Understanding the synergistic effect of pea protein and rice bran protein interaction in stabilizing palm kernel oil‐in‐water emulsion assisted by high‐pressure homogenization

Abstract

BACKGROUND

Plant‐based beverages have recently seen a significant increase in market demand. However, many of these products suffer from poor emulsion stability and low protein content. Gums have commonly been used to enhance emulsion stability but they do not improve the amino acid profile. This study investigated the use of multiples plant proteins to enhance both the stability and nutritional value of plant‐based beverages.

RESULT

Pea and rice bran proteins both enhanced emulsion stability. Pea protein enhanced the viscosity of the continuous phase whereas rice bran protein lowered interfacial tension. When applied synergistically, competitive adhesion occurred. Rice bran protein gradually displaced pea protein from the oil droplet surface as its concentration increased, leading to emulsion destabilization due to the displaced pea protein. The use of high‐pressure homogenization further enhanced the stability of the emulsion by unfolding protein partially. However, increasing homogenization pressure (>500 Bar) and homogenization cycle (>2 cycles) led to protein aggregation due to excessive exposure of its hydrophobic core. The emulsion formed was resistant to coalescence at 4 °C for 28 days and was stable under high pH and low ionic conditions.

CONCLUSION

The synergistic combination of plant proteins and the effective utilization of co‐processing (homogenization) can enhance the functionality of the individual proteins significantly, leading to the formation of a stable emulsion. The use of plant protein mixture as a stabilizer not only improved the emulsion stability but also ensured a plant‐based beverage with a complete amino acid profile for the vegan community. © 2024 The Author(s). Journal of the Science of Food and Agriculture published by John Wiley & Sons Ltd on behalf of Society of Chemical Industry.

INTRODUCTION

The use of plant‐based beverages as alternatives to replace dairy milk has attracted the attention of researchers because individual plant proteins cannot provide emulsion stability and nutritional content like dairy protein. Most plant‐based beverage products on the market use gums as the main stabilizers of the emulsion system. They increase the bulk viscosity and provide the desired mouthfeel. ¹ The poor amino acid profile of plant‐based beverage is therefore not well addressed. Alternative approaches include blending plant protein with dairy protein or other plant proteins. ² , ³ The former is not suitable for application in plant‐based beverages whereas the latter involves a complicated processing step that consumes time and energy. The development of a rapid and efficient approach is therefore crucial to address this challenge and to enable upscaling within the food industry, all while maintaining excellent nutritional value. This will be instrumental in meeting the growing consumer demand for healthy and sustainable food products.

The use of animal proteins in food emulsions has been well documented. The use of plant proteins has been gaining attention recently. ⁴ , ⁵ , ⁶ Nevertheless, the use of plant protein often suffers from drawbacks such as poor solubility and an unbalanced amino acid profile. ² A study by McCarthy et al. ⁷ suggested that pea protein was poorly soluble in water whereas studies on rice bran and palm kernel protein showed that there is a decrease in functional properties at pH levels close to their isoelectric points. ⁶ , ⁸ These drawbacks often limit the applicability of plant protein in food hydrocolloids. To explore the use of plant protein further as functional ingredients for the production of vegan food, efforts were made to improve the performance of plant proteins by mixing them with other functional ingredients.

Extensive research has investigated the effect on their functional properties of associating proteins with other ingredients. Recent studies were conducted in coupling plant protein with antioxidants such as proanthocyanidins and saponin. ⁹ , ¹⁰ The complex performed better in terms of emulsion stability and oxidative stability. However, this coupling does not resolve the issue of imbalanced amino acid profiles in plant protein.

Another alternative is to combine different proteins with complementary amino acid profiles to improve their functional and nutritional properties. McClements et al. ¹¹ suggested that using multiple proteins through covalent conjugation, physical complexes, or simple mixing can enhance emulsion stability. Nonetheless, dairy and plant proteins are often blended, ² which is not a suitable vegan option.

Limited studies have been conducted using dual plant proteins for the production of superior emulsifiers. A study by Wang et al. ³ successfully developed superior dual protein (rice and walnut) nanostructures that created a high internal‐phase emulsion with superior stability and resistance towards freeze–thaw challenge. The problem associated with this method is that the process involved high pH levels, which can alter the structure of proteins permanently, causing the loss of their bioactivity. The utilization of acids and bases can introduce additional salt that can impact the sensory and stability properties of the final products. Thus, there is a need for a more versatile method for the production of an effective emulsifier that counters the problem of poor functional properties and nutritional value.

Other than the protein coupling mentioned above, the use of different processing treatments can also counter the issue of poor protein functionality by promoting the formation of protein aggregates of different sizes and morphology. ¹² High‐pressure homogenization (HPH) is a potential technique for this issue that is commonly adopted in the food industry. The shear forces generated during the process promote the unfolding of the protein, improving its solubility and functional properties. ¹³ A study by Melchior et al. ¹⁴ also showed improved digestibility of pea protein upon HPH treatment, suggesting the role of HPH in improving not only the appearance and stability but also the nutritional value of a food emulsion. As the use of HPH treatment displays promising potential in solving the issue of poor functional properties of plant protein, the low nutritional value of plant protein can be countered by the simple blending of multiple different plant proteins to achieve a more balanced amino acid profile. Nevertheless, no study has been performed previously on the coupling of different plant proteins using HPH treatment but other functional ingredients such as gum, lipids, and antioxidants were used instead.

Hence, this study aimed to explore the synergistic functionality of pea and rice bran protein blend and co‐processing HPH treatment in stabilizing palm kernel oil‐in‐water emulsion. Pea protein and rice bran protein were selected as both demonstrate a promising effect in stabilizing an oil‐in‐water emulsion when used alone and they have a complementary amino acid profile to palm kernel. Both of them also have hypoallergy protein sources and rice bran protein is source from an underutilized food by‐product. An accelerated shelf‐life study under varying environmental stress tests (pH and ionic challenges) was performed to assess the emulsion performance under different processing conditions.

The potential route of destabilization was analyzed and compared across different storage temperatures. This study elucidates the molecular mechanisms by which pea and rice bran proteins act synergistically to stabilize the oil‐in‐water emulsions. It investigates the effectiveness of high‐pressure homogenization in this combined protein system. The findings are important for the development of nutritious, shelf‐stable, vegan‐friendly, and clean‐labeled food emulsions using a rapid and convenient processing route that eliminates the need for gums.

MATERIALS AND METHODS

The palm kernel was generously sponsored by Sime Darby Plantation (Carey Island, Selangor, Malaysia). Rice bran protein was purchased from Undersun Biomedtech (Shaanxi, China) and pea protein was procured from MyProtein (Selangor, Malaysia). All chemicals and solvents used were of analytical grade.

Preparation of palm kernel milk emulsion

Palm kernel milk emulsion was prepared according to the method described in utility innovation UI 2023001869. In brief, extracted palm kernel protein (96.5 mL) was mixed with palm oil (3.5 mL) and emulsified using ultra Turax homogenizer (IKA T25, Staufen, Germany) at 10 000 rpm for 1 min. The resulting coarse emulsion was further subjected to high‐pressure homogenization (PandaPlus 2000, GEA Group, Düsseldorf, Germany). The final emulsion was then pasteurized at 63 °C for 30 min.

The effects of individual protein concentrations (1, 5, 10, 25, 50 g L⁻¹) and binary protein mixes (5:1, 1:1, 1:5) were examined. The effects of different homogenization conditions were also studied; the homogenization was conducted using a Panda PLUS 2000 high‐pressure homogenizer (GEA Group, Düsseldorf, Germany) under different homogenization pressures (0, 250, 500, 750, 1000 Bar) and homogenization cycles (1, 2, 3, 4, 5). ¹ , ² , ³ , ⁴ , ⁵

The prepared emulsions were stored at 4 °C for a day before being analyzed for surface protein adsorption fraction, interfacial tension, oil droplet particle size, zeta potential, viscosity, and microstructure. Emulsion stability index (ESI) analysis was performed on the sample after 7 days of storage at 4 °C.

Protein characterization

Interfacial tension

Interfacial tension was measured using a Sigma 702 Attension force tensiometer (Biolin Scientific, Stockholm, Sweden). The density of the aqueous phase and oil phase were measured using a density probe followed by a manual calculation of the density difference between the two interacting phases. The interfacial tension between the aqueous phase and oil phase was measured using a platinum Du Noüy ring at room temperature. The measured interfacial tension was then corrected using the Huh–Mason correction.

Protein intrinsic fluorescence

The intrinsic fluorescence of protein was measured according to Jorgensen et al. ¹⁵ with slight modification using fluorescence spectroscopy (FL6500, Perkin Elmer, Shelton, US) in a quartz cuvette at a perpendicular angle to the light source. Scanning of the emission spectra was conducted using an excitation wavelength of 290 nm with a slit opening of 1 nm. The emission spectra were collected between 300 – 400 nm with a slit opening of 10 nm and a scan rate of 240 nm per min.

Protein particle size

The particle size of protein after homogenization treatment was measured using a Mastersizer (Mastersizer 3000, Malvern Instruments Ltd, Malvern, UK) with a refractive index of 1.52 ¹⁶ and 1.33 for dispersed and continuous phases, respectively.

Emulsion characterization

Protein adsorption fraction

The protein adsorption fraction was determined by subjecting the emulsion sample to centrifugation at 7000 x g for 5 min. The cream layer was discarded while the serum sample was vortexed for 1 min and subject to a Kjeldahl analysis of the protein content. The protein adsorption fraction was then calculated using the formula: ¹⁷

$$\text{Adsorped protein fraction}\left(\%\right)=\frac{C_0-C_l}{C_0}\times 100\%$$ (1)

where:

• C₀ = protein concentration in whole emulsion

• C_l = protein concentration in serum layer

Particle size and zeta potential

The particle size distribution and zeta potential of the oil droplets in the emulsion were measured using a Mastersizer 3000 (Malvern Instruments Ltd) and Zetasizer (Malvern Nano‐ZS, Malvern Instruments Ltd), respectively. A refractive index of 1.45 and 1.33 was used for dispersed and continuous phases, respectively. A 100 times dilution was performed using water for the sample before zeta potential measurement.

Viscosity

The viscosity of the emulsion was measured using an MCR 302 modular compact rheometer (Anton Paar, Graz, Austria) with a concentric ring (20 mm diameter) at 25 °C by varying the shear rate from 0.01 s⁻¹ to 100 s⁻¹. The complex viscosity flow curves obtained were then analyzed and fitted onto a Newtonian model using the Anton Paar software to obtain the apparent viscosity.

Emulsion stability index

Emulsion stability was analyzed using turbidity measurements. ⁶ Briefly, the emulsion was prepared freshly and stored at 4 °C for 1 day before centrifugation at 5000 × g for 10 min. Then, 100 μL of the serum phase was withdrawn immediately and diluted into 4 mL of distilled water and absorbance was measured at 500 nm using an ultraviolet‐visible (UV‐visible) spectrophotometer (AHS Laboratory Supplies, Selangor, Malaysia). Twenty microliters was used for the sample after determining the best protein ratio due to an excessively high absorbance from the stable emulsion. The measurement was repeated using samples after storing at 4 °C for 7 days. The emulsion stability index (ESI) was calculated using the formulae as follows:

$$\text{emulsifying activity index},\mathrm{EAI}\left(\frac{{\mathrm{m}}^2}{g}\right)=\frac{2.303\times 2\times A_0}{C\times \varphi }$$ (2)

where, C = protein concentration, $\varphi$ = oil volume fraction, A₀ = absorbance at 500 nm.

$$\text{emulsion stability index},\mathrm{ESI}\left(\%\right)=\frac{{\mathrm{EAI}}_{7\text{days}}}{{\mathrm{EAI}}_{1\mathrm{day}}}\times 100$$ (3)

Microstructure

The fat droplet distribution was observed using a fluorescence microscope (Olympus BX 51, Tokyo, Japan). In brief, 20 μL of the emulsion was dyed with 2 μL of 0.5 g L⁻¹ Nile red and 10 g L⁻¹ rhodamine B before observing using an excitation wavelength of 450 – 480 nm. ¹⁸

pH and ionic challenges

Samples were adjusted to pH 3, 5, 7, 9, and 11 using 0.1 mol/L NaOH or 0.1 mol/L HCl to assess the effect of pH on the stability of the emulsion. An ionic stability test was also carried out by adding different concentrations of NaCl (50, 100, 150, 200, 250 mmol/L) into the emulsion at different pH values. The effect of divalent cations on the ionic stability of the emulsion was assessed by replacing NaCl with CaCl₂.

Accelerated shelf‐life study

The shelf stability of the emulsion was determined using an accelerated shelf‐life study. An emulsion containing 3 g L⁻¹ pea protein, and 3 g L⁻¹ rice bran protein was prepared through homogenization at 500 Bar for 2 cycles and split into 15 mL centrifuge tubes for different time points of measurement. Briefly, the samples were stored at three different temperatures: 5, 25, and 45 °C, and analysis was performed on days 0, 1, 3, 7, 10, 14, 17, 21, 24, and 28. Sodium azide (0.2 g L⁻¹) was added to the sample to eliminate the microbial effect on emulsion stability.

Creaming kinetic

A linear regression was used to determine the relationship between creaming rate and time following the equation below

$$C=k_{0}t+C_0$$ (4)

where

• C = creaming index at day t

• C ₀ = creaming index at day 0

• t = number of days

• k ₀ = rate of creaming

Then Q₁₀ was calculated using the formula

$${\mathit{Q}}_{\mathbf{10}}^{\left(\frac{{\mathit{T}}_{\mathbf{1}}-{\mathit{T}}_{\mathbf{2}}}{\mathbf{10}}\right)}=\frac{{\mathit{K}}_{\mathit{T}+\mathbf{20}}}{{\mathit{K}}_{\mathit{T}}}$$ (5)

Destabilization mechanisms

The particle sizes of oil droplets at different storage time points were measured. The calculation of the rate of coalescence, Ostwald ripening, and flocculation was adopted from Buffo and Reineccius ¹⁹ and Wu et al. ²⁰ For the flocculation rate, additional particle size analysis was done using 1 g L⁻¹ sodium dodecyl sulfate (SDS) as a dispersant instead of water. The rates of coalescence, Ostwald ripening, and flocculation were calculated using the formulas

$$\text{coalescence}:\ln \left(\frac{D_1}{D_0}\right)=\frac{1}{3}\mathit{tC}$$ (6)

$$\text{Ostwald ripening}:r^3=\mathit{kt}$$ (7)

$$\text{flocculation}:\mathit{Fl}\left(\%\right)=\left(\frac{d_w}{d_g}-1\right)\times 100$$ (8)

where

• D ₀ = particle size of oil droplets during day 0;

• D ₁ = particle size of oil droplets during day t;

• t = storage time in days;

• C = rate of coalescence;

• r = radius of the oil droplets;

• k = rate of Ostwald ripening;

• D _w = emulsion oil particle size in water;

• D _g = emulsion oil particle size in 1 g L⁻¹ SDS.

Statistical analysis

All analyses were performed in triplicate and results were reported as means ± standard deviations. A one‐way analysis of variance (ANOVA) followed by Tukey post hoc analysis was used to analyze the differences in means between different parameters of the same treatment using GraphPad Prism version 10.0.2 for Windows (GraphPad Software Inc, Boston, Massachusetts, USA). A significance level of 5% (P < 0.05) was used in this study.

RESULT AND DISCUSSION

Protein as stabilizer

Pea protein

The addition of pea protein showed no improvement in the stability of the emulsion, as may be seen from the lower ESI value (Fig. 1(c)). The stability of the emulsion decreased with an increase in pea protein concentration. This follows the study by Sridharan et al., ²¹ who found that 2 g L⁻¹ of pea protein was sufficient to stabilize an emulsion with 100 g L⁻¹ of oil content. As the oil content in this study is only around 35 g L⁻¹, it is expected that excess pea protein existed as free protein. The presence of free protein was supported by the reduced adsorbed protein fraction, indicating that more protein remained unadsorbed to the oil droplets (Fig. 1(b)). The free protein acted as an electrostatic screening barrier, ²¹ lowering the zeta potential and promoting the coalescence of oil droplets into larger droplets. Nonetheless, at low concentrations, the addition of pea protein significantly reduced the interfacial tension (Fig. 1(d)) and marginally increased the viscosity of the emulsion. Such an increase had been reported in a recent study as a result of gel formation. ²² However, the increase in viscosity in the present study was more likely due to the increase in soluble protein content rather than the gelling effect, as the emulsion was not subjected to heat treatment that was intense enough for protein denaturation and formation of a protein gel (>80 °C for 10 min). ²³

Figure 1.

Changes in particle size (a), adsorbed protein fraction (b), emulsion stability index and zeta potential (c), interfacial tension changes by protein solution, and viscosity of the emulsion (d) when different concentrations of pea protein were added as stabilizers. N/A, not applicable. Nil, negligible. Distinct alphabets representing different protein concentrations within the same analysis denote statistically significant differences (P < 0.05) between those protein concentrations.

Rice bran protein

On the other hand, the addition of rice bran protein enhanced emulsion stability at low concentrations (<10 g L⁻¹). However, emulsion stability reduced at higher concentrations (≥10 g L⁻¹). Unlike pea protein, rice bran protein displayed a superior surface‐active property as observed from the significant decrease in oil/water interfacial and higher adsorbed protein fraction (Fig. 2(b),(d)). The surface activity of rice bran protein was previously reported by Sun et al. ²⁴ Chen et al. ²⁵ also found that an increase in rice bran protein caused the aggregation of proteins due to the screening of electric charges by the protein surface charges. This is supported by the reduced zeta potential of emulsion when 25 g L⁻¹ or more of the rice bran protein was added. The protein aggregates forced the droplets into close proximity, producing coalescence and leading to instability. ²⁶ Although there was an increase in viscosity with increasing rice bran protein, it only affected the emulsion stability minimally. Hence, the increasing viscosity with protein concentration is likely due to the presence of protein aggregates instead of a protein gel network that can hinder particle movements. This matches the conclusion of Yeom et al., ²⁷ in which rice bran protein could not form a protein gel at a concentration range of 5–15 g L⁻¹.

Figure 2.

Changes in particle size (a), adsorbed protein fraction (b), emulsion stability index and zeta potential (c), interfacial tension changes by protein solution and viscosity of the emulsion (d) when different concentration of rice bran protein were added as stabilizers. N/A, not applicable. Distinct alphabets representing different protein concentrations within the same analysis denote statistically significant differences (P < 0.05) between those protein concentrations.

Synergistic protein action

As both proteins stabilized the emulsion by different mechanisms, the potential synergistic effect of the two proteins was further evaluated. As Fig. 3 shows, there is an increase in the emulsion stability with an increasing proportion of rice bran protein as supported by the lower interfacial tension of the emulsion (Fig. 3(d)).

Figure 3.

Changes in particle size (a), adsorbed protein fraction (b), emulsion stability index and zeta potential (c), interfacial tension changes by protein solution and viscosity (d), and microstructure (e) of emulsion when different proportion of rice bran and pea protein were added synergistically as stabilizer. Distinct alphabets representing different protein concentrations within the same analysis denote statistically significant differences (P < 0.05) between those protein concentrations. Fluorescence image observed under 200× magnification with Nile red and Rhodamine B as dyes. Scale bar = 30 μm.

During the formation of emulsions, rice bran protein, which is smaller in size (≈24 μm), rapidly deposits onto the oil droplets surface, followed by pea protein (≈90 μm). ²⁸ At low concentrations of rice bran protein, the oil surface was only partially coated, allowing pea protein to adhere to the exposed areas. As the concentration of rice bran protein increased, some pea proteins were displaced as free pea protein. This was confirmed by the observation of red agglomeration in the fluorescence microscopic image, which grew larger as the concentration of rice bran protein was equal to, and higher than, pea protein. Hinderink et al. ²⁹ also suggested the possibility of protein displacement by another protein when the latter is present at a higher concentration. The displaced free pea protein causes an electrostatic screening effect and forms protein aggregates, which reduces the absolute zeta potential and increases the particle size as observed in this study (Fig. 3(a),(c)) with an increase in rice bran protein. When 3 g L⁻¹ of pea protein and rice bran protein was used, the emulsion achieved the highest viscosity, suggesting that this protein combination retained most particles in the bulk system, which restricts flow.

Effect of high‐pressure homogenization

Homogenization plays an essential role in emulsion formation as it provides sufficient shearing force for the disruption of the dispersed phase into the continuous phase. In the present study, emulsion stability increased with increasing pressure. However, beyond 500 Bar, stability started to decline. The introduction of homogenization significantly (P < 0.05) reduced the protein size, as evidenced by the disappearance of the large red particles under the fluorescence microscope (Fig. 4(e)). Homogenization also induced protein unfolding, as indicated by the red shift in the intrinsic fluorescence emission maximum (Fig. 4(a)). Other studies have demonstrated the partial unfolding of pea protein and rice bran protein after high‐pressure homogenization. ¹³ , ¹⁴

Figure 4.

Changes in protein particle size and maximum intrinsic fluorescence emission (a), oil droplet particle size (b), emulsion viscosity (c), emulsion zeta potential and ESI (d), and emulsion microstructure (e) when subjected to different homogenization pressures. Distinct alphabets representing different homogenization conditions within the same analysis denote statistically significant differences (P < 0.05) between those homogenization conditions. Fluorescence image observed under 200× magnification with Nile red and rhodamine B as dyes. Scale bar = 30 μm.

In contrast, greater unfolding of the protein structure was detected under extreme high pressure. This results in emulsion instability, which can be observed by a reduction in the absolute zeta potential and the increase in the oil droplet size (Fig. 4(b),(d)). The increase in the exposure of the hydrophobic site of the protein may result in a bridging flocculation that causes multiple oil droplets to stick in close proximity. ³⁰ The formation of larger oil droplets, in turn, contributed to the increase in emulsion viscosity as they are more prone to shear flocculation, which restricts bulk system movement. ³⁰ Higher emulsion stability recorded for the 0 Bar sample is likely due to the poor emulsion stability that separated at 1 day of storage and resulted in minimum turbidity difference between day 1 and day 7. This is evidenced by a clear solution observed for non‐homogenized samples.

There was a decrease in the particle size of the oil droplets with an increasing homogenization cycle (Fig. 5(b)). A higher cycle of homogenization provides greater shearing forces for the formation of smaller oil droplets, which was reflected in the fluorescence microscopic image with numerous droplets forming a homogenous background (Fig. 5(e)). Nonetheless, the decrease in particle size does not provide better emulsion stability due to a low zeta potential (Fig. 5(d)). A similar result had been reported by Sridharan et al., ²¹ in which pea protein emulsion with a low average particle size displayed a low emulsion stability. The stabilization and destabilization effect of the homogenization cycle can be explained by the reduction in protein particle size at 2 cycles of homogenization and the increase in protein particle size beyond 2 cycles of homogenization (Fig. 5(a)). The increase in protein particle size is likely due to the aggregation and formation of protein lumps, which is supported by the blue shift in the emission maximum of the protein intrinsic fluorescence (Fig. 5(a)), suggesting that the fluorescing amino acids are buried in a more hydrophobic environment. ³¹ The aggregation of protein reduces the exposed electrostatic active surface leading to a reduction in the absolute zeta potential. ²¹

Figure 5.

Change in protein particle size and maximum intrinsic fluorescence emission (a), oil droplet particle size (b), emulsion viscosity (c), emulsion zeta potential and ESI (d), and emulsion microstructure (e) when subjected to different homogenization cycles.  Distinct alphabets representing different homogenization conditions within the same analysis denote statistically significant differences (P < 0.05) between those homogenization conditions. Fluorescence image observed under 200× magnification with Nile red and Rhodamine B as dyes. Scale bar = 30 μm.

pH and ionic stability

As replacements for dairy milk, plant‐based beverages must be versatile for various food applications. Hence, it is vital to evaluate the emulsion stability under different pH and ionic strength. Generally, the emulsion was stable under high pH and low ionic strength as observed under fluorescence microscope (Fig. 6). As the pH decreases below 7, significant creaming (Fig. 7) and an increase in average particle size can be observed. The result is consistent with the findings of Chee et al., ⁴ Zhang et al., ⁶ and Zhu et al. ¹³ who reported pH 4 and pH 5 as isoelectric points (PI) of palm kernel protein, pea protein, and rice bran protein, respectively, leading to the formation of protein aggregates. This is confirmed by the small difference in particle size between the sample measured in water and 1 g L⁻¹ SDS.

Figure 6.

Fluorescence microscope image (200× magnification) of emulsion prepared with different concentration of NaCl at different pH after storage for 1 day at 4 °C. Samples are dyed with Nile red and rhodamine B (lipid = yellow, protein = red).

Figure 7.

Average particle sizes (a), zeta potential (b), and creaming index of samples with 0 mM (c) 50 mM (d) 100 mM (e) 150 mM (f) 200 mM (g) 250 mM (h) of NaCl added at different pH values. The solid line represents particle size measured in water whereas the dotted line represents particle size measured in 0.1% sodium dodecyl sulfate (SDS).

However, the emulsion stability under pH 3 was enhanced when ionic strength increased, as evidenced by the higher absolute ζ‐potential. The ζ‐potential correlates positively with the surface charge density, σ, and is inversely related to the Debye screening length, κ ⁻¹. The latter, on the other hand, showed a direct relationship with the concentration and valency of ions in the electrolyte. ³² Thus, an increase in absolute ζ‐potential with increased salt concentration suggests the occurrence of ionic binding between chloride ions and the amino group on the protein surface, which causes an increase in negativity of the ζ‐potential.

Despite being less stable under a high ionic environment, emulsion stability can be improved when subjected to neutral to alkaline conditions. A higher stability in alkaline conditions occurs because the protein is furthest away from the PI, hence resulting in greater deprotonation of the amine group, increasing the surface charge density, and enhancing solubility as well as emulsification stability. ¹³ This is confirmed by the significantly smaller particle size and higher absolute ζ‐potential of the sample at pH ≥ 7 (in the absence of salt).

Effect di‐valence ions

The effect of cations with different valency (Ca²⁺) on the stability of the emulsion was also examined. This is important as the fortification of minerals in the food systems often involves multivalent ions such as Ca²⁺, Mg²⁺, and Fe^2+/3+.

Figure 8 shows highest emulsion stability was achieved with the lowest CaCl₂ concentration (50 mM). Creaming started to occur at 100 mM and no difference in the creaming index was observed between 150 mM to 250 mM. There was no significant difference in the particle size distribution measured in 1 g L⁻¹ SDS from 100 mM CaCl₂ onward suggesting that the degree of coalescence remained unchanged with increasing CaCl₂, but there was a decrease in the extent of flocculation. The incorporation of CaCl₂ resulted in low absolute ζ‐potentials promoting particle aggregation, resulting in a complete flocculation as indicated by the clear serum phase when CaCl₂ was added. ³³ Hence, all these imply a stronger effect of Ca²⁺ on the stability of the emulsion than Na⁺ potentially due to its higher valency, which had a greater effect on the Debye screening length.

Figure 8.

Change in average particle size, zeta potential and creaming index of emulsion sample with different concentrations of CaCl added at pH 7. Different small letters indicate significant differences between different treatment groups; different capital letters indicate significant differences within the same treatment group.

Shelf‐life study

The shelf stability of the emulsion was assessed using a thermal accelerated condition. As Fig. 9 indicates, emulsion stored at 5 °C showed no creaming throughout 28 days of study. However, a drastic increase in the creaming index was observed at 45 °C starting from day 14 whereas the emulsion stored at 25 °C experienced a slow and consistent increase in the creaming index with storage time. The destabilization mechanisms were further determined using particle size distribution data based on Buffo and Reineccius ¹⁹ and Wu et al., ²⁰ as shown in Table 1. No destabilization mechanism dominated the separation process. The separation under low and high temperatures was mainly caused by Ostwald ripening and coalescence, respectively. This coincides with the study by Binner et al. ³⁴ Under elevated temperature, viscosity and interfacial tension reduced, promoting coalescence. The rates of creaming of the emulsion stored at 5 °C, k_5°C, was further determined by using a Q₁₀ of 1.478 derived from 25 to 45 °C. An Arrhenius plot was obtained and the activation energy, Ea, was calculated to be around 28.6 kJ mol⁻¹. This is similar to the 7.4 kcal mol⁻¹ obtained by Buffo and Reineccius, ¹⁹ working on diluted beverages (<50 g L⁻¹). The high activation energy indicates that a high energy level is required to initiate the separation process and suggesting a more stable emulsion.

Figure 9.

Change in creaming index of sample stored at different temperatures for 28 days.

Table 1.

Creaming regression and destabilization mechanism of samples stored at different temperatures for 28 days, viscosity, and Q₁₀ value of the emulsion

Storage temperature (°C)	Regression equation	R 2 value	Q10	Coalescence		Ostwald ripening		Flocculation
				r 2	k (μm/day)	r 2	k (μm/day)	r 2	k (μm/day)
5	N/A	N/A	N/A	0.1120	−0.0052	0.5607	−0.0032	0.5326	−0.0029
25	y = 1.0657x – 3.2223	0.9445	1.478	0.5864	−0.0142	0.4222	−0.0022	0.5428	−0.0018
45	y = 2.3243x – 3.0033	0.8589		0.6234	−0.0189	0.4265	−0.0013	0.4168	−0.0004

Viscosity (mPa.s)
Storage temperature (°C)	Storage temperature (°C)
	0	1	4	11	14	21	28
5	1.60 ± 0.06a	1.60 ± 0.01a	1.60 ± 0.00a	1.61 ± 0.04a	1.57 ± 0.01ab	1.52 ± 0.00b	1.55 ± 0.02ab
25	1.60 ± 0.06a	1.56 ± 0.01ab	1.53 ± 0.00bc	1.48 ± 0.01c	1.33 ± 0.00d	1.37 ± 0.00d	1.48 ± 0.01c
45	1.60 ± 0.06a	1.43 ± 0.01b	1.40 ± 0.01b	1.40 ± 0.01b	1.41 ± 0.01b	1.41 ± 0.01b	1.42 ± 0.01b

CONCLUSION

In conclusion, HPH treatment enhances the functionality of the pea and rice bran protein mixture effectively as an emulsifier for stabilizing palm kernel milk emulsion. The emulsion displays strong resistance towards coalescence at low temperatures and superior stability at high pH levels against ionic challenge.

Pea protein functions mainly by increasing the viscosity of the emulsion, slowing the Brownian motion of fat droplets, whereas rice bran protein functions as a surface‐active agent to lower the interfacial tension, making the system more thermodynamically stable. The difference in the routes of action of the two proteins significantly enhanced the stability of the emulsion when they were coupled together. Increasing homogenization intensity was observed to increase the emulsion stability by dispersing the oil droplet into tinier droplets and exposing the hydrophobic core of the protein to promote the attachment of the protein to the oil droplet surface. Nevertheless, at extreme homogenization intensity, excessive protein unfolding promoted protein aggregation that leads to phase separation.

The palm kernel emulsion showed decreasing stability as the pH value decreased from pH 11 to pH 3 as it was closer to the PI of the proteins presented in the system. The increase in ionic strength of the system further destabilizes the emulsion by neutralizing charges on the protein surface, inducing ionic aggregation of the protein. Nevertheless, the effect of ionic strength diminishes at higher pH presumably due to the high concentration of hydroxyl ions masking the surface charges of protein, preventing ionic binding. Yet, when ions with higher valency were introduced, destabilization was observed at all pH ranges, presumably due to the greater effect of divalent ions in disturbing the Debye screening length. For storage stability, phase separation happened after 10 days of storage at 45 °C whereas no significant separation was observed when emulsion was stored at 5 °C for 28 days. The main separation mechanisms that contributed to the separation were coalescence at high temperatures and Ostwald ripening and flocculation at low temperatures.

The present findings demonstrated that excessive protein impact emulsion stability, which could limit the development of high protein formulation beverages. The emulsion's sensitivity to acidic and salty conditions also limits its application. Future research could focus on potential pre‐treatment or protein hydrolysis to enhance protein solubility and emulsion stability. Utilizing acid and salt soluble protein fractions may also improve overall protein performance.

CONFLICT OF INTEREST

The authors declare no conflict of interest.

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request.