Valorization of Agro-Industry-Rejected Common Bean Grains for Starch Film Development: Advancing Sustainable and Comprehensive Resource Utilization

Abstract

This study examines the potential use of rejected and discarded grains from the common bean industry as a starch source for producing plasticized films with glycerol. The observed morphological characteristics of starch granules from discarded grains were diverse, with round, oval, and kidney-like shapes and sizes ranging from 7 to 34 µm. We determined the pasting profile: the pasting temperature (GT) fell between 72 °C and 74 °C, while the peak viscosity (P_v) demonstrated a significant rise at 10% and 15% starch concentrations. To better understand pasting behavior, mathematical modeling was employed to predict P_v behavior, with an R² value of 0.98. All film formulations were successful, yielding transparent, homogeneous, odorless, flexible films with smooth surfaces. Scanning electron microscopy analysis of the films revealed a flawless surface devoid of fissures, cracks, and pores, displaying a rough texture with a consistent structure and some starch granules resembling empty sacks due to amylose and amylopectin leaching. The highest tensile strength was observed with 6% starch and 1.5 mL of glycerol and the lowest with 4.5% starch and 3.9 mL of glycerol. The findings suggest that starch derived from discarded grains from the bean industry has unique characteristics and properties, making it a promising alternative source for intelligent packaging development.

1. Introduction

Environmental pollution has emerged as a significant and escalating global issue, driving the creation of new biodegradable materials for packaging and coatings in the food industry [1]. The primary objective is to develop biodegradable materials that possess properties akin to traditional petroleum-based polymers without increasing production costs [2]. Among the naturally occurring polysaccharides utilized in biodegradable materials, starch stands out due to its natural availability, high yield, and ability to form films [3].

Starch is extensively utilized globally across various sectors, especially in the food industry, where it is used to enhance product properties such as thickness; additionally, it serves as an emulsion stabilizer and a gelling agent. Beyond food applications, starch is employed in industries such as plastics, pharmaceuticals, and paper manufacturing. The worldwide production of starch sourced from cereals such as rice, maize, wheat, and barley amounts to approximately two billion tons annually. By 2050, due to population growth, the global demand for food starch is projected to increase to approximately nine billion tons. Furthermore, urban expansion will constrain the availability of arable land for cultivating starchy crops, potentially jeopardizing overall food security [4].

Starch is a polysaccharide and the primary storage carbohydrate in plants. It is organized into semicrystalline microscopic granules varying from 1 to 100 µm in size depending on the botanical origin. Starch mainly comprises two macromolecules, amylose and amylopectin [5], where the former, a predominantly linear molecule with α-1,4 glycosidic bonds, forms an amorphous region, while the latter, a highly branched macromolecule with α-1,4 and α-1,6 glycosidic bonds, forms a crystalline region. Commonly, the amylose content in starch is 20–30%, and that of amylopectin is 70–80%, depending on the source [6]. Due to the presence of both these molecules, which represent amorphous and crystalline zones, starch is defined as a semicrystalline compound [7].

Agro-industrial residues are currently receiving significant research attention, as many of their components can be utilized as raw materials for creating value-added products. Although there are many reports on the reutilization of agro-industrial wastes and byproducts, most research focuses on bio-refineries, where waste is used to obtain compounds of interest such as butanol, organic acids, and phosphates. Nevertheless, there is a lack of information on the use of this type of byproduct to produce various ingredients, including proteins, fibers, and starches.

Today, the extraction of starch from agro-industrial residues is a rapidly expanding field, motivated by objectives to reduce environmental impact, adopt circular economy principles, and extract value from waste. Residues such as cassava peels, bagasse, plantain peels, mango seeds, potato waste, avocado seeds, and tiger nut horchata residue are recognized as valuable alternative sources of starch. These often discarded or underutilized materials can be transformed into high-quality starch for applications in food, packaging, bioplastics, and bioenergy, thereby contributing to pollution reduction and promoting economic sustainability.

The need for research on novel raw materials based on starch processing residues stems from the fact that the primary sources of this biopolymer, such as corn, potato, and wheat, are the base of the human diet, and their production could be insufficient to meet human needs [8]. In this sense, the starches present in legumes are considered non-conventional sources that can be used as ingredients in the same way as the starches from cereals and tubers given their very similar properties; for this reason, the use of starches from legumes such as common beans is a promising avenue to address the potential shortage in starchy crops.

Exploring alternative starch sources offers advantages such as sustainability and byproduct utilization [9]. However, because of their plant origin, starches have varied characteristics, which poses challenges such as limited solubility, rapid retrogradation, low thermal stability (especially instability during freezing), reduced emulsification capacity, increased syneresis, and low shear strength, restricting their use in food and non-food applications [10]. Native starches are characterized by hydrophilicity and poor mechanical properties, with their hygroscopic nature and low tensile strength limiting their packaging applications. Therefore, modifying starch structure is essential to improving its physical, morphological, and mechanical characteristics [11,12]. A method of altering starch properties is the incorporation of plasticizers, which can enhance flexibility and processability, specifically reducing friction and viscosity, improving plasticity, softening materials, and easing handling [13]. Consequently, plasticizers play a crucial role in producing functional films. Although water is the most effective and widely utilized plasticizer for starch, various other substances can also be used [14], with the most common ones being hydrophilic, low-molecular-weight polyols (glycerol, sorbitol, xylitol, polyethylene glycols, etc.) [15]. Glycerol is a widely utilized, non-toxic, and safe-to-use substance in the food industry that functions as an internal lubricant, reducing the stiffness of starch macromolecules at room temperature [16].

Globally, the common bean is one of the most consumed legumes, resulting in a cultivated area of 78 million hectares with an annual production of 70 million MT. Mexico is the seventh worldwide producer of common bean grains, with approximately 1,288,806 MT. Unfortunately, 3% to 10% of the total production consists of discarded grains that cannot be traded because they are broken, chopped, or reduced in size, thus failing to meet the required quality standards. This byproduct has low commercial value because of its physical properties; however, because the nutritional components of rejected common bean grains remain intact, they can be processed to obtain valuable products. The development of films utilizing starches from conventional and non-conventional sources is the object of many current research studies; however, using starch derived from discarded grains as a basis for film formation needs to be better documented. In this context, the present work contributes to circular economy research by sustainably valorizing agro-industrial residues—specifically, rejected or discarded common bean grains from the Mexican legume industry—as a source of starch. The recovered starch is evaluated for its potential in forming biodegradable films with the solution casting method. A comprehensive understanding of the functional properties and applications of legume-derived starch can unlock new opportunities in the legume sector, promoting resource efficiency, waste reduction, and the development of high-value bioproducts.

2. Materials and Methods

2.1. Materials

Agro-industry-rejected grains (RGs) of common bean (Phaseolus vulgaris L.) var. Pinto Saltillo from the November 2019 harvest were sourced directly from producers in Vicente Guerrero, Durango, Mexico. The raw beans were manually cleaned, washed, and dried at room temperature. Then, they were ground into flour and sieved to attain a maximum particle size of 420 μm, and the flour was stored in sealed plastic bags at room temperature for further analysis.

2.2. Starch Extraction and Composition

Starch was extracted according to the methodology described by Lee et al. [17] with some mild modifications. Briefly, 100 g of RG flour was dispersed in 1000 mL of distilled water (the flour/water ratio was 1/10 weight/weight), and the pH was adjusted to 9.5 with 0.1 M NaOH under constant stirring. The resulting solution was sieved using different mesh sizes (10, 20, 40, 60, 80, 100, and 120 mesh number) to separate fiber from starch. The solution obtained was centrifugated at 3500 rpm for 20 min, and the supernatant formed above the pellet was discarded. The starch precipitate was collected, washed twice with distilled water, and dried at 40° C for 24 h. Proximal analysis was conducted by employing the Association of Official Analytical Chemists’ methods to determine the contents of moisture and total solids (930.04:2007 [18]), ash (930.05:2007 [19]), fat (930.09:2007 [20]), and protein (978.04:2007 [21]).

2.3. Determination of Amylose/Amylopectin Contents

Amylose/amylopectin contents were determined using the International Organization for Standardization method. We prepared a test solution by mixing 100 mg of starch with 1 mL of 95% ethanol and 9 mL of 1.0 M NaOH, and distilled water was added to obtain a total volume of 100 mL. Next, 5 mL aliquots of the starch dispersion were added to 100 mL volumetric flasks together with 1 mL of 1 M acetic acid and 2 mL of 2% iodine, and distilled water was further added to dilute each mixture to 50 mL. Absorbance was measured at 720 nm by using a Jenway 6320D spectrophotometer (Jenway Limited, Essex, UK).

2.4. Granule Morphology

The morphology of the starch granules was examined using optical microscopy (OM) with an Axio-Lab A.1 system (ZEISS, Goettingen, Germany). Starch was dispersed in water, and granules were observed at 10× and 40× magnifications with and without polarized light. Starch granules were also observed with a JSM-6510LV scanning electron microscope (JEOL, Tokyo, Japan). Starch sample preparation for SEM analysis involved placing starch powder on double-sided carbon adhesive tape attached to a base and gold-coating it (Desk V, Denton Vacuum, Moorestown, NJ, USA, EEUU) for two minutes; the samples were observed by operating the scanning electron microscope at 25 kV.

2.5. Starch Pasting

Starch pasting properties were analyzed with a Discovery Hybrid Rheometer 3 (TA Instruments, New Castle, DE, USA, EE. UU.) at a rotational speed of 30 rad/s maintained constant throughout the process. Initially, the starch samples were maintained at 30 °C for 120 s and then heated to 90 °C with a heating ramp of 5 °C/min; the temperature was kept constant at 90 °C for 300 s, and then, the samples were cooled to 30 °C at 5 °C /min and maintained at this temperature for 300 s.

2.6. Starch Film Formulation

Starch films were formulated using the solution casting method according to the methodology described by Nogueira et al. [22] with some modifications. The effects of different concentrations of starch and glycerol on the properties of the formulated films were evaluated based on a central composite experimental design with a total of 13 experimental runs (Section 2.11). Solutions were formed by dispersing starch in distilled water (w/v) for 20 min and then heating the mixture to 90 °C in a magnetic stirrer. Glycerol was added to the solution according to the experimental design and homogenized, and the mixed solution was cooled to room temperature under constant stirring to remove possible air bubbles. A volume of 40 mL of the resulting solutions was transferred into 12 × 12 cm square acrylic plates, and the films were dried for 16 h at 40 °C.

2.7. Starch Film Morphology

The starch film microstructure was examined with the scanning electron microscope (JSM-6510LV, JEOL, Tokyo, Japan). Film samples were placed on double-sided carbon adhesive tape attached to a base and coated with gold (Desk V, Denton Vacuum, Moorestown, NJ, USA, EEUU), and they were then observed by operating the scanning electron microscope at 25 kV.

2.8. Starch Film Color Parameters

The starch films were analyzed using a Colorflex #G C04C 1360 colorimeter (HunterLab, Reston, VA, USA, EE. UU.) to determine the color parameters. The equipment was assembled utilizing a ring and an opaque lid and was calibrated before measurements with the calibration tiles (black, white, and green) provided by the manufacturer. The color parameters are indicated as follows: L * (luminosity), a * (red/green), and b * (yellow/blue). The films were evaluated in triplicate.

2.9. Water Activity

Water activity (a_w) was measured using a water activity analyzer, Hygrolab C1 (Rotronic, Hauppauge, NY, USA, EE. UU.). The film samples were placed in a cell and evaluated in triplicate [23].

2.10. Tensile Strength

Mechanical analysis was performed using a texture analyzer model, TA.XTplus (Stable Micro Systems, Surrey, UK), according to ASTM D882-02 [24]. The film samples were cut into rectangular strips (length: 4 cm; width: 1 cm) and placed between tensile grips (A/MTG, Stable Micro Systems, Surrey, UK), where the initial grip distance and velocity were 40 mm and 0.5 mm, respectively. The tensile strength in g was recorded during film stretching until rupture, and a minimum of three samples were analyzed for each film.

2.11. Statistical Analysis and Design of Experimental Approach

The DOE method was created to investigate how starch and glycerol concentrations influence the tensile property of film strength. Initially, a screening study was performed to evaluate different levels of both variables; once the interest region was determined, a Rotatable Central Composite Design (RCCD) was applied to maximize the tensile strength of the starch film. The levels of each variable were concentrations of starch from 3% to 6% and of glycerol from 1.5% to 3.5%, and a total of 5 replicate center points were used to estimate pure error for the lack-of-fit test (

Table 1. Coded and uncoded variables studied with RSM based on RCCD.

		Coded Variables		Uncoded Variables
Std Order	Run	Factor 1: Starch	Factor 2: Glycerol	Factor 1: Starch (%)	Factor 2: Glycerol (%)
13	1	0	0	4.5	2.5
3	2	−1	1	3	3.5
5	3	−1.41421	0	2.37868	2.5
8	4	0	1.41421	4.5	3.91421
6	5	1.41421	0	6.62132	2.5
7	6	0	−1.41421	4.5	1.08579
1	7	−1	−1	3	1.5
12	8	0	0	4.5	2.5
4	9	1	1	6	3.5
10	10	0	0	4.5	2.5
2	11	1	−1	6	1.5
11	12	0	0	4.5	2.5
9	13	0	0	4.5	2.5

Significant differences were evaluated with analysis of variance (ANOVA), and Tukey–Kramer analysis was conducted as a post hoc multiple comparison test (p < 0.05).

). The relationship between tensile strength and starch and glycerol concentrations was represented by using a linear second-order polynomial model defined as follows:

$$\begin{array}{ll}y = {\beta }_0 + {\displaystyle \sum _{i=1}^k{\beta }_i}{x}_i + {\displaystyle \sum _{i=1}^k{\beta }_{ii}}{x}_i^2 + {\displaystyle \sum }\hfill & {\displaystyle \sum _{i<j}{\beta }_{ij}}{x}_i{x}_j + \epsilon \end{array}$$

(1)

where y, represents the predicted response, in this case, tensile strength; k represents the number of variables; x_i and x_j are the independent variables (factors); β₀ defines the mean response when all the predictors are zero; β_i, β_ii and β_ij, are the linear, quadratic, and interaction coefficients, respectively; and ε represents the error term.

Significant differences were evaluated with analysis of variance (ANOVA), and Tukey–Kramer analysis was conducted as a multiple comparison test. To analyze the obtained results, ${\mathit{R}}^{\mathbf{2}}$ and ${\mathit{R}}_{\mathit{a}\mathit{d}\mathit{j}}^{\mathbf{2}}$ were recorded to check model adequacy. For the experimental matrix, we used Desing Expert software, version 13 (Stat-Ease, Inc., Minneapolis, MN, USA).

3. Results and Discussion

3.1. Starch Extraction and Chemical Composition

It is well known that the most important quality factors in starch isolation are yield and purity, where the latter is determined by the contents of lipids, ash, and protein in starch, as well as its color, which serves as a visual clue. The physicochemical features of starch determine its applications, so understanding its granule composition and its shape, functional, and nutritional qualities is crucial to determining and allocating specific functions in various industries [25]. In this study, the starch obtained from RGs was a bright-white, odorless powder, yielding 21 g of starch per 100 g of RG bean flour. The purity obtained was determined based on the starch chemical composition, where carbohydrates represented 98%, protein 0.87%, ash 0.54%, and fat 0.52%.

Usually, legume starches are obtained from whole commercial grains, for which yields can vary between 30 and 37% according to some reports, for example, 30–31% for chickpeas, 34–37% for lentils [26], 30% for black beans, and 29% for “Pinto Durango” beans [27]. The starch yield obtained in this research study is lower than the values in the studies referenced above. However, the present work used RGs (i.e., all those grains that did not meet the necessary quality criteria, such as size, shape, and color). Therefore, if we compare our result with data from studies using other noncommercial and byproduct sources, such as turmeric rhizome wastes (3%), mangalo bean (6%) [28], Indian arrowroot (7%), and mango ginger (7%) [29], we can see that the obtained yield value is significantly higher, indicating that common bean RGs represent an excellent alternative starch source with high potential.

Despite amylose being a minor component of most starch granules, it significantly affects their physical and functional properties, particularly in pasting and product applications. RG starch had an amylose content of 27.0 ± 1.22%, aligning with findings on starch from other sources and species, including chickpeas [30], black beans, smooth peas, lentils [31], and potatoes and sweet potatoes [32].

3.2. Morphological Characteristics of Starch Granules According to OM and SEM

When characterizing powdered products such as starch, analyzing their morphology is essential because it significantly affects their physicochemical properties. As can be seen in Figure 1A, starch granules from discarded grains, compared with other sources, display interesting morphological characteristics: a round, oval, kidney-like shape and sizes varying from 7 to 34 μm. This is consistent with the findings obtained by Hayat et al. [33], who reported sizes around 10.5 to 42.5 for “Pinto” bean starch granules, showing that there are no differences between the starch from discarded grains obtained in this work and that derived from commercial bean grains. Additionally, Figure 1B shows that under polarized light, the granules display central lines, known as the “Maltese cross,” a well-known phenomenon of birefringence, i.e., the ability of crystalline regions in materials to refract polarized light; this pattern is typically used to identify the presence of starch in many food or non-food systems. Scanning electron microscopy shows (Figure 1C) that, at 1000× magnification, the granules have regular and continuous surfaces without pores, deformities, or markedly angular shapes. When viewed at 3000× magnification (Figure 1D), the surface of the granules is smooth, without the presence of fissures, but it shows a marked equatorial groove; this characteristic has been previously reported by Ovando-Martínez et al. [27] in “Pinto Durango” starch and by Hoover et al. [34] in pigeon pea starch.

3.3. Starch Pasting Analysis

A key starch characteristic that can be evaluated in heat-treated aqueous dispersions by assessing the distinct changes in viscosity as a function of time and temperature is pasting, a physical phenomenon that refers to changes in starch structure before and after gelatinization [35]. Figure 2 shows the effect of different concentrations on the pasting profile of RG starch at 3%, 5%, 10%, and 15% concentrations. It is evident that the dispersions at different percentages led to pastes with distinct pasting properties.

The resulting graphics, known as pasting curves (PCs), track viscosity across various times and temperatures, and their analysis allows us to extract key parameters that are essential to starch applications [36]. For instance, peak viscosity (P_v), which denotes the maximum viscosity attained through the interaction of swelling and polymer leaching, is an indicator of the water absorption capabilities of starch granules. Trough viscosity (T_v) is the minimum viscosity observed after P_v, and final viscosity (F_v) is the viscosity of the paste post-cooling; finally, setback viscosity (SB_v) can be derived from these parameters by subtracting T_v from F_v.

In PCs, the pasting temperature is defined as the point at which viscosity increases as the temperature rises. In this study, the pasting temperature (GT) was between 72 °C and 74 °C. The GT is typically associated with the swelling of starch granules, their partial breakdown, and amylose leaching. Variations in the pasting profiles and gelatinization characteristics of starch from various sources arise due to several factors, including amylose content, the extent of crystallinity, granule size, and chain interactions [37].

Table 2. Gelatinization temperature, peak viscosity, and setback viscosity for starch obtained from rejected grains from the common bean industry.

[Starch] %	GT (°C)	Pv (Pa·s)	SBv (Pa·s)
3	74.44 ± 0.04	0.032 ± 0.00	0.01 ± 0.00
5	74.51 ± 0.11	0.047 ± 0.00	0.01 ± 0.00
10	72.38 ± 0.05	0.340 ± 0.00	0.59 ± 0.01
15	72.55 ± 0.01	1.09 ± 0.02	5.11 ± 0.09

GT = gelatinization temperature; P_v = peak viscosity; SB_v = setback viscosity.

compares the effects of different concentrations of RG starch on some pasting parameters, including P_v and SB_v. Starch gelatinization and retrogradation processes are critical to its functional properties, digestive characteristics, and, notably, its applications across various industries. In this context, it is evident that a reduction of 2 °C in the GT value occurred when starch concentrations were increased. This alteration may be associated with the precision of the measurement system, which exhibits greater accuracy at elevated concentrations (greater than 10%) compared with lower levels (less than 10%) of RG starch.

On the other hand, the P_v values showed a considerable increase at concentrations of 10% and 15%, which can be explained as follows: P_v corresponds to the point at which the count of swollen yet intact starch granules reaches its maximum value. In this context, as the concentration increases, the number of granules also rises; thus, P_v increases. This behavior has also been reported by Mauro et al. [38] in maize, potato, and tapioca starches. The P_v values obtained in this work are in agreement with those reported for different common beans, such as white beans, red spotted beans, and Bambara groundnut [39]. P_v reflects the water-binding capacity of starch granules and is often linked to the quality of the final product. Therefore, it is a parameter with a high degree of importance. To gain a clearer insight into pasting behavior, mathematical modeling was carried out following the methodology proposed by Palabiyik et al. [40] to observe how quickly peak viscosity occurs in the initial region of the PCs. This model is essential to comprehending the kinetics of gelatinization (granule swelling) and determining food processing times. The modeling was executed by adjusting the data associated with the viscosity of the starch sample at 15% maintained at 30 °C for 120 s at a rotational speed of 30 rad/s and that of the sample subjected to a heating process, with the temperature increasing from 30 °C to 90 °C at a rate of 5 °C per minute while maintaining a rotational speed of 30 rad/s. The data were fitted to the following equation:

$$\mathit{V}={\displaystyle \frac{{\mathit{V}}_{\mathit{p}\mathit{e}\mathit{a}\mathit{k}}·{\mathit{t}}^{\mathit{S}}}{{\mathit{R}}^{\mathit{S}}+{\mathit{t}}^{\mathit{S}}}}$$

(2)

In this context, V indicates the viscosity of the starch paste throughout the process, V_peak denotes the peak viscosity, t refers to the processing time, R represents the time at which 50% of peak viscosity is achieved, and S is the starch coefficient.

Figure 3 illustrates the observed data, represented as filled circles, and the modeled data, the continuous blue line. The graph indicates that the model sufficiently correlates with the experimental results, particularly as viscosity rises with the increase in temperature.

The V_peak value derived from the regression closely (

Table 3. Coefficient values from the regression analysis of the pasting curves in the initial region.

Coefficients	Value	Confidence Intervals		${\mathit{R}}^2$	${\mathit{R}}_{\mathit{a}\mathit{d}\mathit{j}}^2$
		Low	High
Vpeak	1.19	1.0749	1.3066	0.9814	0.9804
S	32.7991	24.7918	40.8098
R	774.2395	763.4950	778.9840

) matches the P_v value from the experimental data, demonstrating that the kinetic model can effectively predict P_v during starch pasting. The obtained S value (Table 3) is >1, meaning that water penetration into the starch granule positively affects the infiltration of other incoming water molecules, allowing them to enter the granule more efficiently; therefore, the granule swelling rate increases due to water penetration. The S value result (>1) is consistent with the data reported by Fei et al. [41] in mung bean starch. However, the values obtained in this study are significantly greater than those previously reported. This is likely due to the starch granules from RGs facilitating water molecule entry.

3.4. Visual Aspect, Color, and Water Activity Parameters of the Obtained Films

According to the obtained results, all formulations were adequate, since they produced transparent, homogenous, odorless, flexible films with smooth surfaces. There were no bubbles or insoluble particles, as shown in Figure 4, and the films were easy to handle and did not break when removed from the acyclic plate after drying. The visual aspects of the films were the same for the different starch/glycerol formulations, with one side of the films being glossy and the other one matte.

Table 4. Color and water activity parameters obtained for films under different experimental conditions.

Run	Factor
	Starch	Glycerol	L	a	b	aw
1	4.5	2.5	19.16 cd	−0.2 abcd	−2.21 abc	0.28 cde
2	4.5	2.5	19.11 cd	−0.07 abc	−1.72 a	0.35 ab
3	4.5	3.91	18.53 fg	−0.3 abcd	−1.70 a	0.31 bcd
4	2.37	2.5	16.72 i	−0.4 abcd	−3.00 de	0.4 a
5	4.5	2.5	19.6 b	−0.16 abcd	−2.19 abc	0.27 de
6	4.5	2.5	19.5 bc	−0.31 abcd	−2.70 cd	0.22 e
7	3	3.5	18.13 gh	0.12 a	−2.48 bcd	0.25 e
8	6	1.5	18.68 ef	−0.32 abcd	−2.70 cd	0.35 ab
9	6.62	2.5	20.17 a	−0.003 ab	−1.97 ab	0.39 a
10	4.5	1.08	18.9 de	−0.54 abcd	−2.97 de	0.32 bcd
11	6	3.5	19.84 ab	−0.64 bcd	−2.66 cd	0.33 bc
12	3	1.5	15.93 j	−0.82 cd	−3.59 de	0.27 de
13	4.5	2.5	17.91 h	−0.75 d	−3.07 e	0.33 b

The values are the means of three replicates (n = 3). Means in the same column followed by different letters are statistically different according to the Tukey–Kramer test (p < 0.05).

presents the film’s color values. In this table, L indicates luminosity, a* quantifies the shift from red to green, and b* defines the spectrum from yellow to blue. For all films, the a* values are negative and close to zero, indicating a lack of prominent red tones. Similarly, the b* values are negative, indicating that the bean starch films have a blue tint even though they initially appear transparent. Additionally, films with lower starch content show decreased brightness levels.

Tukey–Kramer comparisons were conducted to evaluate the films’ water activity (a_w) results, revealing that the a_w values ranged from 0.2 to 0.4. Comparable values can be found for starch films based on arrowroot, 0.45 to 0.49 [42], and wheat, 0.47. This suggests a potential use of RG starch films in the food industry, considering their low water activity values, which would prevent microbial growth [43].

3.5. Film Surface Study

Figure 5 shows scanning electron microscopy photographs of the films taken at 100× magnification. It can be seen that the surface of the films appears to lack fissures, cracks, and pores, exhibiting a rough texture with a homogeneous structure; there are also some starch granules resembling empty sacks after amylose and amylopectin leaching [44] and starch retrogradation [45]. Figure 5 also shows cross-sections of the obtained films, which are similarly devoid of fissures, cracks, and pores and additionally show no residual granules. This indicates that with the chosen preparation method, along with the characteristics of starch sourced from RGs from the bean industry, we can successfully obtain uniform film structures that may exhibit good plasticization due to the presence of glycerol. This might be because the starch investigated shows strong compatibility with glycerol at the concentrations examined, demonstrating its good film formation ability.

3.6. Mechanical Properties

Pure starch is a brittle polymer, necessitating plasticization with specific compounds for its use in a broader range of applications. Glycerol, a nontoxic, low-molecular-weight polyol, has been shown to have an adequate response [14], making it one of the most widely used options for plasticizer effects.

Table 5. Texture parameters for different film preparations.

		Uncoded Variables		Response
Std Order	Run	Factor 1: Starch (%)	Factor 2: Glycerol (%)	Tensile Strength (Mpa)
13	1	4.5	2.5	3.65 ± 0.12 f,g
3	2	3	3.5	1.73 ± 0.12 h,i
5	3	2.3	2.5	2.10 ± 0.01 h
8	4	4.5	3.9	0.79 ± 0.01 i
6	5	6.6	2.5	9.82 ± 0.13 c
7	6	4.5	1	15.13 ± 0.41 b
1	7	3	1.5	2.57 ± 0.01 g,h
12	8	4.5	2.5	3.54 ± 0.02 f,g
4	9	6	3.5	6.28 ± 0.31 e
10	10	4.5	2.5	3.94 ± 0.75 f
2	11	6	1.5	23.39 ± 0.91 a
11	12	4.5	2.5	4.55 ± 0.27 f
9	13	4.5	2.5	8.20 ± 0.26 d

The values are the means ± standard deviations (n = 3). Means in the same column followed by different letters are statistically different according to the Tukey–Kramer test (p < 0.05).

shows the tensile strength (MPa) of various film preparations based on the experimental design, aiming to assess how starch and glycerol concentrations impact the mechanical behavior of the films. The highest value was achieved with 6% starch and 1.5 mL of glycerol and the lowest one with 4.5% starch and 3.9 mL of glycerol.

To further study the mechanical behavior of the films, tensile strength was chosen as the response variable in the statistical experimental design. Once the analysis was carried out, the regression of a quadratic model in the uncoded equation was performed as presented below:

$$\mathit{T}\mathit{S}=\mathbf{622.33}+\mathbf{529.33}\mathit{S}-\mathbf{489.60}\mathit{G}+\mathbf{133.63}{\mathit{S}}^{\mathbf{2}}+\mathbf{94.52}{\mathit{G}}^{\mathbf{2}}-\mathbf{412.42}\mathit{S}\mathit{G}$$

(3)

where TS represents the tensile strength, S is the starch concentration, and G is the glycerol in the film formulations. According to the analysis of variance, the model was significant, with a p-value < 0.05, and the adequacy of the model was checked with the lack-of-fit test, with no significant result; additionally, ${\mathit{R}}^{\mathbf{2}}$ = 0.98 and ${\mathit{R}}_{\mathit{a}\mathit{d}\mathit{j}}^{\mathbf{2}}$ = 0.9763 values were obtained, indicating that the regression model explains approximately 98% of the response variable. Therefore, the model is sufficiently precise to predict tensile strength. A surface model graph is presented in Figure 6 to analyze the response variable behavior. Overall, the results show that tensile strength tends to increase with higher starch concentrations. Concerning glycerol, as demonstrated in Figure 6, there is a slight rise in tensile strength as its amount increases from 1.5% to 3.5%, while the starch concentration remains at 3%. This behavior is related to what was observed in the pasting profile, particularly to the setback viscosity, since as shown in Table 2, this parameter increases as the starch concentration increases. It is known that setback viscosity is closely related to the retrogradation phenomenon, which refers to the behavior of starch when cooled after gelatinization-driven film formation. This is counteracted by adding glycerol, perhaps due to increased mobility among macromolecules, which reduces the mechanical strength of composites [46]. Glycerol is a low-molecular-weight polyol featuring three hydroxyl groups per molecule, which enable it to readily form hydrogen bonds with the hydroxyl groups in starch macromolecules [47]. This response has been previously reported by Wang et al. [48] in potato starch film reinforced with straw fiber. Taken together, the above findings confirm the close relationship between the pasting profile and the mechanical properties of the films.

On the other hand, it can be seen from the same graph that when glycerol and starch concentrations reach their highest levels (starch = 6%; glycerol = 3.5%), the tensile strength decreases. This a typical pattern that occurs because the glycerol molecules interfere with the interactions between the starch chains, increasing flexibility but reducing rigidity and mechanical strength [49]. Moreover, increasing the starch concentration can enhance rigidity and strength up to a point; however, if it becomes too high, it can hinder interactions with glycerol, resulting in more fragile or less uniform films, which can also lower tensile strength [50]. Additionally, a high starch concentration can increase thickness and opacity, but it does not necessarily improve mechanical strength if the plasticizer amount is insufficient to keep the material flexible [51].

4. Conclusions

The results indicate that starch obtained from byproducts of the bean industry possesses distinct characteristics, positioning it as a potential option for developing innovative packaging; using it in combination with glycerol as a plasticizer would enable the production of transparent, uniform, odorless, and flexible films featuring smooth surfaces. The film structure obtained showed no bubbles or insoluble particles, and the SEM studies showed surfaces free of fissures, cracks, pores, and residual granules. This suggests that with the chosen preparation method, combined with the properties of starch obtained from discarded grains from the bean industry, we can effectively create consistent film structures that could demonstrate good plasticization due to the presence of glycerol. Additional experiments are required to comprehensively assess these films and determine their efficacy as model food packaging.