comparing both coated papers, the results indicate that once the coating is applied, the values for the work of adhesion and contact angle do not show statistically significant differences over the evaluated time. Therefore, the following section presents an analysis of the permeability properties to examine the interaction of water with the paper over time, considering not only surface interactions but also how water molecules permeate through the paper.

3.6. Water Vapor Permeability

Another important parameter to consider when analyzing coatings on paper is the extent to which the permeability values change after the coating is applied. Thus, Figure 5 presents the water vapor transmission rate (WVTR) and water vapor permeance (P) obtained for both natural and coated papers.

The water vapor permeability tests presented in Figure 5 showed that coated papers exhibit greater resistance to water vapor transmission compared to their natural state, with statistically significant differences for both WVTR and P. Initially, Kraft paper had a WVTR value of 29.20 ± 1.135 g/m²·h, which significantly decreased to 7.06 ± 2.80 g/m²·h after coating, representing a reduction of 75.82%. Similarly, natural Bond paper showed a WVTR value of 30.56 ± 0.34 g/m²·h, which decreased to 14.37 ± 5.91 g/m²·h after coating, indicating a reduction of 47.02%.

The decrease in WVTR values can be attributed to the smooth and uniform surface of the CNF/PLA and CNF/CB coatings observed in the FE-SEM images (Figure 3), which show fewer pores on the paper surfaces. Syverud et al. explain that the reduction in porosity results from the formation of a denser material structure, creating a more tortuous path that slows down the passage of fluids through the paper [80]. The P values also demonstrate an improvement in resistance to water vapor transmission, changing from 2.12 × 10⁻² ± 8.22 × 10⁻⁴ and 2.22 × 10⁻² ± 2.47 × 10⁻⁴ for natural Kraft and Bond papers to 5.12 × 10⁻³ ± 2.04 × 10⁻³ and 1.04 × 10⁻² ± 4.28 × 10⁻³ for coated Kraft and Bond papers, respectively. This improvement is attributed to the increased hydrophobicity of the paper after coating. As observed in the contact angle measurements, these coatings create a slightly hydrophobic surface.

Z. Song et al. coated 60 g/m² papers with PLA composites containing CNF modified by grafting with cerium nitrate and ammonium. The coating was applied using solvent casting. The authors reported WVTR values of approximately 8.33 g/m²·h at 23 °C and 50 RH% [14]. As mentioned in this study, we obtained better WVTR values, using less material and without the need for surface modifications to the CNF. The WVTR values found for coated Kraft and Bond papers are lower than those of commercial papers used for packaging yerba mate, which were tested at a temperature of 40 °C and 55 RH%. Ramallo and Albani reported WVTR values for these papers of 450 and 500 g/m²·day with thicknesses of 0.139 and 0.125 mm, respectively [81]. However, to ensure the applicability of these coated papers in food packaging, it is important to test them for other properties, such as air permeability, grease resistance, and coating behavior in different pH environments.

4. Conclusions

The results from this study offer significant advancements in sustainable material technology and paper performance. By utilizing cellulose nanofibers (CNF) from agricultural by-products and combining them with biodegradable PLA and cocoa butter, a waterborne emulsion of PLA and cocoa butter was developed to enhance moisture barrier properties. This was demonstrated by a substantial reduction in water vapor transmission rates and the improved hydrophobicity of the coated papers, highlighting the practical benefits for packaging and specialty papers.

The rheological analysis of the Pickering emulsions CNF/PLA and CNF/CB showed shear-thinning behavior, with the presence of yield stress and gel behavior. These rheological characteristics support the use of the emulsions as suspensions for application as paper coatings via the bar-coating method. When applied to Bond and Kraft papers, the emulsions decreased their hydrophilic character and improved the WVTR values by 75.82% and 47.02%, respectively. This improvement is attributed to the formation of a smooth surface with reduced porosity, as observed in FE-SEM images, resulting in slightly hydrophobic surfaces with contact angles of 82.67 ± 4.03° for Bond paper and 87.30 ± 1.69° for Kraft paper. The WVTR values found for coated Kraft and Bond papers were lower than those of commercial papers used for packaging yerba mate. However, to ensure the suitability of these coated papers for food packaging, it is essential to evaluate additional properties, such as air permeability, grease resistance, and coating behavior in various pH environments.