Rheological and Antimicrobial Properties of Chitosan and Quinoa Protein Filmogenic Suspensions with Thyme and Rosemary Essential Oils

Food packaging faces the negative impact of synthetic materials on the environment, and edible coatings offer one alternative from filmogenic suspensions (FS). In this work, an active edible FS based on chitosan (C) and quinoa protein (QP) cross-linked with transglutaminase was produced. Thyme (T) and rosemary (R) essential oils (EOs) were incorporated as antimicrobial agents. Particle size, Z potential, and rheological parameters were evaluated. The antimicrobial activity against Micrococcus luteus (NCIB 8166) and Salmonella sp. (Lignieres 1900) was monitored using atomic force microscopy and image analysis. Results indicate that EOs incorporation into C:QP suspensions did not affect the Z potential, ranging from −46.69 ± 3.19 mV to −46.21 ± 3.83 mV. However, the polydispersity index increased from 0.51 ± 0.07 to 0.80 ± 0.04 in suspensions with EO. The minimum inhibitory concentration of active suspensions against Salmonella sp. was 0.5% (v/v) for thyme and 1% (v/v) for rosemary. Entropy and fractal dimension of the images were used to confirm the antimicrobial effect of EOs, which modified the surface roughness.


Introduction
Numerous factors affect the original quality of food products [1], and up to date, many synthetic food packaging materials are used due to their good mechanical and barrier properties, but they show long biodegradation processes [2]. Besides providing food protection from the environment, active packaging may also protect from foodborne illness outbreaks [3,4].
Biopolymers are commonly used to produce coatings, such as polysaccharides (starch, chitosan, cellulose), proteins (animal or vegetable), and also lipids (waxes, fatty acids), or a mixture of them [4]. These materials act as barriers against the transport of gases and water vapor, leading to longer shelf life, resolution visualization of their mechanisms of action at cell surface level by atomic force microscopy, which to our knowledge has not been previously reported.
The present work's objective was to develop and characterize a filmogenic suspension based on chitosan and quinoa protein cross-linked with transglutaminase, with antimicrobial effect by incorporation of rosemary and thyme essential oils.

Supplies
Chitosan

Quinoa Protein Extraction
The quinoa seeds were ground to about 200 µm using a coffee grinder (Krups Model GX410011, Solingen, Germany). The flour was defatted by three extractions with ethanol (70% v/v) in the ratio 1:10 w/v (flour:solvent) under constant stirring for 2 h and 25 • C [27,28]. The defatted flour was suspended in distilled water (10%, w/v) adjusting to pH 11 with 1 N NaOH, and stirred for 1 h at room temperature (25 • C). Then, the samples were centrifuged at 3200× g, at 10 • C for 30 min. The supernatant was adjusted to pH 4.5 and stirred for 30 min, followed by centrifugation as before. The precipitate was re-suspended in distilled water at a 5:95 ratio (precipitate: water, w/v) and neutralized with 2 N NaOH, followed by oven drying at 50 • C (ED, Binder, Tuttlingen, Germany). The protein isolates were ground for 2 min using the coffee grinder and passed through a No. 9 mesh (Tyler standard) of 200 µm pore opening [12].

Crosslinking with Transglutaminase
The C:QP FS (Section 2.2.2) before homogenization was adjusted to pH 9 and added with 1.4% (v/v) of TG solution (10% w/v), stirred for 1 h, followed by pH adjustment to 11. After cross-linking, the FS was homogenized by a high-speed mixer (IKA T25-Ultra-Turrax) at 21,500 rpm for 3 min, followed by sonication for 10 min at 150 W and 20 kHz (SONOPULS, HD3200).

Minimum Inhibitory Concentration
The minimum inhibitory concentration (MIC) of the EOs was evaluated against Salmonella sp., and Micrococcus luteus, adjusted to an optical density of 0.08 (10 6 CFU/mL) at 600 nm. Salmonella sp. was chosen due to its highly frequent presence in foodborne outbreaks, being lethal in many cases. Micrococcus luteus is not a severe pathogenic bacterium, but for many years it has been used as a model system for bacterial cell wall study due to low peptidoglycan cross-linking (about 25%) [30,31]. M. luteus sensitivity to cell wall disruption was the reason to choose this microorganism as Gram-positive model. EOs at six different concentrations (0%, 0.5%, 0.7%, 1.0%, 1.5%, and 2.0%, v/v), and Tween 80 at constant concentration of 0.5% (v/v), were added to the TSB culture media. Absorbance readings were recorded every h, and the MIC was considered as the lowest concentration tested that inhibited microbial growth.

Active Filmogenic Suspension
The 1:10 C:QP FS (Section 2.2.2) was cross-linked with TG (Section 2.2.3), and then the active FS was produced by adding the EOs at the previously found MIC and Tween 80 at 0.5% (v/v). The active FS was homogenized by a high-speed mixer (IKA T25-Ultra-Turrax) at 21,500 rpm for 3 min, followed by sonication for 10 min at 150 W, and 20 kHz (SONOPULS, HD3200) [32] 2.2.6. Kinetic Parameters of Tested Microorganisms The effect of the active FS on Salmonella sp. and M. luteus was evaluated by measuring the kinetic parameters, which was associated with the antimicrobial effect. The tested microorganisms were grown in the QP and C biopolymers, C:QP FS, and the FS added with EOs (C:QP:T and C:QP:R). The parameters determined were doubling time (T d ) and specific growth rate (µ) following the Monod model [33], using the GraphPad Prism 5.0 software (San Diego, CA, USA), showing the active FS effect on microbial growth.

Antimicrobial Activity of Active Filmogenic Suspension
Salmonella sp. was activated in TSB broth for 24 h at 37 • C, whereas the same media was used to activate M. luteus for 48 h at 30 • C. Then, the FS of C:QP; and FS with EOs (C:QP-R; C:QP-T) were inoculated with each microbial culture to reach 10 6 CFU/mL. The antimicrobial activity was determined by evaluating the microbial population after 2, 4, 8, 12, 24, and 48 h, using the pour plate method in TSB agar, and incubating as above mentioned.

Particle Size
The mean particle diameter of FS was determined with a Zetasizer Nano-ZS laser diffractometer (Malvern Instruments, Worcestershire, UK) at 633 nm and 25 • C, equipped with a backscatter detector [34]. Z Potential Z potential (mV) was determined by phase analysis light scattering (PALS) with a Zetasizer Nano-ZS laser diffractometer (Malvern Instruments, Worcestershire, UK), which determines electrical charge at interface of droplets dispersed in aqueous phase [34].

Rheological Properties of Emulsion
A rheometer equipped with concentric cylinder geometry (Discovery Hybrid Rheometer TA Instruments, New Castle, DE, USA), was used to determine the rheological properties of the FS at 25 • C. Temperature equilibration and particles settling were allowed for 5 min before steady-state flow measurements were carried out, using a shear rate range of 0 to 100 s −1 . Shear stress, shear rate, and apparent viscosity were evaluated using the TRIOS 4 software (TA instruments); the experimental flow curves were fitted to the Casson model (Equation (1)): where σ is shear stress (Pa), σ 0 is the elastic limit (Pa), γ is the shear rate (s −1 ) and η is apparent viscosity of the fluid [35].

Sample Preparation
Samples were prepared according to Mathelié-Guinlet [36] with some modifications. The bacteria were activated for 5 h, at 37 • C, to reach the log phase. Then, 5 µL (10 6 UFC/mL) of each tested microorganism were inoculated into 2 mL of the FS (control) and 2 mL of FS containing T and R EOs at the previously determined MIC. All mixtures were incubated for 5 h, at 37 • C. An aliquot of 100 µL of each treatment was deposited on glass slides of 26 × 76 mm and 1.1 ± 0.1 mm thick and allowed to dry for 2 h at room temperature in a laminar flow cabinet. Cell adhesion was enhanced by previously adding a layer of 100 µL of FS without EOs to the glass slides and allowed to dry at room temperature.

Cells Topography
This determination was carried out using an atomic force microscope (AFM; Multimode V, Veeco, Plainview, NY, USA) in contact mode to avoid any damage to the samples. The images were obtained at a scanning speed from 0.5 to 1.0 Hz, with a resolution of 512 × 512 pixels and at different areas (50 µm × 50 µm; 2.5 µm × 2.5 µm; 0.5 µm × 0.5 µm) [36].

Texture Image Analysis
Texture image analysis was conducted following Arzate-Vazquez [37] and was applied to quantitatively characterize the microbial surface microstructure treated with the FS C:QP with and without EOs. Three characteristics were selected: entropy, fractal dimension, and roughness. All images obtained by atomic force microscopy (AFM) were converted to grayscale images. Subsequently, the gray level co-occurrence matrix (GLCM) and differential box count (SDBC) algorithms were applied to obtain the texture characteristics from grayscale images, both included in the Image J 1.52 software (NIH, Bethesda, MD, USA). GLCM is a second-order statistical algorithm that compares two neighboring pixels at once and compiles the frequency with which different gray levels can be found within a restricted area. In this algorithm, three variables are considered: the number of gray levels (0-255), the distance of the pixels (d), and the offset angle (θ). The image texture (entropy) was analyzed by studying the spatial dependence of pixel values represented by a co-occurrence matrix Gd, θ, with the input Gd, θ (i, j), which represents the frequency whereby a pixel with intensity i is adjacent to a pixel with intensity j, separated in direction θ. These parameters were measured at a distance d equal to 1 and an angle equal to 0 • , using Equation (2).
The fractal texture was evaluated using the power law scale to obtain its fractal dimension (FD) using the SDBC algorithm based on the surface intensity graph. It is generated from 2D grayscale images by plotting pixel coordinates (x, y) versus their gray level in the z-axis. FD was estimated from the slope of the plot log (number of boxes) vs. log (box size) (Equation (3)), where "N" is the number of boxes and "r" is the length of the box size The FD is an object property that shows how much of the space that contains it is occupied and can acquire continuous values in the space of real numbers between 0 and 3. FD tries to measure the Foods 2020, 9, 1616 6 of 17 extent that a 2D object fills the 3D space, or a one-dimension object resembles a 2D surface allowing the description of the geometry of many natural structures that appear to have great complexity but having the same geometric regularity. According to fractal geometry, the line has FD greater than zero, but less than 1, the FD of a dot is 0, the plane FD is between 1 and 2, whereas that of the cube is between 2 and 3.
FD is directly related to the extent of surface roughness (Rq) (Equation (4)), which was obtained following the protocol of Escamilla-García [12]. Rq is the standard deviation of Zi values indicating roughness (nm); Zi is the difference in the height of i relative to the average height, and N is the number of points in the image.

Statistical Analysis
All tests were performed in triplicate, and data were evaluated by one-way analysis of variance using GraphPad Prism 5.0 software (GraphPad Software, San Diego, CA, USA). Significant differences were determined by Dunnett's test, with a significance level of p < 0.01. Data are presented as the mean ± standard deviation.

Antimicrobial Activity
The antimicrobial activity of R and T free EOs against M. luteus (Gram-positive) and Salmonella sp. (Gram-negative) showed a MIC for both microorganisms of 0.5% (v/v) and 1.0% (v/v) of T and R EOs, respectively ( Figure 1). The EOs were added to the FS of C:QP using these concentrations. All tests were performed in triplicate, and data were evaluated by one-way analysis of variance using GraphPad Prism 5.0 software (GraphPad Software, San Diego, CA, USA). Significant differences were determined by Dunnett's test, with a significance level of p < 0.01. Data are presented as the mean ± standard deviation.

Antimicrobial Activity
The antimicrobial activity of R and T free EOs against M. luteus (Gram-positive) and Salmonella sp. (Gram-negative) showed a MIC for both microorganisms of 0.5% (v/v) and 1.0% (v/v) of T and R EOs, respectively ( Figure 1). The EOs were added to the FS of C:QP using these concentrations.
The    The Td value showed by the C solution was lower than those presented by the C:QP:T and C:QP:R FS (Figure 2a), indicating stronger antimicrobial activity of FS added with EOs, than the C solution. Salmonella sp. (Figure 2b) was significantly inhibited in the presence of EOs, especially with C:QP:R (81.40 ± 7.01 h), which increased about 3.6 times the Td of C:QP:T (22.47 ± 1.27 h), while the inhibition shown by QP, C, and C:QP FS was significantly lower.   (Figure 3a). All FS resulted in significantly different Salmonella sp. specific growth rate (Figure 3b). Both EOs in the FS showed high Salmonella sp. inhibition, but the FS containing R EO showed the highest (µ = 0.085 ± 0.001 h −1 ). The FS producing the least inhibition was C:QP (µ = 0.111 ± 0.001 h −1 ).

Figure 3.
Effect of biopolymers alone, combined, and incorporated with EOs on the specific growth rate of (a) Micrococcus luteus; (b) Salmonella sp. µ: specific growth rate; QP: quinoa protein; C: chitosan; T: thyme essential oil 0.5% (v/v); R: rosemary essential oil 1% (v/v). Data are presented as the mean ± standard deviation. a-e Indicate significant difference (p < 0.01).
Although FS added with T EO showed antimicrobial activity against both bacteria, the effect was greater against M. luteus, which is reflected in a doubling time about twice of that showed for Salmonella sp. (Td = 22.47 ± 1.27 h) (Figure 3). Hosseini [38] reported that T EO had greater activity against Gram-positive than Gram-negative bacteria due to their thick layer of peptidoglycan (90-95%) and the presence of an outer lipopolysaccharide membrane [39]. The outer membrane has also been associated with higher antimicrobial activity of the combination of C and EOs because it restricts hydrophobic compounds [21].
However, the effectiveness of EOs in the presence of C depends on their composition, structure, and functional groups. The C:QP:R FS showed strong Salmonella sp. inhibition, with Td = 81.04 ± 7.01 h, value 3.23 times greater than that of M. luteus.
The FS C:QP did not exhibit inhibition against M. luteus, but FS with EOs exerted an Although FS added with T EO showed antimicrobial activity against both bacteria, the effect was greater against M. luteus, which is reflected in a doubling time about twice of that showed for Salmonella sp. (T d = 22.47 ± 1.27 h) (Figure 3). Hosseini [38] reported that T EO had greater activity against Gram-positive than Gram-negative bacteria due to their thick layer of peptidoglycan (90-95%) and the presence of an outer lipopolysaccharide membrane [39]. The outer membrane has also been associated with higher antimicrobial activity of the combination of C and EOs because it restricts hydrophobic compounds [21].
However, the effectiveness of EOs in the presence of C depends on their composition, structure, and functional groups. The C:QP:R FS showed strong Salmonella sp. inhibition, with T d = 81.04 ± 7.01 h, value 3.23 times greater than that of M. luteus.
The FS C:QP did not exhibit inhibition against M. luteus, but FS with EOs exerted an antimicrobial effect (Figure 4a). There was 2 log population reduction exerted by FS containing either T or R EOs after 4 h of contact time, whereas after 24 h, the reduction increased to 3 log cycles. In contrast, Salmonella sp. population was reduced by the FS containing either R or T EOs about 1 log cycle after 2 h, whereas the maximum reduction of 3 log cycles was achieved after 48 h (Figure 4b). As expected, higher inhibition was observed against the Gram-positive bacterium M. luteus.

Stability
The stability of FS was evaluated by particle size, Z potential, and polydispersity index (IPD). Z potential values were evaluated initially and after 10 d of storage (Table 1), giving values in the range −46.21 to −54.13 mV. After 10 d of storage, the emulsion C:QP:R significantly increased its z potential, whereas all FS showed increased values, although not significant. Absolute values of Z-potential greater than 30 mV indicate high emulsions stability [40], and thus, all the emulsions are stable. This effect may be associated with the presence of the non-ionic surfactant (Tween 80), which permitted a balance of what is known as hydrophilic-lipophilic numbness, allowing the maintenance of stable emulsions, even during storage [41].
The negative charge of the Z potential was attributed to the alkaline pH used to produce the emulsions. Despite chitosan positive charges, aqueous FS at pH > 6 promotes a higher number of deprotonated species followed by aggregation due to hydrogen bonds formation involving the neutralized NH2 groups of the chitosan chains [40]. In addition, a study has shown that Tween can favor negative electrical charges, which are attributed to the presence of anionic impurities [42]. The stability of FS may be enhanced by the resulting protein network, which, through hydrophilic and hydrophobic groups, allows both water and oil interactions, preventing the FS from collapsing [43].
From the results obtained, the particle size presented two main populations for each sample. It

Stability
The stability of FS was evaluated by particle size, Z potential, and polydispersity index (IPD). Z potential values were evaluated initially and after 10 d of storage (Table 1), giving values in the range −46.21 to −54.13 mV. After 10 d of storage, the emulsion C:QP:R significantly increased its z potential, whereas all FS showed increased values, although not significant. Absolute values of Z-potential greater than 30 mV indicate high emulsions stability [40], and thus, all the emulsions are stable. This effect may be associated with the presence of the non-ionic surfactant (Tween 80), which permitted a balance of what is known as hydrophilic-lipophilic numbness, allowing the maintenance of stable emulsions, even during storage [41].
The negative charge of the Z potential was attributed to the alkaline pH used to produce the emulsions. Despite chitosan positive charges, aqueous FS at pH > 6 promotes a higher number of deprotonated species followed by aggregation due to hydrogen bonds formation involving the neutralized NH 2 groups of the chitosan chains [40]. In addition, a study has shown that Tween can favor negative electrical charges, which are attributed to the presence of anionic impurities [42]. The stability of FS may be enhanced by the resulting protein network, which, through hydrophilic and hydrophobic groups, allows both water and oil interactions, preventing the FS from collapsing [43].
From the results obtained, the particle size presented two main populations for each sample. It was observed that the population is representing agglomerated particles comprised most of the FS ( Table 2). The FS with added EOs showed significantly increased particle size, especially considering the more abundant large particles (597.56 ± 37.59 nm to 677.02 ± 35.08 nm) as compared to FS without EOs. However, there was no significant difference among the smaller particles (97.55 ± 10.79 nm to 153.58 ± 13.76 nm). On the other hand, the addition of EOs generated a significant increase in the polydispersity index (PDI). PDI is the ratio of weight average to number average molecular weight, and values close to zero indicate highly monodisperse (homogeneous) FS, while values > 0.7 indicate a broad particle size distribution [44]. The results in Table 2 show PDI values between 0.5 and 0.80, suggesting that the produced FS are heterogeneous, while EOs addition caused greater heterogeneity. Quinoa protein is soluble at pH 11, where it has greater solubility and water absorption capacity. It also contains high amounts of sulfur amino acids, threonine, and tryptophan, although experiencing small protein denaturation [45]. FS were prepared using this pH, which favored the formation of aggregates, which was reflected in suspensions exhibiting two different particle sizes ( Table 2). The C:QP FS revealed one population of particle sizes close to 600 nm, which may correspond to protein aggregates resulting from the formation of hydrogen bonds within the QP and with C, together with hydrophobic interactions and covalent disulfide bonds that are formed in the protein under these conditions [46]. Disulfide bond formation is favored by the presence of TG and by heat processing of the FS, leading to the formation of agglomerates [46]. The particle size of the FS (597.56 ± 37.59 nm) increased in the presence of EOs, which was attributed to the increase in hydrophobic interactions among drops of individual EOs [47].

Rheology
The FS rheology was successfully fitted to the Casson model (R 2 = 0.995-0.998), without significant differences ( Table 3). The C:PQ-R FS showed the lowest elastic limit but was not significantly different from the other emulsions. Similarly, the calculated apparent viscosity of the different FS was not significantly different. Non-Newtonian fluids that behave as elastic solids and show shear-thinning behavior can be fitted with the Casson's rheological model, which shows infinite viscosity at zero shear rate [48].
The FS made with C:PQ-R showed the highest apparent viscosity (3.04 ± 0.002 × 10 −3 Pa.s), which according to Dapueto [49], can be related to the formation of significant amounts of protein aggregates, associated with the increased particle sizes in the presence of EOs. Proteins, like surfactants, form monolayers in aqueous solutions that can have high elastic properties behaving like sticky droplets [43], whose interactions with the EOs promote elasticity reduction.
The apparent viscosity of FS decreased when applying higher shear rates (Figure 5a). However, when reaching a speed of approximately 20 s −1 , as the shear rate increased, there were few changes in apparent viscosity. The apparent viscosity of FS decreased when applying higher shear rates (Figure 5a). However, when reaching a speed of approximately 20 s −1 , as the shear rate increased, there were few changes in apparent viscosity.  Figure 5b shows a directly proportional relationship between stress and shear rate for the FS. However, for both properties, the evaluated FS were not significantly different. Figure 6 shows M. luteus topography for the control and FS with EOs. M. luteus revealed a coccus morphology when growing in the FS without EOs (Figure 6a). However, in the presence of the R EO ( Figure 6b) and T EO (Figure 6c), M. luteus lost its characteristic spherical structure, suggesting an antimicrobial effect. The T EO showed a greater antimicrobial effect on M. luteus by causing massive destruction of its structure. Meanwhile, the effect of R EO was milder because some cells retained their spherical shape, suggesting partial inhibition, as shown in Figure 1.

Antimicrobial Evaluation of Filmogenic Suspension by Atomic Force Microscopy
Salmonella sp. is characterized by presenting the bacillus shape (Figure 7a). The R EO (Figure 7b) and T EO (Figure 7c) exerted an antimicrobial effect on this microorganism by damaging the cell surface. The EOs showed antimicrobial activity against the tested microorganisms (Figure 1), and when incorporated into the FS, they modified the growth kinetic parameters (Figures 2 and 3), exerting antimicrobial effect (Figure 4b), which was visualized by using AFM.
The control treatments exhibited rough and heterogeneous surfaces consisting of agglomerated matter with a structure that coincided with the morphologies of M. luteus and Salmonella sp. The spherical shape of M. luteus showed dimensions of 0.5-3.5 µm (Figure 6a), whereas Salmonella sp. revealed a short bar geometry (Figure 7a) [50,51]. However, when adding EOs, these shapes were lost, which was attributed to damage to the cell wall and cytoplasmic membrane with loss of structural integrity, as confirmed by the population reduction ( Figure 4). This effect might include the disruption of the proton motive force, inhibition of substrate oxidation, and disruption of DNA synthesis [52]. The hydrophobicity of essential oils enables the separation of lipids present in the cell membrane of bacteria, altering their structure, making them more permeable, and causing cell lysis [53].   (Figure 6c), M. luteus lost its characteristic spherical structure, suggesting an antimicrobial effect. The T EO showed a greater antimicrobial effect on M. luteus by causing massive destruction of its structure. Meanwhile, the effect of R EO was milder because some cells retained their spherical shape, suggesting partial inhibition, as shown in Figure 1.
Salmonella sp. is characterized by presenting the bacillus shape (Figure 7a). The R EO (Figure 7b) and T EO (Figure 7c) exerted an antimicrobial effect on this microorganism by damaging the cell surface. The EOs showed antimicrobial activity against the tested microorganisms (Figure 1), and when incorporated into the FS, they modified the growth kinetic parameters (Figures 2 and 3), exerting antimicrobial effect (Figure 4b), which was visualized by using AFM.
The control treatments exhibited rough and heterogeneous surfaces consisting of agglomerated matter with a structure that coincided with the morphologies of M. luteus and Salmonella sp. The spherical shape of M. luteus showed dimensions of 0.5-3.5 µm (Figure 6a), whereas Salmonella sp. revealed a short bar geometry (Figure 7a) [50,51]. However, when adding EOs, these shapes were lost, which was attributed to damage to the cell wall and cytoplasmic membrane with loss of structural integrity, as confirmed by the population reduction ( Figure 4). This effect might include the disruption of the proton motive force, inhibition of substrate oxidation, and disruption of DNA synthesis [52].
The hydrophobicity of essential oils enables the separation of lipids present in the cell membrane of bacteria, altering their structure, making them more permeable, and causing cell lysis [53].

Image Analysis
Entropy (E) is a measure of the heterogeneity of the images, and to quantitatively evaluate the images obtained by AFM; the grayscale co-occurrence matrix algorithm was used. In addition, the fractal dimension (FD), also known as fractal texture through the Differential Box Count (SDBC) algorithm, is a parameter related to the irregularity of the surface.
There are no reports about the E and FD calculated from image analysis evaluation to confirm the antimicrobial effect of T and R EOs on the two microorganisms used in this work, as shown in Figure 4. From Figure 8a, it is observed that the entropy of the samples of C:QP did not significantly change in the FS, incorporating the EOs when Salmonella sp. was inoculated.

Image Analysis
Entropy (E) is a measure of the heterogeneity of the images, and to quantitatively evaluate the images obtained by AFM; the grayscale co-occurrence matrix algorithm was used. In addition, the fractal dimension (FD), also known as fractal texture through the Differential Box Count (SDBC) algorithm, is a parameter related to the irregularity of the surface.
There are no reports about the E and FD calculated from image analysis evaluation to confirm the antimicrobial effect of T and R EOs on the two microorganisms used in this work, as shown in Figure 4. From Figure 8a, it is observed that the entropy of the samples of C:QP did not significantly change in the FS, incorporating the EOs when Salmonella sp. was inoculated.
The lowest entropy was observed for M. luteus in T EO suspensions (Figure 8a), suggesting a more homogeneous image relative to the control, which was attributed to the effect of microbial growth inhibition, which correlates well with the antimicrobial activity shown in Figure 4a. The fractal dimension (Figure 8b) decreased when EOs were added, and this phenomenon was observed in the roughness value (Figure 8c), except for Salmonella sp. in the presence of T EO, which may be associated with cell lysis caused by the EOs.  The lowest entropy was observed for M. luteus in T EO suspensions (Figure 8a), suggesting a more homogeneous image relative to the control, which was attributed to the effect of microbial growth inhibition, which correlates well with the antimicrobial activity shown in Figure 4a. The fractal dimension (Figure 8b) decreased when EOs were added, and this phenomenon was observed in the roughness value (Figure 8c), except for Salmonella sp. in the presence of T EO, which may be associated with cell lysis caused by the EOs.
According to our results, the FD decreased in the presence of EOs. Fractal dimension is a mathematical concept in which a set of multiple scales exhibits the same repeating pattern on each scale, which can be transferred to texture analysis. This parameter is correlated with roughness; the larger the FD, the rougher the texture [53]. The addition of EOs to the C:QP FS generally produced smoother surfaces, and thus, the roughness reduction was attributed to cell lysis promoted by the EOs.
The characteristic structures of M. luteus and Salmonella sp. provided high FD and entropy, considering it as the randomness or degree of disorder showed by the image; whereas the roughness was high for M. luteus, but lower for Salmonella sp.
The entropy is higher when all elements of the co-occurrence matrix are equal and smaller when the elements are different [54]; thus, more homogeneous surfaces are probably the result of the antimicrobial effect of EOs.

Conclusions
Essential oils imparted antimicrobial activity on M. luteus and Salmonella sp., at concentrations of 0.5% (v/v) and 1% (v/v) for thyme and rosemary EOs, respectively. When EOs were incorporated into the FS, thyme EO showed a greater inhibitory effect on Salmonella sp. and M. luteus than rosemary EO.
The FS of C:QP with and without EOs were stable but heterogeneous dispersions. EOs addition significantly increased the particle size distribution, showing two major populations. Casson's rheological model was successfully fitted to the non-Newtonian fluids (FS with and without EOs), that behave like an elastic solid. Formation of protein aggregates allowed greater interaction between protein and aqueous phase, which increased the apparent viscosity of the C:QP-R FS, compared to the control.
AFM permitted the evaluation of the characteristic structures of M. luteus and Salmonella sp. and confirmed the antimicrobial activity of the EOs, which was also monitored by significant changes in their kinetic parameters. Image analysis of FS containing T and R EOs showed that entropy, roughness, and FD changes might correlate with the generation of more homogeneous surfaces and with the antimicrobial effect. The properties of the active FS may be of use as an alternative coating material for food preservation.