Formulation Strategies for Mayonnaise-Type Sauces: The Role of Hydrocolloid Combinations ^†

Abstract

The aim of this study was to investigate the substitution of egg yolk in mayonnaise-type sauces with alternative protein components and to optimize the hydrocolloid composition for improved stability and rheological properties. Mustard powder (1%), soybean flour (1%), casein (2%), and cream powder (1%) blends were employed as emulsifiers. The influence of the ratio of potato starch, carboxymethylcellulose (CMC), pectin, and xanthan gum (0–1% each) on the properties of low-fat mayonnaise formulations with 30% oil content was examined. Sedimentation and thermal stability tests revealed high resistance of all samples (98–99%) after 24 h of storage. Optical microscopy confirmed a homogeneous structure with individual dispersed particles of 100–150 μm corresponding to plant protein inclusions. The particle size distribution D [3,4] exhibited a bimodal profile with peaks at 0.1–1 μm and 2–8 μm, indicating efficient homogenization. Storage experiments demonstrated an increase in particle size by 1.4–1.6 times and a decrease in viscosity, likely due to flocculation and aggregation of polysaccharide clusters into larger agglomerates. Among the tested formulations, the sample containing 0.3% CMC, 0.3% xanthan gum, and 0.4% pectin showed the most favorable physicochemical and sensory properties, highlighting the synergistic effect of hydrocolloid blends in stabilizing reduced-fat mayonnaise-type emulsions.

1. Introduction

In the development of low-calorie emulsified sauces, achieving a balance between microbiological safety, nutritional profile, and specific functional rheology is critical for consumer acceptance. To ensure safety and structural integrity, the system pH must be maintained at or below the isoelectric points of the protein stabilizers, while significantly reducing the oil content to meet low-fat dietary requirements [1,2,3]. Currently, egg-free mayonnaises are gaining significant popularity. The substitution of egg-based ingredients with plant-derived components is a highly promising trend, as such sauces offer enhanced microbiological safety, cost-effectiveness, and improved nutritional profiles. Furthermore, these formulations are particularly appealing to health-conscious consumers and those adhering to a vegetarian or vegan diet [4,5,6,7]. Functionally, emulsion sauces must exhibit sophisticated thixotropic behavior; they require sufficient structural strength and yield stress to remain stable and “hold” their shape when spread onto a surface like bread, yet must demonstrate a controlled reduction in viscosity under shear to flow effectively and provide uniform coating of ingredients when used as a salad dressing. This dual-functionality is achieved through the synergy of hydrocolloid blends [8]. In the technology of emulsified sauces such as mayonnaise, modified starches and xanthan gum are traditionally regarded as the principal gel-forming and structure-building agents, while guar gum is also frequently applied as a thickener to enhance viscosity and water retention [9]. These hydrocolloids are widely used due to their proven ability to stabilize oil-in-water emulsions and to compensate for fat reduction in low-calorie formulations. However, growing interest in diversifying texture-forming systems and improving clean-label perception has motivated the investigation of alternative or complementary polysaccharides. In this context, pectin and carboxymethyl cellulose (CMC) represent promising but comparatively less explored components in mayonnaise technology. Pectin, a plant-derived polysaccharide with emulsifying and gel-forming capacity under acidic conditions, can contribute to network formation and moisture retention in reduced-fat systems [10], while CMC provides viscosity enhancement and emulsion stabilization through its high water-binding capacity and shear-thinning behavior [11]. When combined with established thickeners such as xanthan gum and potato starch, these polymers may exhibit synergistic interactions that improve rheological properties, emulsion stability, and sensory texture. Previous studies on low-fat mayonnaise and salad dressings demonstrate that mixed hydrocolloid systems outperform single thickeners in terms of structural stability and mouthfeel, supporting the rationale for exploring pectin–CMC–xanthan–starch combinations as a functional alternative to conventional guar- or starch-dominated formulations [12,13,14].

2. Materials and Methods

2.1. Materials

The formulation of the low-fat mayonnaise-type sauces consisted of food-grade ingredients obtained from local suppliers and included the following components (wt. %): sunflower oil (Oleina, Bunge, Dnipro, Ukraine)—30.00, casein (Kovelmoloko, Kovel, Ukraine)—2.00, soybean flour (Prodenergo Ltd., Dnipro, Ukraine)—1.00, cream powder (Kovelmoloko, Kovel, Ukraine)—1.00, mustard powder (Ecotechnika CJSC, Kyiv, Ukraine) (1.00), sugar (Astarta-Kyiv, Kyiv, Ukraine)—1.50, salt (Artemsil, Soledar, Ukraine)—1.00, sodium bicarbonate (Bakery Product Ltd., Kyiv, Ukraine)—0.05, vinegar 9% (Koro, Vinnytsia, Ukraine)—8.00, lactic acid 80% (Systopt LLC, Lviv, Ukraine)—0.40, and drinking water. The hydrocolloid complex varied by sample as follows: M1 (0.50 CMC, 0.50 pectin NH, 0.50 xanthan gum), M2 (0.50 CMC, 0.50 pectin NH, 0.50 potato starch), M3 (0.30 CMC, 0.40 pectin NH, 0.30 xanthan gum), and M4 (0.70 pectin NH, 0.30 xanthan gum). The hydrocolloids utilized in this research were sourced from local suppliers, including thermo-reversible NH pectin (IL-Bakery, Kyiv, Ukraine), xanthan gum and carboxymethyl cellulose (Ester LLC, Kyiv, Ukraine), and potato starch (Vimal, Chernihiv, Ukraine).The content of hydrocolloid mixtures in samples M1 and M2 was 1.5%, while in samples M3 and M4 it was 1%. Samples M1 and M2 were acidified using 5 M acetic acid, whereas samples M3 and M4 utilized a combination of 9% spirit vinegar and 80% lactic acid.

2.2. Samples Preparation

Low-fat mayonnaise samples were prepared according to a standardised technological procedure to ensure reproducibility. Initially, protein-based components, including casein, mustard powder, soy flour, and cream powder, were dry-blended with salt, sugar, and sodium bicarbonate. This mixture was dissolved in water and pasteurised at 90 °C for 5 min, followed by hydration for 20–30 min at 65 °C under continuous stirring to obtain a homogeneous, lump-free paste. The oil and acid phases were then incorporated to form a coarse emulsion. To ensure proper functional activity of the hydrocolloids, xanthan gum and pectin NH were pre-dissolved in water at 65–75 °C before being added to the system. The final emulsion was homogenized to achieve the target dispersed structure, followed by packaging and storage in glass containers in refrigerator under 4 °C.

2.3. Methods

The sensory profile included eight key attributes: acidity, pungency, creamy taste, egg-like aftertaste, saltiness, sharpness, creaminess, and graininess. The results were visualized using a spider (radar) plot to compare the impact of different hydrocolloid systems on the organoleptic properties of the samples. Titratable acidity was determined by titration with 0.1 N NaOH using phenolphthalein indicator and expressed as acetic acid percentage. Active acidity (pH) was measured potentiometrically using a pH-150MI meter calibrated with standard buffers (pH 4.01, 6.86, and 9.18). Emulsion stability was evaluated by centrifugation at 1500 rpm for 5 min, followed by a 3 min boiling water bath and re-centrifugation for 5 min, with stability calculated as the volume percentage of the non-separated phase. Particle size distribution was determined using a PSA 1190 particle size analyzer (Anton Paar GmbH, Graz, Austria.). Microphotographs were taken using a Micromed microscope with an integrated 1k Pixelink camera and a 60× objective. Microscopic images were processed and analyzed using ImageJ 1.54g (National Institutes of Health (NIH), Bethesda, MD, USA) software to determine the morphometric characteristics of the emulsion droplets. This software was utilized to measure the diameters of the dispersed phase particles, allowing for the construction of a number-weighted particle size distribution. Rheological properties were tested with a Visco QC 300R rotational (Anton Paar GmbH, Graz, Austria). viscometer equipped with a PTD 175 Peltier temperature control device at 20 °C. Sensory analysis was conducted by a trained panel using a 10-point descriptive scale. To investigate the behavior of aqueous solutions of food-grade gelling agents and describe the flow curves, the generalized flow equation of the Casson rheological model was employed [15].

$${\tau }^{1/2}={\displaystyle \frac{{\tau }_c^{1/2}{\gamma }^{1/2}}{\chi +{\gamma }^{1/2}}}+{\eta }_c^{1/2}{\gamma }^{1/2}$$

(1)

where τ—shear stress applied to the emulsion; γ—shear rate at which the emulsion is deformed; τ_c—parameter characterizing the degree of particle aggregation and the dynamic yield stress; η_c—Casson viscosity. The parameter ${\tau }_c^{1/2}$ characterizes the degree of aggregation and the dynamic yield stress, while χ reflects the tendency toward forming large aggregates and their compactness, effectively distinguishing between plastic (χ = 0) and pseudoplastic (χ > 0) behaviors. Additionally, the viscosity coefficient ${\eta }_c^{1/2}$ represents the system’s internal resistance upon the complete destruction of droplet flocs.

All experiments were performed in triplicate, and results were presented as mean ± standard deviation. One-way ANOVA with Tukey’s multiple comparison post hoc test was performed to assess differences between groups. Differences were considered significant at p < 0.05.

3. Results and Discussions

3.1. Sensory Evaluations

The visual analysis of samples M1–M4 confirms the formation of a stable, homogeneous emulsion system with a characteristic light-cream color (Figure 1a). As shown in the comparative panel, all samples exhibit a smooth surface without visible oil separation or syneresis. The flow behavior and “spoon test” for samples M1 and M2 (Figure 1b) demonstrate a high degree of structural viscosity and thixotropic properties, where the emulsion maintains a stable “peak” and adheres well to the surface, indicating a well-developed stabilizing network formed by the hydrocolloid–protein complex. In M2, the inclusion of potato starch likely provided a “fat-like” lubricating effect, while in M4, the higher pectin ratio combined with lactic acid created a more refined and “round” mouthfeel. While M1 and M3 were physically more viscous due to the CMC/xanthan gum interaction, they were perceived as more “rubbery” or “sticky” rather than “creamy.”

Sensory profiling indicated a significant interaction between taste attributes, where high levels of acidity and saltiness in samples M1 and M2 appeared to mutually enhance each other, leading to a sharp and potentially unbalanced flavor profile. Total score calculations for the eight attributes show that sample M2 achieved the highest cumulative value of 46.5, followed by M1 with 43.0, while M4 and M3 scored 37.5 and 36.0, respectively. Based on the decrease in the total score, the samples can be ranked in the following sequence: M2 > M1 > M4 > M3 (Figure 2).

3.2. Stability and Acidity of Sauces

Sedimentation and thermal stability tests revealed that all investigated samples (M1–M4) maintained an emulsion stability of 99% both after 24 h and following a three-week storage period, exceeding the minimum requirement of 98% specified by DSTU 4560:2006 [16]. To ensure long-term structural integrity, the system pH must be maintained at or below the isoelectric point (IEP) of the protein components to facilitate the formation of a stable interfacial layer [17]. For the proteins utilized in these formulations, the literature defines the IEP values as 5.0 for soy proteins 13, 7.8 and 3.8 for mustard proteins 15, and 4.6–4.8 for casein 16. At pH > pH_IEP and high ionic strength, the system becomes thermodynamically unstable [18]. While samples M1 and M2, acidified with 5 M acetic acid, exhibited high titratable acidity levels of 0.788% and 0.980% respectively, this resulted in an excessively sharp odor and taste. In contrast, the use of 9% vinegar combined with 80% lactic acid in samples M3 and M4 allowed for the achievement of target active acidity (pH 3.70) while maintaining lower titratable acidity 0.474–0.476% (

Table 1. Sedimentation stability and acidity of M1–M4 samples.

Sample	Stability, % of Undestroyed Emulsion	pH	Acidity, %, Converted to Acetic Acid
M1	99	3,80	0.788
M2	99	3,94	0.980
M3	99	3,70	0.474
M4	99	3,70	0.476

). Given the lack of visible phase separation and the consistently high sedimentation stability across all samples, further analysis via rotational viscometry and laser diffraction was employed to detect subtle structural changes during storage.

3.3. Particle Size Estimation

Determining the droplet size distribution is fundamental to assessing the quality and kinetic stability of mayonnaise emulsions, as it directly dictates their rheological behavior, mouthfeel, and shelf-life. While microscopy allows for a direct qualitative assessment of the emulsion’s internal structure and the visual identification of phenomena such as flocculation or oil phase heterogeneity, laser diffraction provides quantitative data and statistical parameters, including specific percentiles like D10, D50, and D90.

The microstructural characteristics and droplet size distributions of the low-fat mayonnaise-type emulsions (M1–M4) were evaluated through optical microscopy coupled with digital image processing via ImageJ software (Figure 3). The resulting number-weighted mean diameter (d [1,0]) values were determined to be 4.99 for sample M1, 4.50 mkm for M2, 4.00 mkm for M3, and 2.82 mkm for M4. Sample M4 exhibited the most refined and homogeneous microstructure, represented by a nearly twofold decrease in droplet size compared to the others. The histograms illustrate a transition from a relatively broad and polydisperse distribution in M1 and M2 toward a narrower, more uniform profile in M4, where the majority of the oil droplets are concentrated within the 1–3 mkm range. This quantitative shift is qualitatively supported by the micrographs, which reveal a dense population of fine droplets in the M4 emulsion, whereas M1 and M2 show larger, more isolated oil globules with a higher tendency toward localized aggregation—floculation.

In agreement with the microscopy results, the diffraction profiles (Figure 4) confirm that sample M4 (red curve) possesses the finest dispersion, with its primary peak shifted significantly toward the lower diameter range compared to the other formulations. Meanwhile, samples M1, M2, and M3 exhibit broader distributions with peaks extending beyond 10 mkm, indicating the presence of larger oil globules or flocculated clusters. Crucially, the laser diffraction analysis reveals a secondary population of sub-micron droplets, specifically a peak around 0.3 mkm, which was not captured in the ImageJ numerical analysis.

The results of particle size analysis of samples M1–M4, obtained using an optical microscope and laser diffraction, are shown in

Table 2. Particle size characteristics of samples M1–M4 determined by optical microscopy and laser diffraction.

	Microscopy	Laser Diffraction
Sample	d [1,0]	d [4,3]	D50
M1	4.99	3.75	4.715
M2	4.50	5.751	4.028
M3	4.00	3.919	3.151
M4	2.82	2.958	2.415

. When characterising the average particle size of these samples, we consider the median percentile D50 to be a more representative parameter than the volume-weighted average diameter d [4,3]. This preference is due to the fact that the observed particle size distributions deviate from the classical monomodal behaviour. In such non-ideal distributions, the use of d [4,3] can give misleading results due to its high sensitivity to minor fractions of large particles or tails of the distribution. Conversely, D50 provides a more reliable measure of central tendency, accurately reflecting the size threshold that divides the total volume of solid particles in half. Therefore, D50 is used as the primary indicator for estimating the average sizes of the systems under study to ensure a more accurate interpretation of the physical characteristics of the samples.

The number-weighted mean diameter (d [1,0]) determined via optical microscopy and ImageJ analysis was found to be remarkably close to the volume-weighted mean diameter (d [4,3]) obtained through laser diffraction. Theoretically, for polydisperse emulsions, d [4,3] should significantly exceed d [1,0] due to the higher sensitivity of the volume-weighted average to larger droplets. This observed convergence is likely attributed to the resolution limits of the optical system at the utilized magnification. Specifically, the inability to detect sub-micron droplets (below 1 mkm) led to an artificial truncation of the lower end of the size distribution, thereby inflating the numerical mean value.

3.4. The Analysis of the Viscometric Measurements for the Investigated Samples

The assessment of viscosity and its dependence on the shear rate is a critical factor in evaluating the structural stability and sensory performance of emulsion systems. The rheological behavior was analyzed by applying the generalized Casson model to the flow curves, allowing for a physical interpretation of the structural transformations within the samples [19,20]. The calculated coefficients for the investigated samples, presented in

Table 3. Casson model coefficients for describing the viscosity dependence of samples on shear rate.

Sample	${\mathsf{\tau }}_{\mathit{c}}^{1/2}$	χ	${\mathsf{\eta }}_{\mathit{c}}^{1/2}$	${\mathbf{R}}^2$	${\mathsf{\sigma }}_{\mathbf{e}\mathbf{s}\mathbf{t}}$
M1	11.5 ± 0.9	0.153 ± 0.021	1.24 ± 0.66	0.9994	0.372
M2	4.93 ± 0.17	0.072 ± 0.008	1.60 ± 0.13	0.9977	0.309
M3	8.37 ± 0.75	0.063 ± 0.022	1.34 ± 0.65	0.9980	0.414
M4	3.74 ± 0.69	1 × 10−12 ± 0.03	3.02 ± 0.93	0.9901	0.712

, demonstrate that the Casson model provides a high degree of mathematical accuracy, as evidenced by superior correlation coefficients (${\mathrm{R}}^2$) and low standard deviations (${\mathsf{\sigma }}_{\mathrm{e}\mathrm{s}\mathrm{t}}$).

Analysis of the experimental data reveals that sample M3, which incorporates a CMC/xanthan/pectin complex, exhibits a rheological profile nearly identical to the commercial control samples, the rheological profile of the which is characterized by non-Newtonian pseudoplastic behavior with a pronounced shear-thinning effect. Based on the analysis of similar commercial emulsions, these products typically exhibit a significant yield stress and high consistency coefficients, which are essential for maintaining a stable semi-solid structure at reduced oil concentrations [21]. This similarity in the Casson model parameters suggests that sample M3 possesses a structural organization and flow stability comparable to industrial standards, supporting the hypothesis that these rheological metrics can serve as a basis for optimizing the sensory characteristics of developed food emulsions.

The rheological data presented in

Table 4. Rheological properties of the investigated samples: static yield point and dynamic viscosity at different shear rates.

Sample	Yield Point	Viscosity, Pa·s
	τstatical, Pa	3 s−1	50 s−1
M1	112	532.01	3.64
M2	–	19.92	1.59
M3	158	40.32	2.44
M4	–	26.70	1.50

reflects a significant correlation between the polysaccharide composition and the structural behavior of the mayonnaise samples. The direct measurement of the static yield stress using a vane rotor reveals that samples M1 and M3 exhibit high yield points of 112 Pa and 158 Pa, respectively, indicating a robust internal network capable of resisting flow until a critical stress is applied. This plastic behavior is primarily attributed to the synergistic interaction between xanthan gum and the other stabilizers, which facilitates the formation of a stable, three-dimensional gel-like matrix. In contrast, the static yield stress for samples M2 and M4 was unmeasurable, signifying that these samples flow under negligible force and possess a more liquid-like nature. The absence of a yield point in M2 suggests that potato starch, even in combination with CMC and pectin, fails to replicate the structural integrity provided by xanthan gum at these concentrations, while the M4 results indicate that the pectin–xanthan ratio without CMC is insufficient to establish a solid-like threshold.

These differences are further supported by the viscosity data obtained at different shear rates. At low shear rate (3 s⁻¹), which may be considered a quasi-rest or low-deformation state, sample M1 displays markedly higher viscosity compared to the other formulations, consistent with its high yield stress and the synergistic thickening effect of the combined CMC, pectin, and xanthan gum system. Sample M3, despite a lower total polysaccharide content, also shows elevated viscosity, indicating efficient network formation due to the presence of xanthan gum, which is known to strongly enhance elasticity and resistance to flow when combined with pectin [22]. In contrast, M2 and M4 exhibit substantially lower viscosities at low shear, in line with their inability to develop a measurable yield stress and their more fluid-like nature.

At higher shear rate (50 s⁻¹), representative of the shear conditions associated with oral processing and spreading of mayonnaise in the mouth, all samples show a pronounced decrease in viscosity, confirming shear-thinning behavior typical of emulsion-based systems [23]. However, M1 and M3 retain higher viscosities relative to M2 and M4, suggesting that even under strong deformation their internal structure is not fully disrupted. The significantly lower viscosities of M2 and M4 at this shear rate indicate rapid structural breakdown and easy flow, which can be attributed to the absence of xanthan gum in M2 and to the dominance of pectin without sufficient network reinforcement in M4.

4. Conclusions

This study demonstrates that a combination of mustard powder, soybean flour, casein, and cream powder effectively substitutes for egg yolk in low-fat mayonnaise-type sauces while maintaining high emulsion stability (99%). The integration of hydrocolloid blends proved essential for tailoring the structural and sensory profiles. Specifically, samples containing xanthan gum (M1 and M3) exhibited a robust three-dimensional network with significant static yield stress (112–158 Pa), ensuring the desired plastic behavior. In contrast, formulations with potato starch or high pectin content without CMC (M2 and M4) displayed more liquid-like properties. Sample M3, featuring a CMC/xanthan/pectin complex, showed the most balanced physicochemical properties and rheological parameters comparable to commercial standards. These findings highlight the synergistic potential of hydrocolloid–protein systems in developing health-oriented, egg-free emulsified sauces.