EMULSIFIED FILMS PRODUCED WITH PROTEINS EXTRACTED FROM WHITEMOUTH CROAKER BYPRODUCTS : DEVELOPMENT AND CHARACTERIZATION

This study aimed to develop and characterize biodegradable films from myofibrillar proteins from whitemouth croaker (Micropogonias furnieri) using stearic acid (SA) to improve its technological properties. The proteins extracted were lyophilized, characterized, and used to prepare the films. The filmogenic solutions were dried on a silicone stand at 35 °C for 14 h in an incubator oven according to the casting method. The region with the best mechanical, physical, and barrier properties in the films was defined by a full factorial design. The lyophilized myofibrillar proteins (LMP) had protein content of 96.03%. The analysis of the experimental design results indicated the best conditions to prepare the optimized film were 2.84% LMP, 3.18% SA, and 78.41% SDS (sodium dodecyl sulfate). The control films were prepared with 2.84% LMP and 30% plasticizer. The optimized film had significantly lower water vapor permeability (5.87E-11 g m m-2 s-1 Pa-1), higher tensile strength (6.35 MPa), and lower elongation (235.60%) compared with control (p≤0.05). It also had lower transparency values and excellent UV barrier property, which indicates a tendency to opaque. Thermal stability was good and the microstructure revealed a structural change in the filmogenic matrix, confirmed by x-ray diffraction, which indicates influence of SDS and SA on film crystallinity. Solubility increased by 22% and swelling decreased slightly in the optimized film compared to the control. The results obtained represent a positive contribution with the use of fish byproducts by applying alternative, sustainable technologies.


INTRODUCTION
Excessive use of synthetic packaging materials from non-renewable sources and their disposal in the environment has been causing growing concern and requires alternatives to control and/or decrease the use of this type of material.On the other hand, the fish industry produces large amounts of byproducts such as skins, scales, and parings during fish filleting.Fish filleting parings contain proteins that can be used to create biodegradable films with several applications, besides minimizing the environmental impact generated by the disposal of these residues (PASCHOALICK et al., 2003;PIRES et al., 2011;HALAL et al., 2016;TAO et al., 2015).
The use of biodegradable films as food packaging is a global trend that has significantly expanded in recent years, driving studies on the development and characterization of such films (KHWALDIA et al., 2010).Biodegradable films obtained from biological materials act as a barrier against external elements, protecting packaged foods from physical and biological damage, decreasing compound volatilization and moisture loss, and extending product shelf life.
Films prepared with proteins have excellent mechanical and optical properties and are good barriers to O 2 and CO 2 in environments with low relative humidity.However, in the presence of high moisture, the film may become susceptible to deterioration, which is not technologically viable (GALLO et al., 2000;DAVANÇO et al., 2007).The water vapor barrier property is considered one of the most important because, depending on the water vapor transfer rate, spoilage processes may begin in the packaged food (PRODPRAN et al., 2007;CARPINÉ et al., 2015).
Researches on the preparation of composite films (added with hydrophobic components) have been carried out to improve permeability, strength, flexibility, and nutritional value.A hydrophobic component is added to the film-forming suspension of emulsified composite films, where the lipid component acts as water vapor barrier and the protein or polysaccharide provide the mechanical characteristics required for a good film and oxygen barrier (ANKER et al., 2002).When the lipid is applied onto the protein film layer, the resulting film is called bilayer (GALLO et al., 2000).
Among the lipid materials studied, waxes, long-chain saturated fatty acids, and fatty alcohol were the most effective in providing moisture barrier properties to hydrocolloid films (YANG and PAULSON, 2000).However, emulsifying or surfactant agents are often required to improve the stability of lipid particles in the protein matrix.Surfactants are compounds that have interface activity between two phases, such as air-water, oil-water, and on the surface of solids as they have two distinct regions of the same molecule: one polar hydrophilic and the other non-polar hydrophobic (DAVANÇO et al., 2007).
This study aimed to develop and characterize biodegradable films from myofibrillar proteins from byproducts (parings) from whitemouth croaker (Micropogonias furnieri) processing using stearic acid (SA) to improve its technological properties.

METHODS
Filleting parings from whitemouth croaker (Micropogonias furnieri) were donated by the industry G-Pesca, located in the city of Bragança, PA, Brazil.The parings were placed in polyethylene packaging, stored in an isothermic box with flake ice, and transported to the laboratory.They were then hygienized by immersion in 5 ppm (mg L -1 ) chlorinated water for 5 min and the spines and skin residues were removed and manually pressed to remove excess water and obtain the muscle.Next, the muscle was ground in a cutter food processor (Filizola, São Paulo, SP, Brazil) for 60 s, placed in polyethylene bags, vacuum-packaged, and frozen at -18 °C.

Extracting lyophilized myofibrillar proteins
The methodology proposed by ZAVAREZE et al. (2012) was used, with modifications, to obtain the lyophilized myofibrillar proteins (LMP).The ground muscle was added with five volumes of 50 Mm NaOH at 7 °C for 5 min, centrifuged at 10,000 rpm for 3 min at 4 °C in a refrigerated centrifuge (Thermo Fisher, Multifuge X1R), and filtered.This process was repeated three times.After those steps, the myofibrillar proteins were placed onto stainless steel trays, frozen at -18 °C, and lyophilized (Liotop, L101) at -60 °C for 48 h.The lyophilized proteins were then sieved (Tyler 20 0.84 mm mesh), weighed, vacuum packaged, and kept at -18 °C.Myofibrillar protein yield was obtained from the ratio between the lyophilized protein and the initial amount of raw material (parings), expressed in g LMP 100 g -1 muscle (KAEWRUANG et al., 2013).

Characterizing the raw material (parings) and LMP
Analyses of proteins, lipids, moisture, and ash (AOAC, 1997) were carried out on the raw material (parings) and on the myofibrillar proteins.

Experimental design
In order to obtain the optimized film, a full factorial 2 3 experimental design was carried out with 17 experiments (Table 1), being eight factorial assays (combination between levels ±1), three assays at the central point (three variables at level 0), and six assays at the axial levels ±α.The input (independent) variables were concentrations of LMP, AE, and SDS.The dependent variables were water vapor permeability (WVP), tensile strength (TS), and percent elongation (E).The definition of the levels of the variables studied was based on the literature (DAVANÇO et al., 2007;LIMPAN et al., 2010).
The response surface methodology was used to verify the behavior of the system, combining the independent and dependent (response) variables.For each response, the significance of the variables or interactions in the polynomial equation described (Equation 1) was verified.After the exclusion of the non-significant effects, the equations and influence plots of the independent variables on the dependent ones were obtained.The results enabled establishing the concentrations of myofibrillar proteins, fatty acid, and surfactant to obtain a film with low WVP and high TS and E values.Thus, some criteria were followed based on the restrictions mentioned for the minimum, intermediate, and maximum values of each response aiming to find the optimal operational values of the indepen dent variables that simultaneously meet the requirements of the response variables (desirability function).

Film preparation
The films were prepared according to the methodology described by DAVANÇO et al. (2007) with adaptations.Initially, the filmogenic solutions with different concentrations of LMP, SA, and SDS were prepared and 30% glycerol was added to each solution according to the full factorial design (Table 1).
The pH of the solution was adjusted to 12 with 2 M NaOH, the solution was homogenized at 10,000 rpm for 5 min (Turratec/Tecnal, TE-102), and then placed in a water bath at 70 °C/30 min.According to the casting method, 130 mL of the filmogenic solution obtained were added to a silicone support with 22 cm diameter and 3 cm height and dried in an incubator oven (Quimis, Q315M) at 30 °C for 16 h.Next, the films were placed in polyethylene packaging and stored at room temperature.For comparison purposes, a control biofilm was prepared with the same concentrations of LMP and plasticizer as the optimized film, however, with no added SA or SDS.

Thickness
Film thickness was measured using a digital micrometer with 0.001 mm resolution (Insize, model IP54).Eight random spots were measured around each film while observing a distance of 60 mm from the edge (ZAVAREZE et al., 2012).

Mechanical properties
Film TS and E were measured in a texture analyzer (QTS, Brookfield) at room temperature (ASTM, 1996).The initial gap of the grips and probe velocity were 50 mm and 1 mm.s -1 , respectively

4/13
( ZAVAREZE et al., 2012) and the samples were cut into 100 mm x 25 mm strips.TS was calculated as TS = Fm/A (Fm=maximum force at the moment of film rupture (N) and A=area (m 2 ) of the film cross-section).Elongation was calculated as E=d T /d initial x 100 (d T =total distance at the moment of rupture (mm) and d initial =initial distance of 50 mm of grip separation).

Water vapor permeability
The modified American Society for Testing and Materials (ASTM, 1996) method, described by ARFAT et al. (2014), was used.The films were placed onto a glass permeation container with 4.5 cm diameter and 7.0 cm height containing 10 g silica gel (0% RH; 0 Pa water vapor pressure at 30 °C) and fixated with silicone adhesive.Next, the permeation containers were placed in desiccators with distilled water at 30 °C (99% RH; 4,244.9Pa vapor pressure at 30 °C) and weighed every hour for 10 h.WVP (g m -1 s -1 Pa -1 ) was calculated by WVP=W.X/At.ΔP (W=weight gain by the desiccant (g); X=film thickness (mm); A=surface area of the exposed film (m 2 ); t=incubation time (h); ΔP=partial pressure difference (Pa)).Three specimens were used for the tests.

Solubility and swelling
In order to determine solubility, the initial dry matter of the films was established in an oven at 105 °C/24 h and they then were immersed in 50 mL water.The system was stirred in a refrigerated shaker incubator (Cielanb, model CE-725B) at 150 rpm for 24 h at 25 °C.After this period, the samples were removed and dried (105 °C for 24 h) to determine the dry matter undissolved in water (GONTARD et al., 1994).
In order to determine swelling, a sample 2 cm in diameter was weighed, immersed in 75 mL distilled water, and placed in an incubator oven (Quimis, Q315M) at 25 °C for 6 h.After this procedure, excess water in the films was removed with filter paper and the sample was weighed again.Swelling was determined as the weight gain of the sample divided by the initial weight of dry solids (LEE et al., 2004).

Light transmittance and transparency
Light transmittance was measured in the ultraviolet and visible ranges (200 and 800 nm) in specimens cut into rectangles and placed in the inner side of the cuvette.Transparency was determined at 600 nm and calculated by T=-logT/e (T=transmittance at 600 nm; e=film thickness in mm) according to the ASTM D1746 method cited by ARFAT et al. (2014).Both parameters were quantified using a spectrophotometer (Biospectro,.

Differential Thermal Analysis (DTA) and Thermogravimetry (TG)
The thermal analyses were carried out in a SHIMADZU DTG-60AH device at 25 to 600 °C using heating rate of 10 °C/min and nitrogen flow at 40 mL/min in an alumina crucible.Data were acquired and treated in the software TA60 version 2.21.

X-Ray Diffraction (XRD)
The samples were placed onto the glass stand of the device and exposed to Cu radiation (Kα1=1.540598Å) submitted to 40 kV voltage and 40 mA current of l=1.78897 x 10 -1 nm in a BRUKER D8 Advance diffractometer with Bragg-Brentano geometry and LynxEye detector with 0.4 s scan time in the Bragg-Brentano geometry of 5<θ<60, 0.6 mm divergence slit, 2.5° Soller slit, and Kβ Ni filter.The diffractograms were collected with 0.02° step angle and 0.4 s step time.

Film surface microscopic characterization
The upper surface microstructure of the film was analyzed in a scanning electron microscope (SEM) (Tescan, VEGA3) with acceleration voltage of 5 kV, electron beam current of 85-90 µA, and 15 mm work distance for the secondary electron images.The samples were metallized twice with gold/palladium with 5 mA for 120 s in a Quorum Technologies SC7620 sputter coater.

Statistical analysis
The results underwent statistical analysis using the software Statistica  10.0 (STATSOFT, 2011) by assessment of effects, analysis of variance, and response surface.

RESULTS
The raw material extracted and lyophilized had high myofibrillar protein content (96.03% dry basis), important for the formation of the biopolymer matrix, and low lipid content (0.94% dry basis), which were partially removed in the successive rinsings during extraction and centrifugation.However, ash concentration (2.21% dry basis) compared to the raw material/parings (2.88% dry basis) did not decrease much due to the presence of salt residues from the myofibrillar protein extraction process.Yield (13.25%) is highly influenced by the extraction of myofibrillar proteins, which eliminates lipids, minerals, bloods, sarcoplasm proteins, and stromal proteins.Moisture decreased between the raw material in natura (7.21% dry basis) and the lyophilized myofibrillar proteins (0.06% dry basis) due to the lyophilization process.

Film optimization
The results of the full factorial 2 3 design for whitemouth croaker myofibrillar protein film optimization is presented in Table 1.
Figure 1a shows the WVP level curves generated by the models proposed (Table 2) considering the mean points of the effects.It is noteworthy that this study expects to obtain low-permeability 5/13 films.The lowest WVP values are obtained at 60 to 80% SDS and 3.18 to 10% SA (or above 30%) combined with 1.5% LMP.A negative correlation of SDS with WVP can be observed, i.e., when the concentration of this surfactant increases, WVP decreases.The opposite effect occurs with LMP concentration on WVP, where higher protein concentrations increase permeability.

6/13
Figure 1b shows the negative influence of the SA x SDS interactions since the highest TS values are observed with the highest SDS and low SA concentrations.A similar behavior was observed for the LMP x SA interaction.The LMP x SDS interaction indicates the region of the highest levels (+1) in both effects favor the maximum TS value.This study aims to obtain films with high TS values and the best combination obtained for the three effects are SDS (60 to 80%), SA (3.18 to 15% or 25% to 36.82%), and LMP (2% to 2.84%).
Figure 1c shows that the highest E values are at the central point of SDS (60% to 80%) associated with the region with low SA level (3.18% to 10%) and high LMP level (2% to 2.84%).The increase in whitemouth croaker LMP concentration improves the mechanical properties (TS and E) of the film (Figure 1b and 1c).

Definition of the optimal process conditions: desirability function
After the statistical analysis of the results, criteria were adopted to determine the minimum, intermediate, and maximum values of each response of the independent variables that simultaneously meet the requirements of the response variables (desirability function) with high values for the mechanical properties (TS and E) and low values for permeability (WVP).
Based on the results of the desirability function, it was verified that, in order to obtain the maximum value of this global function (1.0), LMP, SA, and SDS concentrations must be 2.84%, 3.18%, and 78.41%, respectively.The film prepared with these values will have minimum WVP values (5.64E-11 g m -1 s -1 Pa -1 ) and maximum TS (5.88 MPa) and E (253.58%) values.

Characterization of the optimized film
The results of mechanical, physical, barrier, and color characterization of the control and optimized films are presented in Table 3.
The parameters thickness and breaking force of the films analyzed (optimized and control) did not significantly differ (Table 3).The optimized film had the highest TS and the lowest E compared with control (p≤0.05).Adding SA and SDS impacted WVP reduction in the optimized film (p≤0.05) as the control film had 31% lower WVP.The solubility of the control film was 22% lower than that of the optimized one (p≤0.05).Water solubility of the films is directly related to the intermolecular interactions among its components, according to their structure and chemical characteristics (hydrophilicity and hydrophobicity), thus, the solubility of the optimized film was expected to decrease with the addition of SA and SDS.
Table 3 shows that the integrity of the optimized film was reduced by approximately 16%, with no significant difference compared with control.
Color parameters were impacted by SA and SDS addition to the film (p≤0.05).Color difference (ΔE) in the optimized film was greater than in the control, which indicates a more opaque film.

Light transmittance and transparency
Table 4 shows the effect of adding SDS and SA to the optimized film based on whitemouth croaker myofibrillar proteins on the barrier properties against radiation in the UV/Vis region as a function of transmittance at different wavelengths (200-800 nm) and transparency at T600 mm -1 .
In the wavelength range assessed, the transmittances obtained for the optimized film were significantly lower (Table 4), which indicates less transparent films with barrier property in the UV/Vis region.This property is considered important for food packaging since UV light causes oxidative spoilage of foods, leading to nutrient loss, loss of color, and off flavors (MARTINS et al., 2012).The optimized film had lower transparency (p≤0.05)than the control since the lower the value, the higher the transparency (Figure 2).Thus, adding SA and SDS directly impacted the light transmittance and transparency of the films.

Differential Thermal Analysis (DTA) and Thermogravimetry (TG)
The thermal behavior of the control and optimized films are shown in Figure 3 and the events described, in Table 5.Both films had four stages (events) of mass loss.The first event, The second event, corresponding to exothermic peaks, was observed at peak temperatures of 169.42 °C (P 2 =15.73%) and 159.16 °C (P 2 =9.84%) for the control and optimized films, respectively, and is related to the breakdown of the protein fractions attributed to oxidation and combustion of this organic matter.
The third event (Tp 3 =472.42°C/P 3 =54.92%for the control film and Tp 3 =402.93°C/P 3 =58.37%for the optimized film) and the fourth event (Tp 4 =555.57°C/P 4 =58.15% for the control film and Tp 4 =449.90°C/P 4 =61.22% for the optimized film) are associated with the breakdown of glycerol, SA, and SDS.
Endothermic events (383.61, 481.72, 528.39, and 566.61 °C) are observed in the control film and one peak (396.91 °C) in the optimized film due to mass loss.Exothermic peaks are also verified for both films, at 492.91 and 533.54 °C for the control and 452.15 °C for the optimized.Such anomalies observed in the TG-DTA curves were due to the combustion of organic matter, in which the sample temperature exceeds that of the oven (of reference).Moreover, both films had residual mass, which represents the content of animal charcoal at 600 °C (Table 5).

X-Ray Diffraction (XRD)
The control film has a predominance of polymer structure and a single peak can be identified, even enlarged, at an angle of 9.36° (d spacing=9.45Å), which indicates the formation of a quite amorphized structure, i.e., disordered and little detectable by x-ray diffraction with very little degree of crystallinity (Figure 4).
In the optimized film (Figure 4), the crystalline regions are clearly observed in the highest intensity peak of 2ϴ=6.73°(d spacing=13.13Å), showing the shift of peak d001 by the presence of SDS in the structure, as well as in the four less expressive diffraction peaks (2ϴ) referring to SA at angles of 20. 42°, 20.77°, 21.9°, and 22.21° (4.34, 4.28, 4.06, and 4.00Å of d spacing, respectively).The variation in interplanar spacing in the optimized film was between D001=13.13 Å and D001=2.28Å, indicating crystallinity as verified by the presence of well-defined peaks in the diffractogram (Figure 4).The observations indicate the influence of SDS on the crystallinity of the optimized film was greater than that of SA, suggesting the protein matrix conformation changed.

Film surface microscopic characterization
The surface micrographs of the control and optimized films are shown in Figures 5 and 6, respectively.
Visually, the control film (Figure 2) had homogeneous, transparent, flexible aspect with no evidence of air bubbles or protein clumps.However, the microscopic analysis (Figure 5) revealed that the film has heterogeneous structure with empty spaces, bubbles, and protein clumps (or clusters), besides disperse insoluble particles in low-concentration (0.05 M) saline solution.
Figure 6 shows the influence of adding SA and SDS to the structure of the optimized film.A structural change in the Table 5. Thermal events related to the thermogravimetric analysis of the control and optimized films.

Films
Event    10/13 filmogenic matrix can be observed compared to the control film.SDS concentration allowed incorporating SA, homogenizing the filmogenic matrix, which is compatible with the low WVP found for this film.Nonetheless, in spite of the presence of grooves and lumps in this film and of its rough, irregular surface, the film had excellent physical and mechanical characteristics.
SDS improved the water vapor barrier of the films in this study, which, according to RHIM (2012), is a desirable attribute in emulsified films to be used as food packaging.A similar behavior was also observed by DAVANÇO et al. (2007), who reported adding surfactants was effective in increased the water vapor barrier in the film.
In the present study, the best TS values were found with high protein concentrations since an increase in the number of protein chains per surface unit usually leads to a higher number of potential intermolecular interactions (CUQ et al., 1996 andKAEWPRACHU et al., 2016).
The elongation results show that, of all effects assessed, LMP had the greatest influence on this response (Figure 1c) as the higher protein content results in increased intermolecular aggregation (KAEWPRACHU et al., 2016).Similar results have been reported for films made with blue marlin sarcoplasmic protein (IWATA et al., 2000) and tilapia myofibrillar proteins (KAEWPRACHU et al., 2016).
Adding SDS and SA did not impact the thickness of the optimized film, however, PRODPRAN et al. (2007) andOLIVEIRA et al. (2012) reported increased thickness in fish myofibrillar protein films added with palm oil and gelatin films with coconut oil and Tween 20 surfactant, respectively.The increase in thickness may be influenced by the non-association of the surfactant with protein chains (GONTARD et al., 1994).This study used protein concentrations (2.84%) which led to thicker films (0.147 and 0.176 mm).DAVANÇO et al. (2007) obtained 0.095 mm thickness for gelatin films added with 10% SA and 70% SDS.The type of application of the film will indicate its ideal thickness.ZAVAREZE et al. (2012), when analyzing films with 3% whitemouth croaker myofibrillar proteins, obtained similar TS (4.09 Mpa) and thickness (0.137 mm) as the control film (Table 3).MONTERREY-QUINTERO and SOBRAL (2000) and PIRES et al. (2013), when studying king weakfish myofibrillar protein films added with thyme oil and Nile tilapia films, obtained TS values of 2.2 N and 6.67 N, respectively, which are lower than the results in this study.KAEWPRACHU et al. (2016) andROSTAMZAD et al. (2016) found WVP values close to those of the control film for myofibrillar protein films of tilapia and silver carp, respectively (Table 3).PIRES et al. (2013), when studying king weakfish protein films added with essential oils, obtained WVP of 4.2E-11 g m m -2 s -1 Pa -1 , similar to the value of the optimized film.DAVANÇO et al. (2007) stated that the permeability of biodegradable films strongly depends on their structure and that saturated fatty acids, such as SA, are the most effective in controlling moisture migration than unsaturated ones.
The solubility result of the optimized film can be compared with the studies by DAVANÇO (2006), ARTHARN et al. (2009), andPIRES et al. (2013), who also found increased solubility when adding lipids to fish-protein films.The denaturing property of SDS reduces the intermolecular forces among between polymer chains, thus destabilizing the film structure (DAVANÇO, 2006).Film solubility may also depend on factors such as species, type of muscle, and film preparation (PIRES et al., 2013). Nevertheless, HALAL et al. (2016) noticed that adding palm oil did not significantly impact the solubility of whitemouth croaker protein films.
The lower swelling is a consequence of the higher hydrophobicity of the film with SA and SDS (p>0.05),likely favored by a matrix with higher density and, consequently, lower molecular diffusivity of water and lower swelling capacity, which is confirmed by the low WVP value of the optimized film and by the mechanical analysis results.The swelling index in films allows verifying beforehand the degradation outlook, which is related to the level of hydration of the system.
High-solubility films, such as the ones in this study, can be used to store individual food portions that will be dissolved during preparation, to package fresh fruits and vegetables, and to be placed in trays to absorb exudates from cuts of poultry, beef, and/or fish (GONTARD et al., 1996;RHIM, 2012).
Several authors, when analyzing the color parameters of films added with oils and surfactants, have found similar results as the present study.PIRES et al. (2013) also obtained high luminosity (92.45 to 93.72%) for king weakfish protein films added with thyme oil.ARTHARN et al. (2009), TONGNUANCHAN et al. (2011b, 2015), reported that tilapia films were more yellowish than the control film as evidenced by the increase in b* and ΔE values.This characteristic may favor the protection of packaged foods against the oxidative spoilage caused by light, although films should preferably be colorless to simulate the appearance of synthetic films (MARTINS et al., 2012).
Adding SA and SDS to the films directly influenced their light transmittance and transparency by increasing the barrier property in the UV/Vis region, an important feature in food packaging as UV light may cause oxidative spoilage and lead to nutrient loss, loss of color, and off flavors (MARTINS et al., 2012).ARAÚJO (2015) and ( 2016) found transmittance at the same wavelength of 78.73 to 87.97% and 75.2 to 87.6%, respectively, for fish myofibrillar protein films, i.e., slightly more transparent than the 11/13 films in this study.PIRES et al. (2013) andHALAL et al. (2016) also found lower transparency in fish protein films when oils were added.That can be due to the presence of lipid droplets dispersed in the film and the physical state of the lipid (solid) at room temperature (HALAL et al., 2016).
The presence of marked peaks in the diffractogram of the optimized film at wide angles (WAXS) indicates the presence of crystalline regions (FAKHOURI, 2009).A similar behavior was found by CORTEZ-VEGA (2011) for whitemouth croaker protein films.AKANDA et al. (2015) found a peak at 2ϴ=21.55° and ALMEIDA et al. (2012) found peaks at 2ϴ=20.60, 21.71, and 24.05° for SA in fish-protein films.ELSAYED et al. (2014) detected a peak at 2ϴ=5.27° while YU et al. (2011) reported a peak at 2ϴ=6.5° for SDS.
The control film had higher mass loss at lower breakdown temperature than the optimized films in the first decomposition event of the thermal analyses, which may be attributed to its higher moisture control (LIU et al., 2009;MARTINS et al., 2012).That indicates greater evaporation of free water from the structure of the control film, which was also described by TONGNUANCHAN et al. (2014), CARPINÉ et al. (2015), and BATISTA (2016).The optimized film had higher hydrophobicity due to the addition of SA, which is corroborated by HALAL et al. (2016).
The third event is associated with the breakdown of the fraction of myofibrillar proteins of higher molecular weight and of glycerol since the boiling point of this plasticizer is around 182.0 °C (HALAL et al., 2016).In the optimized films, it is also associated with the thermal breakdown of SA, whose peak breakdown temperature is 236.33 °C (MATOS, 2012,) and with the thermal breakdown of SDS at 213 °C (YU et al., 2011).
SEM shows the action of SDS in incorporating SH into the optimized film.A similar behavior was found by DAVANÇO et al. (2007) when verifying the efficiency of SDS (60%) in the incorporation of SA into the filmogenic matrix because microscopy reveals fat droplets in films without the addition of this surfactant.The same was reported by ROSTAMZAD et al. (2016) for silver carp protein films and ARAÚJO (2015) for films based on myofibrillar proteins of gilthead bream.SOUZA et al. (2004) suggested that the presence of chinks and/or irregularities in film structure may compromise its integrity by changing its functional properties and those related to film surface such as sheen and water absorption.Those factors confirm the high solubility found in this study for the optimized films as well as the results obtained for the color parameters, such as ΔE.

CONCLUSION
Adding stearic acid and dodecyl sodium sulfate to the optimized film led to a reduction by 31% in water vapor permeability, greater tensile strength (6.35 MPa), and lower elongation (235.60%)compared with the control film (p≤0.05),producing, however, strong and flexible films.The lower transparency and more yellowish color of this film suggest it can be used in food sensitive to oxidation due to its better UV barrier property.
The optimized film had good thermal stability despite the higher mass loss compared with control, with 22% higher solubility and slightly lower swelling.The results obtained represent a positive contribution in the field of emulsified films with the use of fish byproducts and allow for the application of alternative sustainable technologies to obtain quality products, besides minimizing environmental impact.

Figure 3 .
Figure 3. TG and DTA curves of the control and optimized films.

Table 1 .
Coded matrix of the full factorial 2 3 design for whitemouth croaker myofibrillar protein film optimization.

Table 2 .
Reduced model for WVP, TS, and E as a function of the independent variables, F-test, R 2 , and P.

Table 3 .
Mechanical, physical, barrier, and color characterization of the control and optimized films.

Table 4 .
Light transmittance and transparency of the control and optimized films.