Impact of Three-Fluid Nozzle Emulsification on the Physicochemical and Thermodynamic Properties of Avocado Oil Microcapsules Obtained by Spray Drying

Featured Application

The findings highlight the potential of spray drying with a three-fluid nozzle as a strategy for microencapsulating oils. Future research could focus on its incorporation into different food matrices and on optimizing storage conditions and controlled release.

Abstract

This study investigated the production and characterization of avocado oil emulsions generated with a three-fluid nozzle (3FN) and the physicochemical and thermodynamic properties of the resulting microcapsules obtained by spray drying. The emulsions showed a bimodal size distribution with a main peak at 0.893 µm and PDI values below 0.70 indicate a mid-range polydispersity. Despite their shear-thinning behavior, emulsions exhibited limited stability, as indicated by ζ-potential (−23.9 mV) and increasing TSI values. Spray drying with 3FN achieved a yield of 71.7% and an encapsulation efficiency of 57.8%, with moisture content below 4%, meeting commercial requirements. The microcapsules displayed unimodal particle distributions (D[3,2] = 8.38 µm; D[4,3] = 11.14 µm) and irregular spherical morphologies with surface folds and roughness. Adsorption isotherms followed a type II pattern, well described by the GAB model, with monolayer moisture content (0.043–0.060 g H₂O/g solids) defining critical stability conditions. Thermodynamic analyses identified a “minimum entropy zone” corresponding to enhanced structural stability, while glass transition data confirmed that encapsulated oil did not act as a plasticizer. Overall, the use of a three-fluid nozzle enabled the development of avocado oil microcapsules with favorable physical and thermal attributes, supporting their potential for long-term stability in functional food applications.

1. Introduction

Microencapsulation has emerged as an effective technological strategy for preserving bioactive ingredients susceptible to degradation, particularly those present in oils and lipophilic compounds [1]. Avocado oil, recognized for its high nutritional value, attributable to its content of unsaturated fatty acids and bioactive compounds [2,3,4]. However, its composition makes it particularly vulnerable to oxidation induced by environmental factors such as light, oxygen, temperature, and pH, which limits its stability and application in the food industry [5].

In this context, spray drying has emerged as a highly efficient strategy for oil encapsulation, as it reduces susceptibility to oxidation and extends shelf life [6]. This technique relies on the entrapment of oil within a protective matrix, commonly referred to as the wall material. Among various encapsulating agents, proteins have been identified as high-performance matrices in spray-drying processes, owing to their functional properties as emulsion-stabilizing capacity and affinity for lipophilic compounds [7,8]. Within this group, whey protein has gained particular attention due to its ability to interact with both hydrophilic and hydrophobic compounds, as well as its high mass transfer rates, which facilitates uniform coating of oil droplets during the microencapsulation process [9].

For the production of microcapsules, wall and core materials are dissolved in an aqueous medium to form an emulsion, which is subsequently atomized through a nozzle into a drying chamber. During atomization, the emulsion is exposed to a stream of hot-air (150–220 °C), resulting in rapid dehydration and the formation of solid particles [10]. The emulsification step is crucial in the microencapsulation of oils, as the stability, composition, and characteristics of the emulsion significantly influence the properties of the resulting powder, including surface oil content, encapsulation efficiency, internal structure, oxidative stability, and various physical attributes [11,12,13]. This stage also largely determines the particle size of the final product, under the assumption that it remains unchanged throughout the process. However, during atomization, the high velocity of the injected gas generates shear and friction that fragment oil droplets, potentially compromising emulsion stability and, consequently, the characteristics of the final product [14,15,16]. In this context, the type and design of the atomizing nozzle constitute a critical factor.

Among the most commonly used pneumatic nozzles are two-fluid nozzles (2FNs), which feature concentric channels for compressed gas and preformed emulsion [17,18]. Their operation is based on the principle of pneumatic atomization, where high-velocity gas generates a pressure gradient and shear forces that break the liquid into fine droplets. The size of the atomized droplets depends on the relative velocity between phases, the orifice geometry, gas pressure, and gas to liquid ratio parameters that determine drying efficiency and the final particle size distribution. Their ease of operation, low cost, and compatibility with a wide range of formulations have made them the most widely used in the food and pharmaceutical industries. However, they require a prior emulsification step to ensure the stability of the feed liquid, which can increase process time and cost. In contrast, three-fluid nozzles (3FNs) feature an innovative configuration with one channel for compressed gas and two independent channels for feeding different materials, such as the core and the wall material. This configuration allows the emulsion to form in situ at the nozzle tip, eliminating the need for prior emulsification and offering potential advantages in terms of efficiency and cost [10].

Several studies have investigated the influence of nozzle type on the physical and functional properties of microcapsules. Tatar Turan et al. [19] evaluated the use of an ultrasonic nozzle versus a two-fluid nozzle (2FN) for the encapsulation of blueberry juice, observing that ultrasonic technology provided greater protection of phenolic compounds and improved storage stability. Similarly, Legako and Dunford [20] compared 2FN, 3FN, and a sonic atomizer in the encapsulation of fish oil using whey protein, concluding that nozzle design significantly affects encapsulation efficiency and particle morphology. Also, Cai et al. [10] reported that microcapsules produced using a 3FN exhibited greater uniformity and oil retention capacity compared to those obtained with a 2FN, using soy oil encapsulation systems with pectin and maltodextrin.

In contrast, other authors have focused on the effects of the atomization process on the stability of emulsions prior to drying. Taboada et al. [21] observed that during the atomization of whey protein-based emulsions, the size of oil droplets undergoes changes that can be modulated by the addition of low-molecular-weight emulsifiers. Similarly, Villalobos-Espinosa et al. [22] reported that atomization can alter droplet size distribution, which directly impacts emulsion stability and, consequently, the properties of the final product. In agreement with these findings, Munoz-Ibáñez et al. [23] investigated the influence of various atomization parameters on the particle size distribution of spray-dried emulsions, and proposed guidelines for selecting operating conditions that preserve specific droplet sizes, thereby allowing for greater control over the characteristics of the encapsulated product.

Despite the aforementioned advances, there remains limited scientific evidence specifically addressing the impact of 3FN use on emulsion stability and physicochemical properties of the resulting microcapsules, particularly those related to water vapor sorption and resulting thermodynamics parameters. This limited availability of studies restricts comprehensive understanding of the mechanisms determining encapsulation process efficiency and final product stability. It is crucial to explore complementary approaches that provide insight into the stability of encapsulated systems under varying conditions. In this regard, the thermodynamic relationships between water and food components play a fundamental role, as they allow estimating the energetic requirements of dehydration processes and prediction of optimal storage conditions that ensure physicochemical and microbiological stability of products [24]. Sorption isotherm analysis is a key tool for determining thermodynamic parameters such as differential enthalpy and entropy, which elucidate the nature and intensity of interactions between water and the encapsulating matrix and supports understanding adsorption dynamics and evaluating the hygroscopic stability of the system [25,26].

Considering the fundamental role of these factors in product quality and process design, the objective of this work was to characterize the stability of emulsions and the physicochemical and thermodynamic properties of microcapsules obtained by spray-drying, using a three-fluid nozzle (3FN). This will help us to further understand their physical attributes, discuss advantages of its use, and evaluate their influence towards the microencapsulation of lipophilic compounds.

2. Materials and Methods

2.1. Materials

Extra virgin cold-pressed avocado oil (Avocare^®, Poole, UK) was selected as the core component. Whey protein isolate with a protein content of 90% (WPI, Meli-Natura^®, Mexico City, Mexico) was used as the wall material.

2.2. Obtaining the Emulsion with a Three-Fluid Nozzle

To examine the influence of the atomization mechanism on the physicochemical properties of the precursor emulsions, the method described by Villalobos-Espinosa et al. [22], was employed. Briefly, a three-fluid nozzle (3FN, Büchi; compressed gas/internal liquid/external liquid) was used to generate the emulsions. To fully understand the operating principle of the 3FN as compared to that of the two fluid-device (2FN), in Figure 1, diagrams of both nozzle types are presented.

In situ emulsification at the nozzle tip was achieved by simultaneously feeding the oil and aqueous phases through independent channels. Specifically, avocado oil was delivered using a syringe pump at a flow rate of 0.10 mL/min, while the 20% (w/w) WPI solution was supplied using a peristaltic pump at 2.78 mL/min, maintaining a 1:6 ratio.

The atomization pressure was set at 3.5 bar. An aerosol formed by emulsion particles in an air stream was produced at the tip of the 3FN. Emulsions were collected in a plastic container at a fixed distance of 30 cm from the nozzle tip.

2.3. Physicochemical and Rheological Characterization of Emulsions

2.3.1. Droplet Size (DS), Polydispersity Index (PDI), and Zeta Potential (ζ-Potential)

DS, PDI, and ζ-potential of the emulsions were determined by using a Zetasizer Nano ZS analyzer (Malvern Instruments, UK). Prior to analysis, the collected samples were diluted in deionized water (0.01%, v/v) to minimize multiple scattering effects and ensure optimal measurement conditions [22,27].

2.3.2. Emulsion Stability Analysis Using the Turbiscan Stability Index (TSI)

Emulsion stability was evaluated by using a Turbiscan^® Lab analyzer (Formulation Smart Scientific Analysis, L’Union, France). After preparation, 20 mL aliquots of each emulsion were transferred into specific glass cells for analysis. Samples were monitored over a 4 h period at regular intervals.

TSI values were obtained using Equation (1) established by Banasaz et al. [28]:

$$\mathrm{T}\mathrm{S}\mathrm{I}=\sqrt{{\displaystyle \frac{{\sum }_{\mathrm{i}=1}^{\mathrm{n}}{\left({\mathrm{X}}_{\mathrm{i}}-{\mathrm{X}}_{\mathrm{B}\mathrm{S}}\right)}^2}{\mathrm{n}-1}}}$$

(1)

where $X_i$ is the average backscattering at each min of measurement, $X_{BS}$ is the mean of $X_i$, and n is the number of scans. Higher TSI values indicate greater overall variation in backscattering over time and, consequently, greater emulsion destabilization. In contrast, values close to zero reflect minimal variation and thus higher physical stability of the emulsion.

2.3.3. Kinetic Modeling of TSI Data

To describe the destabilization kinetics of the emulsions, the evolution of the TSI over time was modeled using a two-parameter empirical equation (Equation (2)), reported by Villalobos-Espinosa et al. [22]. This equation assumes a saturating behavior over time, where the TSI approaches a theoretical maximum instability:

$$\mathrm{T}\mathrm{S}\mathrm{I}={\displaystyle \frac{m_{1}t}{m_2+t}}$$

(2)

Within the applied kinetic model, TSI represents the stability index at time t, $m_1$ indicates the theoretical maximum instability limit as time approaches infinity (t_∞), and $m_2$, defined as the time required for the system to reach half of its maximum instability (TSI/TSI_max = 0.5), serves as a kinetic parameter characterizing the rate of destabilization.

2.3.4. Rheological Characterization

The rheological characterization of the emulsions was carried out using a HR20 rheometer (TA Instruments, Montreal, QC, Canada), following the methodology described by Lopez-Hernández et al. [29]. A plate–plate measurement system (60 mm diameter) was used, and tests were performed at a controlled temperature of 25 ± 2 °C. The flow behavior of the emulsions was evaluated over a shear rate range from 0.01 to 100 s⁻¹.

2.4. Preparation of Microcapsules by Spray-Drying

2.4.1. Spray Drying with a Three-Fluid Nozzle (3FN)

The drying process was carried out using a laboratory-scale spray dryer (Büchi B-290), following the methodology described by Cai et al. [10]. Briefly, it was operated at an inlet temperature of 180 ± 2 °C, with 100% aspiration and an atomization pressure of 3.5 bar. The feeding of the samples to the nozzle was carried out according to the procedure described in Section 2.2. After drying, the resulting powders were hermetically packed in vacuum-sealed bags for storage until the corresponding analyses were performed.

2.4.2. Process Yield

The yield (%) of the spray-drying process was quantified using Equation (3), which defines it as the percentage of total solids recovered in the final powder relative to the amount of solids initially present in the atomized emulsion [30,31].

$$\mathrm{Yield}(\%)={\displaystyle \frac{\mathrm{Mass} \mathrm{of} \mathrm{collected} \mathrm{solids}}{\mathrm{Injected} \mathrm{solid} \mathrm{mass}}} \times 100$$

(3)

2.4.3. Encapsulation Efficiency

The encapsulation efficiency (EE) was determined by following the methodology described by Altuntas et al. [32]. Briefly, two grams of microencapsulated powder was suspended in 20 mL of hexane. The mixture was stirred for 10 min at room temperature to facilitate the extraction of surface oil. Subsequently, the suspension was filtered using Whatman No. 1 filter paper, and the residue was washed with hexane until reaching a final volume of 25 mL. The solvent was then removed by evaporation in an oven at 60 °C for 12 h. The mass of surface oil was determined gravimetrically as the difference in the flask’s weight before and after solvent evaporation. The EE value was calculated using Equation (4).

$$\mathrm{E}\mathrm{E}={\displaystyle \frac{\mathrm{T}\mathrm{O}-\mathrm{S}\mathrm{O}}{\mathrm{T}\mathrm{O}}}\times 100$$

(4)

where TO represents the theoretical total oil content (g), and SO corresponds to the surface oil content associated with the non-encapsulated fraction (g).

2.5. Characterization of Spray-Dried Microcapsules

2.5.1. Moisture Content

The moisture content of the microcapsules was determined gravimetrically according to the method described by Bajac et al. [33]. One gram of sample was placed in a convection oven (9053A, ECOSHEL, Covington, KY, USA) at 105 °C until constant weight, and the moisture content was calculated based on the weight loss.

2.5.2. Characterization of Particle Size and Morphology

Particle size was determined using laser light scattering, employing a Mastersizer 2000 instrument (Malvern Instruments Ltd., Malvern, UK) with the SIROCCO 2000 dry dispersion unit [34]. The microcapsules were dispersed at 65% vibration intensity and an air pressure of 3.75 bar. Morphology of the microcapsules was evaluated using a dual-beam FEI Quanta 3D FEG system (SEM/FIB, Thermo Fisher Scientific, Waltham, MA, USA) equipped with a field emission gun (FEG) at an acceleration voltage of 5 kV. The microcapsules were mounted on cylindrical holders using double-sided conductive carbon tape to ensure proper electrical contact [35]. A working distance of 10.1 mm was used, operating in low-vacuum mode to preserve the surface characteristics of the samples.

2.5.3. Evaluation of Powder Flow Function

The flow function of the microcapsules was evaluated by using a Powder Flow Tester (Brookfield PFT, Middleboro, MA, USA) according to the methodology of Deng et al. [36]. Axial and rotational speeds were set at 1.0 mm/s and 1 rev/h, respectively, and tests were conducted at room temperature (~20–25 °C). The equipment was automated using Powder Flow Pro v1.3, USA software, which provided the unconfined failure strength (UFS) as a function of the major principal consolidated stress (MPCS). UFS indicates the force required for the sample to yield under a specific stress, reflecting its cohesion or tendency to cake, while MPCS corresponds to the pressure applied to compact it.

2.5.4. Moisture Sorption Isotherms

Moisture sorption isotherms were determined by the gravimetric method described by Lang et al. [37], conducted at controlled temperatures of 25, 35, and 45 °C. Briefly, triplicate samples of 1.00 g each were placed in hermetically sealed chambers containing saturated salt solutions capable of maintaining relative humidity levels ranging from 11% to 85% [38]. The samples were periodically weighed until equilibrium was reached, which was defined as a change in mass of less than 0.001 g of water per g of dry solids between two consecutive measurements.

2.5.5. Modeling of Water Sorption Isotherms

The relationship between the equilibrium moisture content and water activity was described using the GAB model [39], as presented in Equation (5):

$$M={\displaystyle \frac{M_{0}CK{a}_w}{(1-K{a}_w)(1-{Ka}_w+CK{a}_w)}}$$

(5)

where M denotes the moisture content (g of water per g of dry solids), while M₀ corresponds to the monolayer moisture content. The variable $a_w$ represents the water activity of the system; C is a dimensionless parameter related to the sorption heat in the monolayer region, and K is a parameter associated with the sorption energy in the multilayer region.

2.5.6. Determination of Thermodynamic Parameters

Differential Properties

The changes in differential enthalpy at the water–solid interface during the various stages of adsorption were calculated using the Othmer equation (Equation (6)) [40]:

$$ln{P}_V={\left({\displaystyle \frac{H_v\left(T\right)}{H_v^0\left(T\right)}} \right)}_M ln\left(P_V^0\right)+C_1$$

(6)

where $H_v\left(T\right)$ is the isosteric heat of water adsorption, $H_v^0\left(T\right)$ is the heat of condensation of pure water, M is the moisture content, and C₁ is the adsorption constant. Plotting lnP_V against $ln\left(P_V^0\right)$ yields a straight line if the ratio $H_v\left(T\right)/H_v^0\left(T\right)$ remains constant within the studied temperature range.

The net isosteric heat or differential enthalpy ${\left(∆H_{diff}\right)}_T$ is defined by Equation (7).

$${\left(∆H_{diff}\right)}_T={\left({\displaystyle \frac{H_v\left(T\right)}{H_v^0\left(T\right)}}-1\right)}_M\left(H_v^0\right)$$

(7)

Using the values obtained from the estimation of enthalpy changes, the differential entropy (∆S_dif) can be calculated through Equation (8).

$$(∆S_{dif})=S_1-S_L={\displaystyle \frac{-{\left(∆H_{dif}\right)}_T-RTln\left(a_w\right)}{T}}$$

(8)

where S₁ = (∂S_t)/(∂N₁)_T,P is the differential molar entropy of water adsorbed in the food; $S_L$ is the molar entropy of pure water in equilibrium with the vapor; S is the total entropy of water adsorbed in the food; N₁ is the number of moles of water adsorbed in the food; R is the universal gas constant; $a_w$ is the water activity; and T is the temperature (K).

Integral Properties

The integral molar enthalpy ${\left(∆{\mathrm{H}}_{\mathrm{i}\mathrm{n}\mathrm{t}}\right)}_{\mathrm{T}}$, is calculated by using an expression analogous to that employed for the differential enthalpy (Equation (9)), while maintaining constant diffusion pressure (Φ):

$${\left(∆{\mathrm{H}}_{\mathrm{i}\mathrm{n}\mathrm{t}}\right)}_{\mathrm{T}}={⌊{\displaystyle \frac{{\mathrm{H}}_{\mathrm{v},\mathrm{i}\mathrm{n}\mathrm{t}}\left(\mathrm{T}\right)}{{\mathrm{H}}_{\mathrm{v}}^0\left(\mathrm{T}\right)}}-1⌋}_{\mathsf{\phi }}{\mathrm{H}}_{\mathrm{v}}^0\left(\mathrm{T}\right)$$

(9)

where ${\left(∆{\mathrm{H}}_{\mathrm{i}\mathrm{n}\mathrm{t}}\right)}_{\mathrm{T}}$ is the integral molar heat of water adsorbed in the food, and (Φ) can be calculated using Equation (10) or Equation (11).

$$\phi ={\mu }_{ap}-{\mu }_a=RT{\displaystyle \frac{W_{ap}}{W_v}}{\int }_0^{a_w}Xd ln{a}_w$$

(10)

$$\phi ={\alpha }_{1}T{\int }_0^{a_w}Xd ln{a}_w$$

(11)

where µ_a is the chemical potential of the adsorbate in the condensed phase, µ_ap is the chemical potential of the adsorbent, W_ap is the molecular weight of the adsorbent, and W_v is the molecular weight of water; the constant φ/α₁ is analogous to the process at constant φ. The heat of condensation of pure water, ${\mathrm{H}}_{\mathrm{v}}^0\left(\mathrm{T}\right)$, was determined using Equation (12), as established by Wexler [41].

$$H_v^0 \left(T\right)=6.15\times {10}^4-{10}^4-94.14 T+17.74\times {10}^{-2}{T}^2-2.03\times {10}^{-4}{T}^3$$

(12)

Once the values for ${\left(∆H_{int}\right)}_T$ are obtained, the changes in integral molar integral entropy ${(∆{\mathrm{S}}_{\mathrm{int}})}_{\mathrm{T}}$ can be calculated using Equation (13):

$${(∆{\mathrm{S}}_{\mathrm{int}})}_{\mathrm{T}} = {\mathrm{S}}_{\mathrm{int}}-{\mathrm{S}}_{\mathrm{L}} = {\displaystyle \frac{-{\left(∆{\mathrm{S}}_{\mathrm{int}}\right)}_{\mathrm{T}}-\mathrm{RTln}\left({\mathrm{a}}_{\mathrm{w}}\right)}{\mathrm{T}}}$$

(13)

2.5.7. Compensation Theory

According to the enthalpy–entropy compensation theory, there is a linear relationship between enthalpy and entropy. This relationship is expressed through the compensation law, which correlates the values of ${∆\mathrm{H}}_{\mathrm{int}}$ and ${∆\mathrm{S}}_{\mathrm{int}}$ [42], as shown in Equation (14):

$${∆\mathrm{H}}_{\mathrm{int}} ={ \mathrm{T}}_{\mathsf{\beta }}{∆\mathrm{S}}_{\mathrm{int}} + {∆\mathrm{G}}_{\mathsf{\beta }}$$

(14)

where $T_{\beta }$ is the isokinetic temperature, corresponding to the condition under which all reactions occur at the same rate; ${∆\mathrm{S}}_{\mathrm{int}}$ represents the integral entropy, and ${∆\mathrm{G}}_{\mathsf{\beta }}$ corresponds to the Gibbs free energy at the isokinetic temperature.

The verification of the compensation theory is performed by comparing the isokinetic temperature (${\mathrm{T}}_{\mathsf{\beta }}$) with the harmonic mean temperature (${\mathrm{T}}_{\mathrm{hm}}$), which is determined using Equation (15) [43]:

$${\mathrm{T}}_{\mathrm{hm}} = {\displaystyle \frac{\mathrm{N}}{{\sum }_1^{\mathrm{N}}(1/\mathrm{T})}}$$

(15)

where N is the total number of isotherms used.

2.5.8. Glass Transition Temperature (Tg)

The glass transition temperature (Tg) of the microcapsules equilibrated at a water activity of 0.33 was determined by differential scanning calorimetry (DSC, Discovery, TA Instruments, New Castle, DE, USA), using a nitrogen purge rate of 50 mL/min. Approximately 3 mg of sample were placed in 20 µL aluminum pans, hermetically sealed, with an empty pan used as reference. Analyses were performed in triplicate over a temperature range from −80 to 160 °C, at a heating rate of 5 °C/min. Tg values were obtained from the midpoint of the change in heat flow recorded in the thermograms [44].

2.6. Statistical Analysis

The experiments were conducted in triplicate, and the results are reported as mean ± standard deviation. Statistical analysis was performed using Microsoft Excel (Microsoft Corp., Seattle, WA, USA) and Minitab software, version 19 (Minitab Inc., State College, PA, USA). Water sorption isotherm modeling was carried out with KaleidaGraph 4.5 Synergy software.

3. Results and Discussion

3.1. Effect of Nozzle Type on Droplet Size (DS), Polydispersity Index (PDI), and ζ-Potential

Figure 2 illustrates the droplet size distribution of avocado oil emulsions stabilized with whey protein isolate (WPI), obtained using a three-fluid nozzle (E3FN). The system exhibited a bimodal distribution, characterized by a predominant peak in the micrometric range, with an average diameter of 0.89 µm, and a secondary fraction consisting of droplets smaller than 0.20 µm. PDI values below 0.70 (

Table 1. Average droplet size, peak intensity (%), ζ-potential, and PDI of avocado oil emulsions with WPI obtained using three-fluid nozzle (3FN). Values represent the mean of three replicates.

Droplet Size (µm)
Sample	Average Size (µm)	Peak 1	Peak 2	ζ-Potential (mV)	PDI
E3FN	0.89 ± 24.03	0.89 (78.2)	0.15 (21.8)	−23.90 ± 1.69	0.63 ± 0.04

) indicate a mid-range polydispersity [45].

The particle sizes obtained were comparable to those reported by Espinosa-Solis et al. [3], who reported ranges of 0.36–0.85 µm in Pickering emulsions based on avocado oil, prepared using different emulsification methods. This similarity can be attributed to the fact that the 3FN configuration promotes in situ emulsification at the nozzle tip, reducing residence time and, consequently, pre-atomization instability phenomena, which favors the formation of smaller droplets [17,46]. These findings reinforce that both nozzle geometry and operating conditions are key factors in determining droplet size distribution during the atomization process [22,47,48].

Regarding the ζ-potential, the evaluated emulsions showed an average value of −23.90 ± 1.69 mV, which suggests moderate stability, according to the range proposed by Cano-Sarmiento et al. [49]. The ζ-potential reflects the contribution of electrostatic repulsion to emulsion stability; in this regard, a higher absolute value of the surface charge promotes the generation of repulsive forces between particles, thereby preventing their coalescence [50].

3.2. Description of Emulsion Stability Using TSI and Its Modeling

Figure 3 shows the evolution of the Turbiscan Stability Index (TSI) of avocado oil emulsions obtained using the three-fluid nozzle (3FN) during 120 min of monitoring. During the first 60 min, the emulsions exhibited an initial destabilization kinetics with a TSI value of 1.75, which subsequently increased progressively, reaching 2.25 at the end of the test. This behavior indicates a gradual instability attributed to classical phase separation mechanisms, such as creaming, sedimentation, flocculation, and coalescence, characteristic of emulsions [51].

The fitting of the experimental data to the two-parameter asymptotic kinetic model (Equation (2)) was highly satisfactory, with a coefficient of determination of R² = 0.999 and a root mean square error (RMSE) of 0.002 (

Table 2. Model fitting results for TSI monitoring over a two-hour period. (E3FN) Emulsion produced using a three-fluid nozzle.

Emulsions	Parameters of the Model
	m1	m2 (min)	R2	RMSE
E3FN	3.15	48.14	0.999	0.002

), confirming the suitability of this model to describe the observed destabilization dynamics. Within this framework, parameter m₁, representing the theoretical maximum instability, reached a value of 3.15, while m₂, defined as the time required to reach 50% of such instability, was 48.14 min.

The moderate magnitude of m₁ and the intermediate rate associated with m₂ indicate that the emulsions obtained with the 3FN exhibit a relative stability that can be considered adequate. This result suggests that the atomization conditions play a decisive role in droplet interactions and in the kinetic evolution of the system, limiting accelerated coalescence processes and promoting resistance against destabilization [15,21]. Taken together, this evidence supports the ability of the three-fluid nozzle to produce emulsions with good stability, providing a solid basis for its consideration as a promising strategy in the design of more robust and functional emulsion systems. The experimental TSI values were fitted to the two-parameter asymptotic kinetic model described in Equation (2). The fit was outstanding, with a determination coefficient of R² = 0.999 and a root mean square error (RMSE) of 0.002 (Table 2), confirming its suitability for representing the process kinetics.

3.3. Rheological Characterization of the Emulsions

Figure 4 shows the apparent viscosity (η) as a function of the shear rate ($\stackrel{.}{\gamma }$, 0.1–100 s⁻¹) for the emulsion obtained using the three-fluid nozzle (E3FN). Maximum η values were observed at low shear rates, which progressively decreased as $\stackrel{.}{\gamma }$; increased. This behavior is characteristic of pseudoplastic or shear-thinning fluids.

This rheological behavior has been widely reported in oil-in-water (O/W) emulsions [52,53,54], and is commonly attributed to the alignment of polymer chains in the direction of flow, which reduces the system’s internal resistance [55]. Moreover, shear-thinning may also be associated with the disruption of the three-dimensional (3D) network formed by the oil droplets, as hydrodynamic forces promote directional rearrangement and decrease flow resistance [53].

Fitting the experimental data to the power-law (Ostwald–de Waele) model yielded correlation coefficients above 0.9 (

Table 3. Fitting parameters of the power-law model for the rheological curves of emulsions obtained using three-fluid nozzle (E3FN).

Samples	K (Pa·sn)	n	R2
E3FN	0.02 ± 0.003	0.82 ± 0.03	0.997

), supporting the suitability of this model to describe the rheological behavior of the emulsions studied. In all cases, the flow index (n) was below unity, confirming the pseudoplastic nature of the emulsions [56]. Regarding the consistency coefficient (K), relatively low values were obtained, indicating a limited capacity of the system to increase its viscosity [57]. The analysis of rheological properties is essential to predict the stability and functional performance of emulsions in various technological applications [58], where apparent viscosity constitutes a critical parameter, reflecting both the macroscopic appearance and the system’s structure [57].

The progressive decrease in viscosity with increasing shear rate reflects the system’s ability to reorganize and dissipate energy through molecular alignment and partial disruption of the three-dimensional network formed at rest. The good fit to the Ostwald–de Waele model confirms that this behavior follows typical non-Newtonian fluid laws, where the reduction in K indicates lower internal resistance and, consequently, a more fluid consistency under applied stress. This suggests that the internal structure is organized yet flexible, ensuring stability at rest and good processability under flow. Such behavior is consistent with that reported for O/W emulsions and represents a functional advantage for technological applications requiring a balance between stability and ease of handling [13].

3.4. Encapsulation Efficiency and Yield

The previously characterized emulsions were subjected to spray drying (Ti = 180 °C), and the physical properties of the resulting microcapsules, including process yield, encapsulation efficiency (EE), moisture content, and particle size, are presented in

Table 4. Properties of avocado microcapsules obtained by spray drying using three-fluid nozzle (3FN).

Properties	3FN
Yield (%)	71.73 ± 1.59
Encapsulation efficiency (%)	57.83 ± 0.07
Moisture (% db)	3.63 ± 0.41

and

Table 5. Particle size distribution of avocado oil microcapsules obtained by spray drying (SD) using three-fluid nozzle (3FN).

	Particle Diameter (um)		Volume Diameter (um)
	D[3,2]	D[4,3]	d10	d50	d90
3FN	8.38 ± 0.17	11.15 ± 0.18	4.73 ± 0.10	9.76 ± 0.21	19.55 ± 0.27

. The microcapsules obtained by spray drying (SD) with the three-fluid nozzle (3FN) exhibited a process yield of 71.72% and an EE of 57.82%.

When contrasting these results with those reported in the literature, significant differences are evident, attributable to both the formulation and process design. Legako et al. [20] reported encapsulation efficiencies of 83.3% and 85.8% for microcapsules produced by SD using 3FN and 2FN, respectively, employing fish oil as the core and 20% (w/w) whey protein isolate (WPI) as the wall material, with a core-to-wall ratio of 1:2. Similarly, Bae and Lee [59] encapsulated avocado oil using 20% WPI with SD in a 2FN configuration, achieving an EE of 52.62%, a value even lower than that obtained in the present study. More recently, Cai et al. [10] reported efficiencies close to 90% in the encapsulation of soybean oil by SD with 2FN and 3FN, attributing this high performance to the nature of the wall materials used, consisting of blends of maltodextrin and sugar beet pectin, whose high viscosity favored rapid formation of the protective matrix and, consequently, greater core retention. These findings confirm the determining influence of the functional properties of the encapsulating agents on process efficiency.

The relatively low efficiency obtained in the present study (57.82%) can be explained by the drying dynamics inherent to protein-based systems. During spray drying, the rapid evaporation of the aqueous phase progresses faster than the internal diffusion of solutes, leading to surface enrichment of the oil [60]. Moreover, in smaller particles, the increase in specific surface area intensifies this effect, facilitating core migration to the surface and reducing encapsulation efficiency [61,62]. This effect is further stressed by the three-fluid nozzle configuration, in which the core and wall material mix only at the tip of the atomizer; therefore, the initial coverage of the oil depends strictly on immediate mixing, which dictates the final distribution of the components and limits the retention efficiency [63].

Regarding the process yield, values of 71.72% were achieved. These values fall within the range reported in various studies on oil microencapsulation by spray drying, in which yields have been reported to vary between 40% and 88%, depending on the operational conditions and the nature of the wall materials used [30,64,65].

With respect to moisture content, the microcapsules exhibited a value of 3.6%. This value is below 4%, not only meeting the range established for commercial products obtained by spray drying (3–6%), but also favoring physicochemical stability during prolonged storage by reducing the likelihood of undesirable phenomena such as oxidation of the encapsulated compounds and microbial growth [33,66,67].

The final application of microcapsules in food matrices critically depends on properties such as quality, appearance, and powder flow, which are in turn strongly influenced by the size and distribution of the particles comprising the microcapsules [68]. In the present study, the microcapsules obtained by spray drying with 3FN showed average Sauter (D[3,2]) and Brouckere (D[4,3]) diameters of 8.38 and 11.14 µm, respectively (Table 5). Additionally, the d₁₀, d₅₀, and d₉₀ percentiles confirmed a unimodal and relatively narrow distribution (Figure 5), indicating that most particles are concentrated within a reduced micrometric range.

When contrasting these findings with the literature, Zhu et al. [65] reported D[3,2] and D[4,3] values of 5.93 and 26.29 µm for soybean oil microencapsulation, with a maximum distribution peak close to 30 µm, higher than that recorded in the present study (~9 µm). In contrast, Bajac et al. [33] reported a D[4,3] of 9.59 µm when encapsulating juniper berry essential oil, a value closer to that found in this study.

From a phenomenological perspective, the microcapsules obtained with the 3FN nozzle exhibited a size distribution concentrated in the lower micrometer range, with a d₅₀ close to 10 µm and a d₉₀ around 20 µm. This unimodal and uniform distribution indicates that the atomization and solidification process promoted the formation of spherical particles of intermediate size, avoiding both ultrafine fractions and large agglomerates. Such behavior enhances the powder’s flowability, stability, and dispersibility, increasing its technological functionality in food matrices. Overall, these attributes represent an advantage in manufacturing efficiency, sensory quality, and acceptance of the final product [33,68,69].

The morphology of the microcapsules was analyzed by SEM (Figure 6). The micrographs revealed that the particles exhibited morphological features characterized by predominantly deformed spheres, with rough surfaces, pronounced folds, and irregular protrusions. These surface deformations are mainly associated with conditions inherent to the drying process, particularly the high temperature and rapid moisture loss during the initial stages, which induce stresses in the polymeric matrix and cause partial wall collapse [65,70]. Additionally, the formation of agglomerates could be attributed to moisture reabsorption from the environment, which increases adhesive interactions between particles [71].

3.5. Powder Rheology

The results showed that the capsules exhibit a relatively cohesive behavior, with UFS (unconfined failure strength) values of 1.65 and 9.0 kPa at major principal consolidated stresses (MPCS) of 2.57 and 21.50 kPa, respectively. These values indicated that the powder tends to agglomerate, hindering free flow under low stresses, while flowing more easily under high stresses. Similar trends have been reported for magnesium stearate, a lubricating additive commonly used in food applications to prevent caking of powders subjected to high stresses [37]. Protein-rich powders are particularly prone to developing particle–particle interactions, which increase cohesion forces and, consequently, reduce their flowability. The obtained data allow for the interpretation of flow functions at different consolidation stresses and the classification of powders according to various flow regimes, ranging from free-flowing to cohesive.

Additionally, the high Tg (140.34 °C) observed (Section 3.9), indicated that obtained powder particles will not stick due to thermal effects, which aids in preserving flowability characteristics of emulsion produced by the 3FN.

3.6. Adsorption Isotherms and Thermodynamic Properties

Figure 7 shows the adsorption isotherms corresponding to avocado oil microcapsules with WPI obtained by spray drying using a three-fluid nozzle (3FN) at 25, 35, and 45 °C. The results show that, at a constant temperature, the moisture content of the powders increases as the water activity (a_w) rises, whereas for a given a_w, the moisture content decreases with increasing temperature, an effect attributed to the higher kinetic energy of water molecules [72]. The curves exhibited the characteristic sigmoidal shape of type II isotherms, typical of non-porous or mesoporous powder systems with weak water–matrix interactions [73]. Modeling using the GAB equation (

Table 6. Parameters obtained from the fitting of the GAB models to the water adsorption isotherms of microcapsules stored at 25 °C, 35 °C, and 45 °C, obtained by spray drying (SD) with three-fluid nozzle (3FN).

	GAB	Temperature (°C)
		25	35	45
3FN	M0	0.06	0.05	0.04
	C	8.87	7.291	7.965
	K	0.81	0.87	0.90
	R	0.997	0.998	0.991
	E (%)	0.444	0.435	0.007

) yielded a statistically satisfactory fit (R² values close to unity and E < 5%), confirming the suitability of this model to describe adsorption in the microcapsules. The monolayer moisture content (M₀) ranged from 0.04 to 0.06 g H₂O/g dry solids, corresponding to a_w values between 0.18 and 0.37. This parameter constitutes a key indicator of active sorption sites, as it quantifies strongly adsorbed water, whose presence sets a critical threshold beyond which water mobility decreases and, consequently, the progression of degradative, oxidative, and enzymatic reactions is limited [74,75,76].

These findings indicate that the three-fluid nozzle generates particles with a surface and internal structure of low affinity for water, which explains the decrease in moisture content with increasing temperature and the observation of type II isotherms. The consistency of this behavior is further supported by the decreasing trend of the C parameter, which confirms that the thermal energy acquired weakens the interactions between water and the active sites of the matrix, reducing the stability of adsorbed water and the equilibrium moisture content. This effect is associated with changes in the balance between polar and hydrophobic groups in the matrix, limiting water retention and decreasing the overall hygroscopicity of the system [5,72]. Meanwhile, K values close to unity indicate that the water molecules in the multilayer behave similarly to liquid water, marking the transition between strongly bound water and more mobile water [77,78]. Taken together, the combination of M₀, C, and K not only explains the adsorption dynamics at the molecular level but also provides strategic information for assessing the technological performance of the three-fluid nozzle, whose ability to produce microcapsules with low hygroscopicity and greater moisture resistance constitutes a decisive advantage in designing stable powders with extended shelf life.

3.7. Thermodynamic Properties

Figure 8 shows the enthalpy and entropy responses as a function of moisture content in avocado oil microcapsules obtained by spray drying with a three-fluid nozzle (3FN) at 25, 35, and 45 °C. The estimates of differential enthalpy (ΔH_dif) and integral enthalpy (ΔH_int), derived from the GAB model, serve as indicators of the intensity of water–matrix interactions, with the former describing the pointwise energetic variation and the latter representing the average energetic behavior [79].

An initial increase in ΔH_dif and ΔH_int was observed with rising moisture content, reaching a maximum at ≈0.05 g H₂O·100 g⁻¹, followed by a progressive decrease. Within this range, the microcapsules exhibited values between −18.73 and −18.37 J·mol⁻¹ (ΔHdif) and between −12.94 and −12.69 J·mol⁻¹ (ΔH_int), reflecting intense energetic interactions at low moisture content, associated with a higher water–matrix affinity and the greater energy required to desorb firmly adsorbed molecules.

It is noteworthy that the maximum value of ΔH_int closely coincided with the estimated monolayer moisture content, confirming that at low moisture levels, water binds to high-energy active sites, whereas increasing moisture induces the saturation of these sites and adsorption at less energetic positions, favoring multilayer formation [74,80]. The point of maximum water–matrix interaction was located at the intersection of ΔH_dif and ΔH_int, at ≈ 0.06 g H₂O·100 g⁻¹, coinciding with the minimum integral entropy. This phenomenon, previously described in chayote juice and canola oil [79,81], is associated with the transition from specific interactions at primary sites toward less energetic associations at secondary sites.

Figure 9 shows that the increase in moisture content led to a decrease in integral entropy until reaching a minimum (≈ −32 J·mol⁻¹), coinciding with the crossover point with differential entropy, identified as the region of maximum system stability. At this point (≈ 0.06 g H₂O·100 g⁻¹; aw ≈ 0.3), which corresponds to the theoretical monolayer value, water molecules exhibited a higher degree of structural order and lower availability to participate in degradative reactions [82]. Moreover, a “minimum entropy zone” was identified, ranging from 0.05 to 0.09 g H₂O·100 g⁻¹, within which the system would preserve its structural and functional stability during storage [79,82,83].

Microcapsules obtained with the three-fluid nozzle exhibit an adsorption pattern characterized by initial high-energy interactions, followed by adsorption processes at secondary, less intense sites, which explains the decreasing trend of ΔH_dif and ΔH_int after reaching their maximum values. The coincidence between the points of minimum entropy and the theoretical monolayer value indicates that the system achieves its maximum thermodynamic stability within this range, where adsorbed water remains strongly immobilized with a higher degree of structural order. This finding is particularly relevant, as it allows the identification of a critical moisture window in which the microcapsules retain their physicochemical and functional integrity, reducing susceptibility to degradative processes. From a technological perspective, defining this “minimum entropy zone” provides a strategic criterion for the design of storage and preservation conditions, while also validating the effectiveness of the three-fluid nozzle in producing microcapsules with energetic properties that favor the long-term stability of avocado oil powder.

3.8. Enthalpy–Entropy Compensation

Figure 10 shows the enthalpy–entropy compensation of avocado oil microcapsules obtained by spray drying with a three-fluid nozzle (3FN) at 25, 35, and 45 °C. This analysis serves as an essential tool to determine whether water vapor adsorption is governed by entropic or enthalpic effects, thereby elucidating the a_w range in which the system is most stable [74]. The applicability of the compensation theory was confirmed by the observation that the isokinetic temperature (T_B) differed significantly from the harmonic mean temperature (T_hm = 307.93 K), thus validating the model for the studied system.

The results revealed the existence of two clearly differentiated linear domains: T_B1, associated with low water activities (a_w = 0.05–0.29, 0.06–0.38, and 0.06–0.40), and T_B2, linked to high water activity intervals (a_w = 0.29–0.85, 0.40–0.89, and 0.48–0.99), both determined at the three evaluated temperatures. The T_B1 region is predominantly governed by entropic control (T_B1 = 124.43 ± 3.90 K < T_hm = 307.93 K), whereas the T_B2 region is characterized by enthalpic dominance (T_B2 = 426.70 ± 3.50 K > T_hm = 307.93 K). The transition point between both regions coincided with the minimum integral entropy (≈ −32 J·mol⁻¹, a_w ≈ 0.33), suggesting a change in the water–matrix interaction mechanism. Overall, although the microcapsules exhibited entropic and enthalpic domains of comparable magnitudes, the greater adsorption observed in the enthalpic region highlights the decisive role of energy interactions in the hygroscopic stability of the system [82,83]

The results provided clear evidence that the hygroscopic stability of the microcapsules cannot be interpreted solely from a quantitative perspective, but rather depends on the nature of the thermodynamic forces involved. The entropic domain reflects a structural response governed by conformational constraints, whereas the enthalpic domain corresponds to stronger, purely energetic molecular interactions [79,80]. The identification of this biphasic behavior, also reported in other protein encapsulation systems [81], suggested that it is a generalizable phenomenon in complex bioactive matrices. In this context, the dominant role of enthalpic interactions emerges as the main controlling factor of hygroscopic stability, positioning this parameter as a key criterion in the design and evaluation of encapsulating materials.

3.9. Glass Transition Temperature

Figure 11 shows the glass transition temperature (Tg) of the microcapsules and the control (WPI at 20% solids). The WPI exhibited a Tg of 137.55 °C, while the microcapsules recorded a slightly higher value of 140.34 °C. These results are consistent with previous studies, where Tg values close to 111 °C were reported for WPI when analyzing the relationship between protein stability, water content, and glass transition [84].

The incorporation of encapsulated oil did not significantly modify the Tg of the microcapsules, suggesting that the segmental mobility of the protein matrix is mainly governed by the protein and its monolayer-associated moisture. Consequently, the oil does not appear to play a relevant plasticizing role within the structure [85]. It is noteworthy that both samples were conditioned at an a_w of 0.33, corresponding to the monolayer moisture content determined from adsorption isotherms (Section 3.5). Thus, the observed differences in Tg can be primarily attributed to compositional and structural effects derived from the incorporation of avocado oil and the protein mobility restrictions induced during microencapsulation. In agreement, Foster et al. [86] reported similar monolayer values (0.06 g H₂O/g dry solids) for WPI with a comparable protein content (90%), reinforcing the hypothesis that adsorption in the studied microcapsules is strongly determined by the wall material.

The above is also supported from a phenomenological perspective, which indicates that the slight increase in the Tg of the microcapsules reflects additional structural constraints induced during the drying process, in which the protein matrix retains its dominant role in the glass transition, while the oil phase remains confined without significantly altering the overall system dynamics. The relevance of this finding lies in the fact that, in general, even a modest increase in Tg contributes to prolonging the stability of the material against processes that intensify above this threshold, such as stickiness, structural collapse, lipid oxidation, or degradation reactions. Consequently, microencapsulation not only protects the oil but also modifies the molecular dynamics of the protein system, reinforcing the thermal and functional stability of the microcapsules [87].

4. Conclusions

The use of a three-fluid nozzle enabled the preparation of stable emulsions, which were subsequently spray-dried to produce avocado oil microcapsules with adequate yield, high encapsulation efficiency, and low moisture content. The particles exhibited controlled size distribution, characteristic surface morphology, and thermodynamic behavior that defined a stability window associated with monolayer moisture content and minimum entropy conditions. Glass transition analysis confirmed that oil incorporation did not compromise the structural stability of the protein matrix. Overall, the process generated microcapsules with favorable physicochemical and energetic properties, supporting their application as a strategy to improve the stability and functionality of avocado oil in powdered form.