Effect of Sweet Potato Starch on Rheological Properties and Emulsion Stability of Salad Dressings

Abstract

Due to its gelling and thickening properties, sweet potato starch (Ipomoea batatas L.) could be a promising ingredient to improve characteristics such as the viscosity and consistency of foods like dressings. The objective of this study was to use sweet potato starch by adding it to salad dressing-type emulsion formulations. Sweet potato starch was characterized (microscopic appearance, granule size, and thermal properties). Four formulations (F1–F4) were developed incorporating different amounts of sweet potato starch (2 and 4%), and were characterized by particle size, emulsion stability, rheology, and sensory analysis. The starch granules were oval shaped, with a size range of 10–33 μm, and a temperature and enthalpy gelatinization (ΔH) of 69.08 °C and 10.72 J/g, respectively. The formulations were evaluated for 30 days, the particle size had a range of 2.18–13.88 μm, the emulsion stability was 98.89–100%, all formulations presented a creaming index at 0%, and the coalescence rate obtained values between −2.33 × 10⁻⁸ and 7 × 10⁻⁸ Kc (s⁻¹) showing a significant difference. The consistency coefficient (K) was obtained, 2.477–35.207 Pa·sⁿ, and there was no significant difference between F1 and F2 with respect to a commercial dressing. In the sensory analysis, F2 presented greater acceptance. The values obtained suggest that sweet potato starch could be used in this type of food, showing similarities to the commercial brand.

1. Introduction

Sweet potato (Ipomoea batatas L.) is a tuberous root of great economic importance due to its caloric source and high nutritional value, in addition to its ease of cultivation, with production in Asian countries, Central and South America, and tropical and subtropical regions around the world [1,2,3]. Worldwide, the crop had a production of 71 million tons in 2022, with China being the leading country in the production of the crop [4]. In Mexico, in 2023, annual sweet potato production was almost 77 thousand tons, with the state of Michoacán having the highest production [5]. The sweet potato is an important source of nutrients and functional ingredients such as proteins, carbohydrates, dietary fibers, phenols, and carotenoids [2]. The most abundant carbohydrate is starch, which represents about 80% of its dry matter; this component is a polysaccharide composed of amylopectin and amylose, the ratio of which depends on the botanical source, the maturity of the tuber, and the cultivar. Starch is an ingredient that has been widely used as a raw material for the manufacture of chemical additives, foods, pharmaceuticals, and bio-foods [1,2].

Sweet potato pulp can have a variety of colors such as white, beige, orange, and purple, and the starch granules can present different morphologies and plastering properties, which allows their application to contribute to the textural properties of various products as an ingredient in the food industry such as sauces, pasta, soup, snacks, and bread [3].

Currently, consumers and the food industry have increased their interest in seeking low-fat foods, as fat is one of the main nutrients in foods that provides the highest caloric density. However, product development has been challenging because fat plays a diverse role in some primary sensory attributes, such as texture (creamy, rich, and smooth), appearance (creamy and milky), and flavor (satiating and desirable). Consequently, it is essential to develop new formulations that can contribute to reducing the fat content of food products while preserving desirable sensory attributes. Starches are widely used as ingredients in food products to improve overall acceptability, including texture. Their main functions are to confer gelling, adhesion, thickening, stabilizing, moisture-retaining, film-forming, texturizing, and anti-caking properties; in gluten-free products, starch is used to improve some of these properties [6,7]. Conventional starches such as those obtained from corn, rice, potato, wheat, and cassava have been the focus of most research because they are the most common sources for commercial starch extraction worldwide; however, studies on starch from less conventional sources are required due to a growing demand for starch-rich foods with a low glycemic index (GI), which allows for gradual metabolism and lower blood glucose [8]. The use of starches obtained from alternatives such as cassava, sweet potato, and yam has been studied for the production of salad dressings due to their functional and chemical properties; they are considered an alternative due to the high viscosity and whiteness they can provide to the food product [9]. Other research has used sweet potato flour as a replacement ingredient for wheat flour at various levels in baked goods such as donuts, bread, and cookies. It is sensorily acceptable and provides various desirable physicochemical and functional properties, addressing the challenge of a population that restricts the consumption of cereal products due to celiac disease [10]. Donaldben et al. [11] evaluated the functional properties of four starches obtained from cream/orange-fleshed sweet potato and white/water yam as potential raw material substitutes for the food and pharmaceutical industries; they concluded that, due to the high amylose content of these starches, they could contribute to textural attributes, especially in food products that require a thick paste consistency, elasticity, and high gel strength, making them alternative sources of starch due to their particular characteristics and use in various products. Although starch is a natural polymer with emulsifying properties, the stabilization of food O/W emulsions by gelatinized starches is uncommon [12]. In the food industry, mayonnaise is one of the products that seeks innovation among its ingredients. Low-fat products are occasionally associated with deficiencies in flavor, appearance, texture, stability, and palatability. It is necessary to reformulate the fat component with functional molecules to obtain a product with sensory attributes similar to those of the original product. The replacement of oil with sweet potato powder in mayonnaise has been evaluated to reduce fat levels in this food product, showing improvements in rheological attributes, sensory acceptability, and nutritional content, in addition to contributing to shelf life stability due to the antioxidant activity presented, highlighting a possible alternative for food emulsions [13,14]. Due to its ability to improve texture and stability, gelled sweet potato starch is used as a functional additive acting as a thickening and gelling agent in various food applications [15]. This research aimed to study the effect of using sweet potato starch as a thickening and stabilizing agent and as a substitute for xanthan gum, evaluating the emulsion stability, particle size, rheology, and sensory acceptability of the dressings through the development of salad dressing formulas.

2. Materials and Methods

2.1. Extraction of Sweet Potato Starch

For the extraction of the starch, sweet potato was washed, hulled, and cut into cubes, then mixed with three volumes of distilled water and poured into a blender (Osterizer, Acuña, Coahuila, Mexico^®) for 5 min at maximum speed. Then, the suspension was centrifuged at 1500 rpm/5 min and then decanted, with the remaining residue being re-liquefied with two volumes of water and then left to stand for 60 min and decanted. The sediment was liquefied again for 1 min at maximum speed and then left to stand for 60 min. Subsequently, the sediment was filtered through a sieve (Manitox, Monterrey, Mexico^®) with a 105 μm mesh opening. The remaining solids were removed and the filtrate was washed until the wash water was translucent, indicating that the starch had been extracted. Subsequently, the fraction in which the starch was contained was dried in an oven (Riossa Digital, Monterrey, Nuevo León, Mexico^®) at 40 °C for 24 h and then stored in glass containers [16].

2.2. Microscopic Appearance

The morphology of the sweet potato starch granules and their size were determined following the methodology presented by Rojas et al. [17]. For this, 1% sweet potato starch suspensions (w/v) were prepared and, with the help of a Pasteur pipette, a drop of starch solution was taken onto a slide. A drop of iodine/potassium iodide solution (0.1%) was added, after which the cover object was placed and observed directly under the microscope (LEICA ICC50) via 40× magnification with the help of Application Suite LAS EZ, version 2.0.0 (LEICA microsystems, Wetzlar, Germany) and the mean particle size diameter of granules that appeared on a field was determined. Micrographs allowed the shape and size of the granules to be determined using the Image J. software 1.54g [18].

2.3. Chemical Composition

The proximal composition was determined according to the official methods described by the AOAC [19], comprising the following analyses: moisture (method 925.09), crude protein (method 954.01), crude fat (method 920.39), crude fiber (method 926.09), ash (method 923.03), and total carbohydrates such as nitrogen-free extract.

2.4. Functional Properties

Thermal Properties

Temperatures and enthalpy of gelatinization were determined by differential scanning calorimetry (DSC), according to the method of Hernández et al. [20]. This was performed on a TA Instruments machine (DSC Q-2000, Crawley, UK) where 2 mg of sample was placed in the aluminum cells for DSC, and 10 μL of distilled water was added. Following this, the cells were sealed and kept in equilibrium for 1 h. They were then swept with a heating speed of 10 °C/min, from 30 to 120 °C. An empty cell was used as a reference for all measurements. As a result, thermograms were obtained in which a plot of heat flux (Y-axis) against temperature (X-axis) was observed. These were analyzed with Universal Analysis software (2000 version 4.7A) to obtain the % peak gelatinization temperature (Tp), initial temperature (Ti), final temperature (Tf), and the gelatinization enthalpy (ΔH).

2.5. Sweet Potato Starch Dressings (Formulations)

The dressings (F1, F2, F3, and F4) were made, varying the percentage of boiling water (36.4–38.7%), xanthan gum (0–0.3%) Sigma–Aldrich (Sigma Chemical Co., St. Louis, MO, USA), and sweet potato starch (2–4%) (

Table 1. Formulations of dressings.

Ingredients		Formulations (%)
		F1	F2	F3	F4
	Water	38.4	36.4	36.7	38.7
	Egg yolk	4	4	4	4
Emulsifiers	Sweet potato starch	2	4	4	2
	Xanthan gum	0.3	0.3	0	0
	Oil	45	45	45	45
	Vinegar	1.4	1.4	1.4	1.4
	Salt	1.4	1.4	1.4	1.4
	Sugar	4.2	4.2	4.2	4.2

). Briefly, the emulsifiers (egg yolk, sweet potato starch, and xanthan gum) were hydrated with the cooking water and constantly stirred (500 rpm, IKA T50 Digital Ultra Turrax, company IKA, Staufen im Breisgau, Germany) for half an hour in a beaker. After gel formation (69 °C), the other ingredients (sugar, salt, and vinegar) were added. Finally, the oil was added slowly in a continuous manner, allowing it to be incorporated into the mix. The mix was stirred for 3 more min. at 3000 rpm and then 2 more min. at 500 rpm, while the temperature was 5 °C. Each formulation containing xanthan gum (XG) and sweet potato starch in different proportions was evaluated for coalescence rate, creaming index, and rheological parameters [consistency coefficient (k) and flow behavior indices (n)]. The formulations were stored at a refrigeration temperature of 4 °C in containers with lids.

2.6. Characterization of Salad Dressings

2.6.1. Microscopic Emulsion Photographs

In a glass tube, 1 g of each formulation was taken, then 9 g of water was added, and stirred vigorously, until a homogeneous solution was formed. A drop was then taken with a Pasteur pipette and placed in a slide holder and, on top of it, an object cover was placed. It was observed at 100× magnification for 30 days using the LEICA ICC50 microscope with the help of the Application Suite LAS EZ program, Version 2.0 (Leica Microsystems, Wetzlar, Germany).

2.6.2. Creaming Index

Creaming results are reported as the Creaming Index (CI) and measured in the following way: in a graduated plastic conical tube (15 × 125 mm), 10 mL from each emulsion was added and stored at room temperature during the time of the analysis, according to Charoen et al. [21]. At the end of the time of storage, the calculation was carried out with the following Equation (1):

$$\mathrm{C}\mathrm{I}={\displaystyle \frac{\mathrm{h}\mathrm{e}\mathrm{i}\mathrm{g}\mathrm{h}\mathrm{t}\mathrm{o}\mathrm{f}\mathrm{i}\mathrm{n}\mathrm{t}\mathrm{e}\mathrm{r}\mathrm{f}\mathrm{a}\mathrm{c}\mathrm{e}}{\mathrm{h}\mathrm{e}\mathrm{i}\mathrm{g}\mathrm{h}\mathrm{t}\mathrm{o}\mathrm{f}\mathrm{t}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l}\mathrm{e}\mathrm{m}\mathrm{u}\mathrm{l}\mathrm{s}\mathrm{i}\mathrm{o}\mathrm{n}}}\times 100$$

(1)

The emulsion with greater stability is the one closest to zero.

2.6.3. Emulsion Stability

To measure physical stability, the methodology of Karshenas [22] was utilized, whereby 15 g (F₀) of the sample was taken in the centrifuge tubes with a specific weight, then, the HERME Labnet Z326 centrifuge was used to centrifuge the tubes for 30 min at a speed of 500 rpm. Finally, the oil layer was removed and the weight of the sediment (F₁) was measured. This test was performed in triplicate, and the stability of the emulsion was calculated according to the percentage using Equation (2).

$$\mathrm{E}\mathrm{m}\mathrm{u}\mathrm{l}\mathrm{s}\mathrm{i}\mathrm{o}\mathrm{n}\mathrm{s}\mathrm{t}\mathrm{a}\mathrm{b}\mathrm{i}\mathrm{l}\mathrm{i}\mathrm{t}\mathrm{y}=\left({\displaystyle \frac{{\mathrm{F}}_1}{{\mathrm{F}}_0}}\right)\times 100$$

(2)

2.6.4. Rheology

The viscosity of the emulsions was measured at the beginning and end of the monitoring month. A ReolabQC (Anton Paar, Graz, Australia) rotational rheometer with the CC-27 concentric cylinder geometry was used, with a cut off rate of 1 to 100 s⁻¹. The flow curves were obtained, as well as the consistency coefficient (K) and flow index (n). Flow parameters were obtained using the power law equation (Ostwald–de Waele) [23].

$$\sigma =K{\left(\mathsf{\gamma }\right)}^n$$

(3)

where σ is shear rate (Pa·s), K is consistency coefficient Pa·sⁿ, γ is shear rate (1/s) is, and n is flow rate.

2.6.5. Particle Size

The particle size of the D_(3,2) dressings was measured using the Mastersizer 3000 (Malvern Instruments, Worcestershire, UK) using the hydro 2000S unit, monitoring the particle size of the emulsions for one month, and then calculating the coalescence rate of the emulsion globules. From the variation in the particle size of the emulsion, the coalescence rate with respect to time (30 days) was calculated using the following equation:

$${{\displaystyle \frac{N_t}{N_0}}=\left[{\displaystyle \frac{{\left(d_{30}\right)}_{t=0}}{{\left(d_{30}\right)}_{t=1}}}\right]}^3$$

(4)

where N_t is the number of globules with respect to time (t) and N₀ is the number of globules at time zero. When plotting ln (N_t/N₀) versus time in seconds (s) for the emulsion, it gives a straight line, so the slope is Kc. The smaller the number obtained in the coalescence rate, the more stable the emulsion [24].

2.6.6. Sensory Analysis

Sensory evaluation was performed by affective testing of the study formulations with a commercial brand dressing. It was evaluated by a hedonic test based on the general perception of the following attributes: texture, flavor, odor, oiliness, and acidity. The formulations were evaluated by 60 untrained panelists, in the age range of 19 to 22 years, with both sexes in the same proportion. The panelists evaluated each attribute with a code generated on a structured scale from 1 to 10, where 1 = the lowest score and 10 = the highest score. The samples were presented in identical containers, using random 3-digit codes in a balanced presentation. Crackers were used as a vehicle to taste the salad dressings.

2.7. Statistical Analysis

All measurements were performed in triplicate. To determine statistically significant differences between the values, a one-way analysis of variance (ANOVA) and Tukey’s test were used (p < 0.05). Statistical analysis of the data was performed using IBM SPSS Statistics Version 25 software for Windows.

3. Results and Discussion

3.1. Microscopic Appearance of Sweet Potato Starch Granules

As shown in Figure 1, the sweet potato starch granules were oval in shape. The size of the granules varied in the range 10–33 μm as shown in

Table 2. Characteristics of sweet potato starch.

Yield (%)	3.63 ± 0.31
Granules	Range of size (μm)	10–33
	Shape	Oval

Values are expressed as mean of three repetitions in triplicate ± standard deviation for yield.

. These values found in the starch are within the typical sweet potato range of 10–25 μm [8]. According to the reports of De la Rosa [25] and Moorthy [26], the size of the granule has an effect on the functional properties, such as the degree of gelatinization, swelling factor, and solubility, and on the polymerization of amylopectin chains. Smaller particles have an increased surface area relative to volume; they facilitate more efficient water absorption and heat transfer which leads to faster and more effective gelatinization [15].

Principal starch properties (gelatinization, thermal, and rheological behavior) vary depending on the starch’s genetic background [27]. This allows us to know the variables that influence the cooking processes of starch and its products.

3.2. Chemical Composition

From the research reviewed, it was found that the physicochemical characteristics of sweet potato starch vary depending on the species analyzed, culture development, and synthesis processes [28]. The results are shown in

Table 3. Chemical composition of sweet potato starch (%).

Parameter
Moisture (%)	0.07 ± 0.04
Fat (%)	0.10 ± 0.04
Ash (%)	1.18 ± 0.42
Fiber (%)	ND
Protein (%)	ND
Nitrogen-Free Extract (%)	98.64 ± 0.26

Values are expressed as the mean of three repetitions in triplicate ± standard deviation. ND = not detected.

. Firstly, the percentage of protein is lower than that reported by Guízar et al. [29], which presents a value of 0.67% for Dioscorea remotiflora, and lower than that reported by Hernández et al. [20] of 0.22% for sweet potato. The ash value is higher than that obtained by Guízar et al. [29] of 0.73% present in Dioscorea sparsiflora. Finally, it can be seen that the carbohydrate content is similar to that reported by Hernández et al. [20] of 98.93% for sweet potato. The percentage of fat of sweet potato starch was 0.10%, which is good, because the presence of lipids can affect the hydration and swelling of starch granules, potentially inhibiting gelatinization if water availability is compromised [15].

3.3. Thermal Properties

Table 4. Thermal properties of sweet potato starch.

Thermal Properties
Initial temperature (Ti)	63.10 ± 0.03
Gelatinization temperature (Tp)	69.08 °C ± 0.02
Final temperature (Tf)	76 ± 0.05
Gelatinization enthalpy (ΔH)	10.72 J/g ± 0.12

Values are expressed as mean of three repetitions in triplicate ± standard deviation.

presents the results for gelatinization and enthalpy. The gelatinization temperature is similar to the value of 70.7 °C that was reported by Osundahunsi et al. [30]. The gelatinization temperature of sweet potato starch is in the range of 60–85 °C and moisture affects the gelatinization of sweet potato starch [15].

The enthalpy of gelatinization (ΔH) agrees with what was reported by Moorthy [26] of 10.18 J/g and Osundahunsi et al. [30] of 10.5 J/g. Lower enthalpy values are associated with higher amylose levels, i.e., Czuchajowska et al. [31]. In a previous study, the amylose content in sweet potato starch was quantified at 27.14% [32], which was similar to that found by Tecson [33], who reported a range from 12.90% to 29.70%. Higher amylose content raises the onset temperature of gelatinization from 73.5 °C in wild-type starch to 76.2 °C in waxy varieties [15].

Starch has amorphous regions which are rich in amylose, while the crystalline region is an organized region where amylopectin is present in greater quantity in its structure. Sweet potato starch has a similar amylose content to that of corn starch, so it can be an alternative for its application in food [32].

3.4. Characterization of Salad Dressings

3.4.1. Microscopic Emulsion Photographs

Micrographs obtained by optical microscope of the fats in salad dressings are shown in Figure 2, with the oil droplet sizes of all the samples being in the range of 2.18–13.88 μm. The formulations F3 and F4 since day 1 showed the phenomenon of flocculation. Meanwhile, formulations F1 and F2 remained stable, with percentages of 2 and 4% starch, respectively. Formulation F4 presents the phenomenon of coalescence, which occurs when the droplets of the dispersed phase come into contact and merge to form larger droplets. As two droplets approach, coalescence causes the thin liquid layer separating individual droplets to break [34].

3.4.2. Creaming Index

Creaming can be prevented by employing a high-viscosity continuous phase [34]. The creaming index is related to the stability of the emulsion; the higher it is, the more unstable it is [35]. The results of the creaming index are presented in

Table 5. Stability of dressings.

Analysis	Unit	Commercial Dressing	Formulation
			F1	F2	F3	F4
Rheology	K (Pa·sn)	35.207 ± 0.55 a	28.925 ± 0.44 a	33.187 ± 0.62 a	8.785 ± 0.41 b	2.477 ± 0.82 b
	n	0.283 ± 0.22 a	0.323 ± 0.52 a	0.230 ± 0.35 a	0.456 ± 0.42 a	0.552 ± 0.32 a
Coalescence rate	Kc (s−1)	2.87 × 10−8 b	−2.33 × 10−8 a	−3.67 × 10−8 c	5 × 10−8 d	7 × 10−8 e
Emulsion stability	%	NA	100 a	100 a	98.89 c	99.30 b
Creaming index	%	NA	0	0	0	0

^a–e means there is significant difference (p < 0.05). Values are expressed as mean of three repetitions in triplicate ± standard deviation; K: Consistency coefficient; n: flow rate. NA: not analyzed.

. A value of 0% was found for formulations F1, F2, F3, and F4. The four formulations remained stable during the storage time, which was one month. It can be deduced that it is possible to make this type of formulation by replacing xanthan gum with sweet potato starch. In previous research by Muñiz et al. [32], the emulsifying capacity of sweet potato starch was evaluated, and a value of 35.56% was obtained, which is related to the creaming index. It is mentioned that starch has amorphous regions which are rich in amylose, and this has a characteristic disordered structure, while the crystalline region is an organized region where amylopectin dominates in its structure. In the gelatinization process, water is absorbed in the amorphous intermicellar zones that are less organized and more accessible. Therefore, a starch with more amorphous regions (amylose) would have a lower gelatinization temperature and in the same way, having less amylose will reduce the syneresis phenomenon.

3.4.3. Emulsion Stability

A one-month study was conducted to evaluate the stability of the dressings, whereby F1 and F2 did not show a significant difference, being 100%, and F3 and F4 presented values of 98.89% and 99.3%, respectively. These results are presented in Table 5. This may be due to the gelatinized starch acting as a thickener, which is a positive influence on the physical stability of the emulsion, contributing to the formation of a thicker protective layer at the interface [36]. This type of emulsion can be highly stable when exposed to environmental stress, such as pH, salt, heating, cooling, freezing, and dehydration.

According to Stoke’s law, the rate of settling or creaming is inversely proportional to the viscosity of the aqueous phase [37], which means that the higher viscosity of the continuous phase and smaller particle size are the main reasons leading to the reduction in creaming rate. Therefore, the high stability of the samples was attributed to their particle size and the stability formed by gelatinized gum and starch. Similar results were observed by Worrasinchai et al. [38].

3.4.4. Rheology

The rheological properties of emulsions can be altered by adding polymers or hydrocolloids, such as thickeners and texture modifiers (e.g., pectin, modified cellulose, gum arabic, corn fiber gum, modified starch, polysaccharide–protein complexes, etc.) [34].

Flow parameters were obtained using the power law model. The power law flow indices (Table 5) confirmed high shear thinning and non-Newtonian behavior for all dressing samples over the entire shear rate range used. Upon analyzing the results, it was observed that there is no significant difference between the commercial dressing and F1 and F2 in the consistency coefficient; this related to the interaction of xanthan gum–starch.

In the work of Rahmati et al. [39], it was found that, among several statistical terms, the linear effect of xanthan gum was the most effective term compared with the other hydrocolloids. The relationship between the power law consistency coefficient and the hydrocolloids used shows that the increase in xanthan gum concentration from 0 to 0.3% had a considerably positive influence on the consistency coefficient. In other works, the high capacity of xanthan gum for improving viscosity is reported by Dolz et al., Sanchez et al., and Thomareis et al. [40,41,42]. The high molecular weight of xanthan gum and the formation of aggregates through hydrogen bonding are the reasons why its solutions exhibit high viscosity. It is inferred that the power law consistency coefficient depends on the interactions between xanthan gum and starch. These results are in agreement with those found through previous research, in which a synergistic effect between xanthan gum and starch was observed [36,40].

In the research conducted by Weber et al. [43], no covalent bonds were found between starch and xanthan gum; therefore, they reported that the only interactions that occurred between them were probably formations of hydrogen bonds. With regard to the flow index, the addition of xanthan gum also influenced this result, but it positively benefited from the interaction with starch, since, as seen in Table 5, the flow index of F1, with a starch content of 2%, obtained a lower n than that of F2 with 4% starch content, indicating that the modified starch helps to improve the flow index. This is verified by the results of F3, which had a lower n than F4 since it has a lower percentage of gelatinized starch. This may be due to the interaction between xanthan gum and starch, since xanthan gum makes the continuous phase more viscous, which decreases the value of the flow index, so the higher the proportion of xanthan gum, the lower the value (n) will be. The highest flow behavior indices indicate a weak pseudoplastic behavior and were related to samples containing only starch as a fat substitute, i.e., Rahmati et al. [39]. This coincides with the results in F3 and F4, in which no xanthan gum was used, only gelatinized starch.

3.4.5. Particle Size

The particle size distribution of the salad dressing emulsion was analyzed; the droplet size of the emulsions varied from 2.184 to 3.461 μm. The values of the coalescence rates of all emulsions were in the very stable range of (1 × 10⁻⁸–1 × 10⁻⁹ s⁻¹) according to the Sherman et al. [44] data shown in Table 5. The good results of the interaction between xanthan gum and starch were due to the particle size. In Rahmati et al. [39], it can be observed that the increase in the amount of gelatinized gum and starch changed the distribution pattern of droplet size. With respect to its curves, it was reported that the use of higher contents of different gums, regardless of their type, shifted the curves to the left (smaller particle size). From this result, it can be inferred that the higher contents of gums prevented the flocculation and coalescence of the droplets after the preparation of the emulsion, which resulted in a smaller particle size for the formulations developed. Surface plots for particle size as a dependent variable were also produced, and it was shown that the increase in starch and xanthan gum led to a decrease in droplet size.

This is mainly seen in xanthan gum since, according to the sum of squares test undertaken, the increase in the concentration of guar gum did not statistically change the size of the oil droplets. This can be interpreted as polysaccharides, through water absorption, resulting in a more viscous continuous phase that limits the movements of oil droplets. Therefore, collisions between droplets would decrease, resulting in a smaller size for the scattered droplets. Previous researchers reported similar results for smaller particle sizes for samples with higher viscosity [36,45]. Although, a higher viscosity (Figure 3) was observed for samples containing higher concentrations of xanthan gum (F1 and F2). This is probably due to its negatively charged structure. In general, if enough loaded biopolymer is added to an emulsion containing oppositely charged droplets, it can completely saturate the droplet surfaces and form a stable system since the droplets are completely coated with polymer. Theoretically, the isoelectric point for egg yolk proteins is 5.3 [39]. When there is a pH value lower than the isoelectric pH, proteins have a positive charge. Due to the presence of egg proteins and lipoproteins, the xanthan molecule, as an anionic gum, can be adsorbed at the interface through attractive electrostatic forces. In these multilayer emulsions, the inner part of the layer is built by proteins and the outer section is built by polysaccharides. This result was in agreement with that observed by Bouyer et al. [46], who reported that in the dispersion of beta-lactoglobulin and gum arabic, beta-lactoglobulin was adsorbed at the interface and gum arabic was electrostatically bound to it, leading to the formation of a stable bilayer emulsion.

3.4.6. Sensory Analysis

According to the flow curves shown in Figure 3, it can be seen how formulations F1 and F2 have a behavior very similar to the commercial dressing, while F3 and F4 did not have an expected behavior of a commercial dressing, so it was decided to discard these formulations for sensory evaluation and only leave formulations F1 and F2.

First, a descriptive test was carried out, to determine if there were differences between the samples and the reference (commercial dressing). According to the scale presented to the judges, it was considered that the value of the mean closest to the reference value was the value at which the judge (untrained) considered it to be equal to the reference sample.

Starting from the fact that the value of 5 is the value given to the reference sample, a completely random design was made with a significance of p < 0.05. According to the results presented in

Table 6. Sensory analysis.

Attribute	Commercial Dressing	Formulation
		F1	F2
Texture	7.10 ± 1.80 a	7.03 ± 2.07 a	7.08 ± 1.64 a
Flavor	6.90 ± 2.45 a	6.43 ± 2.53 a	6.71 ± 2.54 a
Smell	6.80 ± 2.28 a	6.55 ± 2.21 a	6.61 ± 2.25 a
Oiliness	7.30 ± 1.9 a	6.83 ± 2.14 a	7.00 ± 1.81 a
Acidity	6.30 ± 2.15 a	6.23 ± 2.04 a	6.25 ± 2.28 a

^a = Equal suprascripts means there is no significant difference (p < 0.05). Values are expressed as mean of three repetitions in triplicate ± standard deviation.

, the samples do not present a significant difference between the sample and the commercial dressing, and since the lower values of the mean is in the color attribute, this means that this is the attribute that is the least close to the commercial dressing, while consistency was the one that showed the closest to the commercial dressing without showing a difference in acceptance between the treatments. According to these results, the formulations F1 and F2 showed that they do not have a significant difference in comparison with the sample of commercial dressing. Therefore, they were used to conduct the pleasure level test.

A palatability test was applied to the dressings developed to measure their acceptance by consumers. The attributes evaluated were taste, smell, oiliness, texture, and acidity. Values with higher averages are more accepted by the population. The judges did not observe significant differences between the dressings developed with respect to any of the attributes. As shown in Table 6, the attribute with the best acceptance was texture, with there being no significant differences between the formulations.

4. Conclusions

The use of sweet potato starch (Ipomoea batatas L.) facilitated the development of physically stable dressings. The addition of xanthan gum to F1 and F2 provided better results in the flow curves, flow index, and consistency coefficient, in addition to obtaining results similar to the commercial brand. The concentration of 2% starch in the dressing (F1) obtained rheological and physical stability results similar to the commercial brand. The formulations developed with the addition of 2 and 4% sweet potato starch (F1 and F2) in the sensory evaluation did not present significant differences to the commercial brand. This research provides results for the use of sweet potato starch as a new additive option for emulsion-type foods such as dressings.