Designing flaxseed oil-in-water emulsion-based stimuli-responsive edible films

2.1 Materials

Gelatin (bovine skin, Type B, 250 Bloom) and soy lecithin were purchased from Qualifirst Foods Ltd. (Etobicoke, ON, Canada) and BulkBarn Food Ltd. (Ottawa, ON, Canada), respectively. The chemical structures of gelatin and lecithin are given in Figure 1. Flaxseed oil was purchased from Millipore Sigma (Oakville, ON, Canada). All pH adjustments were done with 0.1 mM HCl.

Figure 1:

Representative chemical structures of gelatin (top) and lecithin (bottom).

2.2 Preparation of emulsified film-forming solutions and films

The soy lecithin used in this study was first tested for solubility in water and oil phases. The lecithin was soluble in water but not in the oil phase, and therefore, it was suitable for O/W emulsions. Then, trials of O/W type emulsions were prepared using different concentrations of lecithin (1 %, 2 %, and 3 % w/v) + gelatin (1 %, 2.5 %, and 5 % w/v) to evaluate the emulsion stability. In addition, samples containing only lecithin or gelatin at these concentrations did not form stable emulsion or film, and therefore gelatin and lecithin combinations were used in further experiments. Based on the results, the emulsion containing 2 % w/v lecithin + 5 % w/v gelatin resulted in no phase separation over time (2 weeks) and was chosen for further analysis. The stability of the emulsions during the drying (4 days), is the most critical since gelation occurred within 20 min and the dried samples were stable in terms of keeping oil in their structure during storage (3 months).

The aqueous phase of the emulsion (2 % w/v lecithin and 5 % w/v gelatin) was prepared according to the method of Pulatsu et al. [14] with some adjustments. Lecithin was dissolved in Milli-Q water at 300 rpm for 15 min. Gelatin was then slowly added to the lecithin-containing solution at room temperature, and stirring was continued for another 15 min, with the pH adjusted to 4 in the last 5 min for gelatin stability (pH < pI ensures the presence of positive charges). The solution was then transferred to a water bath set at 50 °C and held for 30 min. The mixture was allowed to stand at room temperature for 10 min before adding the oil phase.

Numerous studies have reported incorporating oil at proportions ranging from 0.1 to 7 % (v/v) in emulsions [37], [38], [39], [40], [41], [42], [43]. In this study, the emulsions were prepared with two different O:W ratios (5:95 and 10:90). The two ratios were selected because a larger volume of oil incorporation resulted in thicker films and reduced their swelling performance, which is undesirable for further applications. Flaxseed oil (oil phase) was added dropwise to the aqueous solution (water + lecithin + gelatin), and agitated for 15 min at room temperature, using a magnetic stirrer at 1,000 rpm. The emulsions were then allowed to set for 10 min to remove large bubbles. Thereafter, different volumes (5 mL and 10 mL) of the O/W emulsions (5:95 and 10:90 ratios) were poured into small Petri dishes and left to settle at room temperature for 30 min. Finally, the samples were placed in a fume hood and air-dried for 4 days, where air flow was 80 ft/m and the temperature was 20 ± 2 °C. The samples, explanations, and schematic representation of the experimental setup are given in Table 1 and Figure 2, respectively.

Figure 2:

Flowchart of the film preparation.

2.3 Rheological measurements

The rheological properties of the emulsions were measured in duplicates with a rheometer (Discovery HR-2, TA Instruments, New Castle, DE, USA) equipped with a parallel sandblasted plate geometry (20 mm diameter, Peltier plate steel – 113637) due to the presence of oil (i.e., to reduce slippage), following 20 min of preparation which allowed the sample to form a gel. All measurements were performed at 25 ± 0.1 °C with a 60 s soak time and a gap of 100 μm. Flow sweep curves were recorded under an applied shear rate of 1–100 s⁻¹. The data were fitted to the Herschel Bulkley model (Eq. (1)) as described in a previous study [44].

where σ _y is yield stress, K is the consistency index (Pa s ⁿ ), and n represents the flow behavior index [45].

An amplitude sweep test was first performed at 1 Hz (0.01–100 % strain) to determine the linear viscoelastic region (LVER). The limiting strain within the LVER was then identified to establish the optimal strain amplitude for the frequency sweep test. A frequency sweep test was then conducted over an angular frequency of 0.1–100 rad/s within the identified LVER for each sample (at 1 % shear strain). A power-law model fitting was applied on the complex modulus (G ^*) as a function of angular frequency (ω) obtained from frequency sweep tests to evaluate the internal network in terms of strength and connectivity.

where A _f and z are network strength and connectivity related to the number of junctions/protein area, number of endpoints/protein area, length of a continuous protein particle and the amount and size of network gaps [46].

2.4 Film characterization

The fabricated films (with a diameter of 45 ± 1 mm) were characterized in terms of thickness, moisture content (MC), and bulk density (at least two measurements were taken from the film samples). A digital micrometer was used to measure the thickness of the dried films in at least three places, and results were reported in mm. A moisture analyzer with 0.01 % MC (1 mg) readability (HE53, METTLER TOLEDO, Mississauga, ON, Canada) was used to determine the MC of the dried films at 105 ± 1 °C. The bulk film density was approximated based on the weight (g) and volume (cm³) ratio of the film samples [19].

2.5 Swelling behavior of emulsion-based films in DI water (pH: 4, 7, and 10)

The initial weight (W _i ) and thickness of the dried films were measured based on an adjusted method from a previous study [14]. The films were then immersed in deionized water (25 ± 0.1 °C) adjusted to pH 4, 7, and 10. After immersion, the wet films were removed and gently wiped, and their weights (W _t ) were recorded after 10 min, 20 min, 60 min, and overnight. Thickness measurements were noted after drying only and were performed in duplicates for each film. The amount of absorbed water (swelling %) was calculated as expressed in Eq. (5). The swelling rate (%/min) was determined from the slope of the swelling % versus time (min) graph.

where W _i is the initial weight of the dry film (g), and W _t is the weight of the swollen film at different exposure times (10, 20, or 60 min) in DI water at different pH values.

Swelling kinetics was further analyzed by model fitting via Eq. (6), known as Fick diffusion model.

where W _t and W _e stand for swollen weight of the films at t = 10, 20, and 60 min and overnight. The k is the Fick’s constant and n defines the swelling dynamics. The values n > 0.5 and n < 0.5 for non-Fickian diffusion, pseudo-Fickian diffusion, respectively. The case n = 1 is characterized as zeroth-order release and mode of transport is pseudo-case-II solute transport [50].

2.6 Statistical analysis

The data were reported as mean ± standard deviation, and each measurement was duplicated. One-way analysis of variance (ANOVA) and Tukey’s tests using GraphPad Prism (GraphPad Software, La Jolla, CA) were performed (p < 0.05). The Pearson correlation was employed for some parameters, as applicable.

3.1 Rheological behavior of film-forming emulsions

Understanding the rheological properties of film-forming solutions is critical to their various applications, including in food packaging [51], and emerging 4D printing [52]. The incorporation of oil in emulsion-based films influence their rheological properties, the film formation, and swelling behavior when exposed to stimuli medium (e.g., water). In this study, two different oil phase ratios (O:W, 5:95 or 10:90) were prepared, as described earlier (Figure 2). Figure 3 shows the rheological behaviors of the emulsions at 5:95 (5O95W) and 10:90 phase ratios (10O90W).

Figure 3:

Rheological behavior of the film-forming emulsions. Flow sweep results: (A) stress versus shear rate and (B) viscosity versus shear rate. Dashed lines are for Herschel–Bulkley model curve-fitting. Oscillation tests: (C) modulus versus shear strain, (D) modulus versus angular frequency and (E) modulus versus time.

Table 2 provides the Herschel-Bulkley model fitting parameters (R ² > 0.90). The samples exhibited yield stress values, σ _y , where 10O90W had a higher value than 5O95W (p < 0.05). The n represents the shear sensitivity or deformability of the samples, and K indicates the overall material’s ability to resist deformation [45]. The emulsions exhibited similar K (p > 0.05), but different n values (p < 0.05). The n values of both 5O95W and 10O90W were below 1 (Table 2), indicating non-Newtonian, shear thinning behavior. This suggests that shearing causes the structural breakdown due to the disruption of molecular interactions and rearrangement in the direction of applied shear. The results align with similar studies, where biopolymer solutions and emulsions exhibit shear thinning behavior due to the nature and concentration of the polymers used [14], [19]. According to Zhao et al. [53], increased oil phase in emulsions improved the rheological properties associated with the denser packing of oil droplets. Overall, upon exceeding the corresponding σ _y , the flow behavior shear thinning is more pronounced in the less oil-containing sample (5O95W), but the resistance to deformation remained the same in both samples (Table 2).

In a gelled emulsion, the structural integrity is determined by the type of gel network formed (aggregation of the dispersed oil droplets or polymer matrix gel), the role of gelatin, and the mechanisms of interactions between the oil droplets, gelatin matrix and lecithin [54], [55]. Taking into account that the LVER of both 5O95W and 10O90W is under 10 % shear strain (Figure 3C), the samples were further tested for their frequency dependence, and their results were expressed in terms of the storage or elastic modulus (G′) and loss or viscous modulus (G″). Both samples showed a solid-like behavior (G′ > G″), where some of the deformation is elastic or recoverable depending on the applied stress (Figure 3C and D). In addition, when the oil concentration was at 10 %, both G′ and G″ had higher values, proving the presence of a stronger gel network. A previous study by Van Vliet [56] reported that the change in modulus of filled gels depends on whether bonds are formed between the emulsifying agent (i.e., lecithin) and the gel matrix (gelatin) that can directly influence the binding of oil droplets to the matrix. In addition, the presence of aggregated oil droplets can increase the volume fraction, contributing to the modulus [57]. According to Dickinson [54], the mechanical properties of gelatin can be enhanced by the adsorption of oil droplets in the aqueous phase. Oil droplets act as active fillers and strengthen the droplet-matrix interactions, reinforcing the gel structure. G′ values were not dependent on the frequency, characterized by a linear line over the studied range (Figure 3D). In contrast, G″ had frequency dependent behavior, pronounced with increased frequency (Figure 3D). A typical behavior of a viscoelastic solid is a crossover at higher frequency, which is estimated beyond 100 rad/s for these samples.

The Power law model fitting parameters of the G ^* values, known as the sample stiffness, within the studied oscillatory frequency range were used to describe the matrix effect of the film-forming emulsions (R ² ≥ 0.90). More oil incorporation resulted in larger A _f and smaller z values (p < 0.05) (Table 2). Network connectivity, z, was higher for less oil-containing samples, which can be explained by more protein and emulsifier due to a larger fraction of the water phase. Hence, oil strengthens the network while the water phase contributes to the extent of interaction, which aligns well with the G′ and G″ discussion in the previous paragraph.

Oscillatory time sweep tests were run to observe time-dependent change in the samples, as given in Figure 3E. The modulus of 5O95W increased and reached a plateau around ∼20 min, indicating the gel setting. The absence of significant changes in the G′ and G″ trajectory values, thereafter, with the elastic modulus remaining predominantly higher suggests that the sample demonstrates good O/W structural elastic stability. Similarly, Bertasa et al. [58] reported stable G′ and G″ over 60 min in agar hydrogels, demonstrating long-term stability for specific application. The 10O90W sample exhibited a more stable modulus values and an earlier onset of the modulus plateau (∼10 min) compared to 5O90W, likely due to larger oil fraction (Figure 3E).

3.2 Film formation and physical properties

Regardless of the different O:W phase ratios and poured volumes of film-forming emulsions, the average density values of the dried film samples were 0.12 ± 0.01 g/cm³. These values are in agreement with findings in similar studies [19], [59]. This approximate density value ensures the uniformity of the resultant film samples.

The physical properties and the survivability of the emulsion-based films in DI medium at different pH values are shown in Table 3. The thickness of the emulsion-based films was dependent on the amount of poured emulsion, of which 10O90W-10 was the thickest (p < 0.05). Similarly, 10O90W-10 had the highest weight among the samples; it has higher volume of oil phase (i.e., O:W ratio) compared to 5O95W samples. This was unexpected since the oil density is less than that of water, in addition to the recorded lower MC for 10O90W samples (p < 0.05). The observed phenomenon can be explained by the presence of more solids (lecithin and gelatin) in larger poured volumes (Tables 1 and 3). These findings are in accordance with the study by Tongnuanchan et al. [60] the increase in thickness with O:W ratio is attributed to the interfering effect of oil on the compact and ordered network of the gelatin film matrix. As water evaporates during the drying, the thickness and weight of the film are expected to decrease considerably. Our recent study [14] demonstrated that, under similar preparation methods, environmental conditions, and the film-forming solution used in Petri dishes, edible films prepared with 5 % w/v gelatin (without the oil phase), were approximately half the thickness of the emulsion-based films in this study. This finding is expected, given that the added oil alters the interactions and increases the molecular distances within the film matrices due to differences in molecular structures, density, and affinity to water.

3.3 Optical properties of the films

The appearance of food and edible packaging is an important attribute influencing consumer acceptability, and a quality control parameter used in the food industry. The refractive index of the phases, biopolymer and emulsifier concentration, and droplet size distributions are critical to their optical properties. To better understand the optical characteristics of the films, the opacity and different CIE color parameters, including lightness (L*), redness-greenness (a*) and yellowness-blueness (b*) were measured. The transparency and color of films are influenced by multiple factors, such as the types of components present, the droplet size of the dispersed phase, the heterogeneity of their distribution, and the internal microstructure development during the drying [61]. A higher value corresponds to a more opaque film. As shown in Table 4, 10O90W-5 had the highest opacity, similar to 5O95W-5, and better light absorbance. The smaller volume of film solution resulted in a thinner dried film (Table 3), which in turn increases light scattering.

All samples displayed lightness (L*) values close to 100 (p > 0.05), indicating a lighter coloration. Meanwhile, in terms of primary color analysis, negative a* values and positive b* values were observed in all samples. a* values were less negative, and b* values were more positive, with a larger volume of poured film solution i.e., 5O95W-10 and 10O90W-10. Specifically, 10 mL dried films were significantly less green but more yellowish compared to 5 mL dried films. These color changes are related to (1) the yellow pigments present in soy lecithin and flaxseed oil [62], [63] and (2) the increased solid content, which enhanced the film color saturation. The films showed distinct yellow color variations. It is known that for ΔE* < 3, no perceptible differences can be found while at ΔE* > 12 different colors are perceived among the samples [49]. These observations are consistent with the ΔE* values (ΔE* > 12) [49], suggesting perceptible color differences in samples. The overall color difference was similar for 5O95W-5 and 10O90W-5, even though the thickest sample was 10O90W-10. A balance may have occurred for larger volume samples because lecithin and oil may have similar contributions to yellowness pigments.

3.4 Cross section microscopic structure of the films

Figure 4 shows the SEM micrographs of the cross sections of the emulsion-based films. Both 5O95W and 10O90W had a slightly uneven and rough structure at magnification 100× due to cracks formed during cutting.

Figure 4:

SEM photomicrographs of emulsion-based films’ cross sections at 100× magnification. Scale bars under μm labels (with red boundary) in the images represent 100 μm.

The O:W ratio impacted the film microstructure, especially the continuity within the polymer matrix (Figure 4). More droplet-like structures appeared in the polymer matrix, which can be seen in 10O90W samples. A more layered cross section was observed in 5O95W films of different thicknesses due to the water-dominated phase (Figure 4). These images are visual evidences of the rheological findings regarding the network attributes. In other words, 10O90W had smaller z but larger A _f than 5O95W, leading to continuity in the microstructure while 5O95W had a layered appearance, indicating discontinuity.

Changes in a film microstructure during the drying are due to the structural rearrangement of the phases, driven by water evaporation. This results in destabilization phenomena, such as coalescence and creaming, leading to oil migration to the surface of the film [38], [64]. Such phenomena were not observed in the film samples for at least 3 months, but after that high oil containing films started to oil off at the ambient temperature kept in the Petri dishes. Emulsion gels were stable over a week owing to the adsorbed lecithin and gelatin molecules at the O/W interface. Destabilization due to coalescence or creaming was not expected (and confirmed via storage at ambient condition for up to 3 months) as the solidification was about 20 min, which limits the movement of oil droplets in the gelled matrix. The time-resolved solidification from preparation to gelation of emulsion is illustrated in Figure 5, confirming the time sweep results in Figure 3E. This analysis shows how quickly they solidify, which is relevant to drying phenomenon and casting. For example, casting would be difficult if the solution or emulsion solidifies too fast.

Figure 5:

Time-resolved solidification images of the film-forming emulsion (5O95W).

3.5 Swelling behavior of the emulsion-based films

The film swelling is similar to hydrogel swelling. It is of significant interest as it is critical to its drug delivery, packaging, and 4D printing applications. This transport phenomenon involves three successive steps: (i) the diffusion of water molecules into the polymer network, (ii) the relaxation of hydrated polymer chains, and (iii) the expansion of the polymer network into the surrounding aqueous solution [65].

The overnight survivability of the samples in DI water at varying pH values revealed that samples survived pH 4 and 7 but not pH 10 medium. Only 5O95W-10 survived at pH 10 for up to 60 min. This finding can be due to longer gel setting time and the presence of more protein and emulsifier allowing better network connectivity. The film swelling rates were determined by linear regression analysis within 60 min of immersion based on Figure 6, which shows the swelling behavior of film samples exposed to DI medium at three different pHs. At alkaline pH, films were solubilized, and the components were disrupted, as expected. Also, proteins are denatured, and lipids become oxidized, leading to disintegration of the film components. On the other hand, the films were observed to swell overnight at pH 4 and pH 7 (Table 3).

Figure 6:

Swelling behavior of emulsion-based films in DI medium at (A) pH 4, (B) pH 7, and (C) pH 10.

When films are immersed in a solution, a small amount of film matter is expected to dissolve in water due to the interaction between the swelling medium and film [47]. Recent studies have demonstrated that gelatin-based films exhibit good water uptake due to their hydrophilic nature [66], [67]. In contrast, the presence of hydrophobic flaxseed oil is expected to reduce the rate of water transport [68]. However, a review article by Atarés and Chiralt [64] revealed that incorporating the oil phase does not necessarily reduce the swelling rate depending on the oil type and distribution, which alter the molecular mobility. Indeed, the microstructure of a film is dictated by molecular interactions and drying conditions. Table 5 presents the swelling rates of the films, where no significant differences were reported (p > 0.05). This could be explained by their low swelling rates compared to a previous study that did not incorporate an oil phase [14]. In addition, no significant differences (p < 0.05) were observed for the overnight change in thickness of the submerged films (Table 5). However, the % overnight change in weight at pH 4 for 5O95W-5 was the highest, similar to pH4 for 5O95W-10. This finding could be explained by the fact that both films and the swelling medium have the same pH, which provided an equilibrium between the sample and the medium. Additionally, the overnight appearance of the emulsion-based films revealed that the swollen films appear to be less saturated (i.e., less yellowish), likely due to the absorption of water and/or the loss of pigments (Figure 7).

Figure 7:

The images of the films before swelling ( t = 0), after overnight immersion in DI at pH 4 and 7. The films did not survive at pH 10, hence, no image was shared.

Table 6 lists the swelling kinetics parameters, associated with the water transport into the films, of the emulsion-based films, which were able to survive 60-min exposure to varied pH conditions. Only swelling media at pH 4 and 7 were suitable for such analysis (Figure 6). All surviving samples exhibited pseudo-Fickian diffusion given n < 0.5 and with statistically similar k values (Table 6). Similar findings were reported for polymer swelling in the presence or absence of polyelectrolytes and at different pH [69], [70]. Diffusion exponents and coefficients are useful for swelling characterization governed by water diffusion and chain relaxation. Psuedo-Fickian case is where the water penetration rate is much below the polymer chain relaxation rate [70]. This is expected due to the presence of oil in the films, which restricted the water transport.