Optimization and mechanistic insights into SiO₂ nanoparticle–CTAB surfactant pickering emulsions for water mobility control

Abstract

Pickering emulsions have gained considerable attention for water mobility control because of their high kinetic stability and resistance to coalescence under harsh reservoir conditions. Here, SiO₂ nanoparticles (NPs) combined with cetyltrimethylammonium bromide (CTAB) were used to stabilize emulsions for flow diversion in dual-permeability porous media. The novelty of this work lies in integrating a Box–Behnken experimental design (BBD) with quantitative analysis to evaluate the individual and interactive effects of NPs concentration, surfactant concentration, and water-to-oil (W/O) ratio on emulsion stability and to develop a predictive correlation for the stability index (SI). The model showed excellent agreement with the experimental data (R² = 0.9849), identifying NPs concentration as the dominant factor and revealing a significant NPs–surfactant interaction, with an optimal W/O ratio near 60/40. Thermal analysis indicated that the emulsions remained stable up to 80 °C; above this temperature, droplet size increased markedly (47 μm to 149 μm), accompanied by a decrease in stability. Rheological measurements confirmed shear-thinning behavior: viscosity increased from 30 to 57 mPa.s with increasing water fraction and decreased to 15 mPa.s at 120 °C. Micromodel flooding demonstrated effective flow diversion, increasing oil recovery from 46% (50/50) to 68% (75/25) under seawater (SW) flooding, whereas high-salinity conditions reduced recovery to approximately 27–34%.

Introduction

The application of emulsions—either preformed prior to injection or generated in situ within the reservoir—has been widely recognized as an effective method for water mobility control and flow diversion in heterogeneous porous media^1,2. By preferentially restricting flow through high-permeability channels, emulsions enhance sweep efficiency and redirect injected fluids toward low-permeability or previously unswept regions^3,4. At the pore scale, this behavior is primarily governed by the balance between viscous drag and capillary forces, where emulsion droplets may become temporarily trapped within pore constrictions and subsequently released when the imposed pressure gradient exceeds the local capillary resistance^5–8.

Among various emulsion systems investigated for mobility control, particle-stabilized emulsions have attracted increasing attention due to their superior kinetic stability and resistance to coalescence under harsh reservoir conditions⁹. Unlike classical surfactant-stabilized emulsions, Pickering emulsions are stabilized by the irreversible adsorption of solid colloidal particles at the oil–water interface, forming a rigid mechanical barrier that suppresses droplet coalescence and enhances structural integrity during flow^10–18. The type and stability of Pickering emulsions are strongly governed by particle wettability, which determines their affinity toward the oil and water phases and ultimately controls emulsion type and flow behavior^19–22.

Silica NPs in combination with cationic surfactants have been widely studied as a versatile platform for tuning NPs wettability and stabilizing Pickering emulsions^23–29. Surfactant adsorption on NPs surfaces modifies surface charge, interfacial tension, and contact angle through concentration-dependent molecular rearrangements, including electrostatic adsorption, lateral aggregation, and bilayer formation^30–34. These interfacial phenomena directly influence emulsion stability, droplet size distribution, and rheological behavior, which in turn control permeability reduction and flow diversion in porous media^35–49.

Despite extensive prior research on NPs–surfactant-stabilized emulsions, most studies have examined the effects of formulation parameters—such as NPs concentration, surfactant dosage, or water-to-oil ratio—independently, without accounting for their coupled and nonlinear interactions^50–54. Moreover, while emulsion stability and rheology are commonly evaluated under static or bulk conditions, experimental validation of flow diversion mechanisms at the pore scale remains limited. In particular, the combined use of statistically optimized emulsion formulation and direct micromodel visualization has received little attention in the context of Pickering emulsions for water mobility control.

In the present study, these gaps are addressed through an integrated experimental and statistical framework. A Box–Behnken design (BBD) is employed to systematically evaluate the individual and interactive effects of SiO₂ NPs concentration, CTAB surfactant concentration, and water-to-oil ratio on emulsion stability, leading to the development of a predictive correlation for the stability index. The optimized Pickering emulsions are further examined through thermal stability analysis, rheological characterization using the power-law model, and glass micromodel flooding experiments to directly visualize flow diversion and droplet entrapment in dual-permeability porous systems. This combined BBD optimization and micromodel validation approach provides new mechanistic insights into the design of NPs–surfactant Pickering emulsions for effective water mobility control and enhanced sweep efficiency in heterogeneous reservoirs.

Materials and methods

Synthetic formation water (FW) and SW were prepared using MgCl₂, CaCl₂, KCl, NaCl, Na₂SO₄, and NaHCO₃. All salts were purchased from Dr. Mojallali Chemical Company (purity > 99%). Each salt was weighed and dissolved in deionized water, followed by continuous magnetic stirring to ensure complete dissolution and homogeneity. The ionic compositions of the FW and SW are provided in Table 1. The FW composition represents the formation brine of the oil field from which the crude oil was obtained, whereas the SW composition corresponds to a standard laboratory seawater formulation reported in the literature⁵⁵.

Table 1.

Ionic composition and physicochemical properties of the FW and SW used in this study⁵⁵.

FW	Na+ (ppm)	Mg2+ (ppm)	Ca2+ (ppm)	Cl− (ppm)	SO42− (ppm)	HCO3− (ppm)	K+ (ppm)	Salinity (ppm)	Ionic strength (mol/l)
	49,898	3248	14,501	111,812	234	162	0	179,855	3.659
SW	12,664	1641	500	23,308	3036	77	432	41,658	0.833

The salinity was calculated as the sum of the individual ionic concentrations, while the ionic strength was determined using the classical expression ​, where ​ and ​ denote the molar concentration and charge of ion i, respectively, based on the reported ionic composition.

The oil phase used in this study was crude oil obtained from one of the oil fields located in southwestern Iran. The physical properties, including density, viscosity, and API gravity, together with the SARA analysis (saturates, aromatics, resins, and asphaltenes), are summarized in Table 2.

Table 2.

Physicochemical properties of the crude oil used in this study⁵⁵.

SARA analysis				Crude oil properties
Resins (wt%)	Asphaltenes (wt%)	Aromatics (wt%)	Saturates (wt%)	TBN (mg KOH g −1)	TAN (mg KOH g −1)	API	Viscosity at 15.56 °C
10.16	6.69	45.68	37.47	1.11	0.15	36.89	3.25

To prepare and stabilize the emulsions, a cationic surfactant (CTAB, purity 99%, Merck, Germany) with a critical micelle concentration (CMC) of 0.82 mM, was employed. For the preparation of Pickering emulsions, SiO₂ NPs supplied by NanoSany Co. (Iran) were used as solid stabilizers. The physical and structural characteristics of this NPs is summarized in Table 3.

Table 3.

Physical and structural characteristics of the NPs used in this study⁵⁶.

	Size (nm)	SSA (m2/g)	Morphology	Density (g/cm3)	Color	Wettability	Purity (%)
SiO2 NPs	20–35	200	Spherical	2.100	White	Hydrophilic	99.98

Evaluation of emulsion stability

Two complementary methods were employed to evaluate emulsion stability: visual observation of phase separation using graduated cylinders and droplet size analysis through optical microscopy. For visual assessment, samples with a fixed volume of 200 mL were transferred into transparent graduated tubes and maintained under constant temperature conditions. At specific time intervals, physical changes such as the formation of a cream layer at the top or a serum layer at the bottom were recorded until equilibrium was achieved. The SI was employed to quantitatively assess the stability of the emulsions, and it was calculated using Eq. 1:

1

To determine the optimal conditions and analyze the influence of key variables on emulsion stability, the BBD was employed using Minitab 16 statistical software. This statistical approach efficiently estimates nonlinear relationships among variables while minimizing the required number of experiments compared with full factorial designs. Three independent parameters were selected: NPs concentration (0.1, 0.55, and 1 wt%), surfactant concentration (0.1, 0.55, and 1 mM), and W/O ratio (25/75, 50/50, and 75/25). The response variable was the SI, measured as described previously. The BBD generated 15 experimental runs, including 12 factorial points and three center-point replicates for error estimation and reproducibility assessment. The experimental conditions corresponding to each run are listed in Table 4.

Table 4.

BBD for evaluating emulsion stability.

Sample	NPs concentration (wt%)	Surfactant concentration (mM)	W/O ratio
1	0.1	0.1	50/50
2	1	0.1	50/50
3	0.1	1	50/50
4	1	1	50/50
5	0.1	0.55	75/25
6	1	0.55	75/25
7	0.1	0.55	25/75
8	1	0.55	25/75
9	0.55	0.1	75/25
10	0.55	1	75/25
11	0.55	0.1	25/75
12	0.55	1	25/75
13	0.55	0.55	50/50
14	0.55	0.55	50/50
15	0.55	0.55	50/50

To investigate the effect of temperature on emulsion stability, optical microscopy was employed to analyze droplet morphology and microstructure. The emulsion samples were imaged, and droplet sizes were measured and recorded. The obtained micrographs were processed using ImageJ software to determine the droplet size distribution. Microscopic observations and droplet size analysis provided valuable insights into the occurrence of destabilization phenomena such as coalescence and Ostwald ripening.

Contact angle measurements

The wettability of CTAB-modified SiO₂ NPs was evaluated through contact angle measurements. modification procedure followed the schematic shown in Fig. 1. Briefly, SiO₂ NPs were dispersed in deionized water by ultrasonication and then magnetically stirred to ensure homogeneous suspension. CTAB was added directly to the suspension and the mixture was stirred to promote adsorption of surfactant molecules onto the NPs surfaces. The modified NPs were then compressed into pellets and placed in a chamber filled with deionized water. An oil droplet was introduced from below onto the pellet surface, and the equilibrium contact angle was recorded using an optical imaging setup (Fig. 1). The captured images were analyzed in ImageJ to determine the contact angle from the droplet profile.

Fig. 1.

Schematic representation of the preparation of surfactant-modified SiO₂ NPs and the contact angle measurement procedure using the optical imaging technique.

Rheological measurements

The rheological properties were measured using a Fann 35 rotational viscometer (Fann Instrument Co.)⁵⁷. Measurements were performed on 350 mL samples placed in the annular gap between the rotor and the cup. The instrument provides six rotational speeds (3, 6, 100, 200, 300, and 600 rpm). The effects of W/O ratio and temperature on apparent viscosity and flow behavior were evaluated.

Glass micromodel flooding experiments

A glass micromodel was fabricated to visualize displacement in porous media. The model consisted of two glass plates, one laser-etched with a predesigned pore network pattern. Inlet and outlet ports were drilled, and the plates were thermally bonded at 710 °C for 2 h with a controlled heating rate of 5.75 °C/min, followed by gradual cooling over 24 h. The micromodel dimensions were 3 × 8 × 0.02 cm³, with a porosity of 44.27%, and a pore volume of 0.2389 mL (Fig. 2). Experiments were conducted at ambient conditions on a leveled surface to minimize gravitational effects. Prior to injection, the micromodel was vacuumed to ensure full initial saturation⁵⁸.

Fig. 2.

Schematic diagram of the experimental setup used for the glass micromodel injection tests. Adapted and slightly modified from reference⁵⁸.

In each experiment, crude oil was injected using a syringe pump to establish the initial oil saturation. The micromodel was then aged under static conditions for one week at reservoir temperature. Brine was subsequently injected at a constant flow rate of 0.002 mL/min, followed by emulsion injection at the same rate. The system was shut in for 1 h to reach equilibrium, after which brine injection was resumed at the same flow rate. The process was monitored by high-resolution imaging, and oil recovery was quantified using ImageJ. The oil recovery factor (RF) was calculated from pixel-based image analysis as defined in Eq. 2⁵⁹.

2

Results and discussion

Results of the emulsion stability tests

To evaluate the stability of the prepared emulsions, fifteen experiments were conducted under ambient conditions according to the BBD. The samples were monitored over a 14-day period to reach equilibrium, during which phase separation behavior and the type of emulsion formed—either oil-in-water or water-in-oil—were examined (Fig. 3a). This variation was primarily governed by the surfactant concentration and the W/O ratio. Specifically, when the water fraction exceeded the oil fraction or when higher surfactant concentrations increased the hydrophobicity of the NPs surfaces, the continuous phase shifted from water to oil, resulting in the formation of water-in-oil emulsions. The quantitative results of the stability tests, expressed in terms of the SI are presented in Fig. 3b. The SI values ranged from a minimum of 0.31 to a maximum of 0.85, indicating a wide variation in emulsion stability behavior under different experimental conditions. The reproducibility of the SI measurements was confirmed by three center-point replicates, which exhibited good repeatability with a standard deviation of 0.0097.

Fig. 3.

(a) Visual assessment of emulsion stability after 14 days of storage at ambient temperature and (b) Emulsion SI for the samples evaluated in the experimental design.

The inversion of emulsion type can be mechanistically attributed to the alteration of NPs wettability caused by surfactant adsorption at increasing concentrations (Fig. 4a). Surfactant molecules, owing to their amphiphilic structure, progressively adsorb onto the NPs surface, modifying interfacial tension, surface charge, and contact angle^60–68. This adsorption involves a sequence of molecular rearrangements driven by electrostatic, van der Waals, and hydrophobic interactions^69–72. As surfactant concentration rises, lateral aggregation and hemimicelle formation occur, followed by bilayer development beyond the CMC, which renders the NPs surface increasingly hydrophobic^73,74. It should be emphasized that this wettability alteration corresponds to surfactant adsorption on the SiO₂ NPs surface, as contact angle measurements were conducted on compressed pellets prepared exclusively from CTAB-modified NPs rather than on glass or bare substrates. The progressive evolution of surfactant adsorption configurations and their corresponding effects on wettability are schematically illustrated in Fig. 4b, which depicts the transition from electrostatically driven monolayer adsorption to hemimicelle formation and ultimately bilayer development as surfactant concentration increases. Consequently, this wettability shift from hydrophilic to hydrophobic conditions alters the affinity of NPs toward the surrounding phases, facilitating the transition from an oil-continuous to a water-continuous system.

Fig. 4.

a) Modification of SiO₂ NPs wettability at different CTAB surfactant concentrations. b) Schematic illustration of the concentration-dependent adsorption of a cationic surfactant on a hydrophilic SiO₂ NPs surface, showing the successive adsorption stages and the corresponding qualitative changes in wettability. Adapted and slightly modified from³¹.

The statistical significance of each factor and their interactions was evaluated through ANOVA, as summarized in Table 5. Among the main factors, the NPs concentration exhibited the most significant effect on emulsion stability (p < 0.01). Increasing the NPs concentration markedly enhanced the SI, which can be attributed to the formation of a stronger physical barrier at the phase interface, effectively preventing droplet coalescence. The surfactant concentration also showed a statistically significant positive effect (p < 0.05), primarily due to the reduction in interfacial tension and the improved dispersion of NPs. Furthermore, the W/O ratio had a significant influence on stability (p < 0.01); an increase in the water fraction initially improved stability up to an optimal level, beyond which excessive water content led to a decline in emulsion stability.

Table 5.

ANOVA results for the BBD showing the linear, quadratic, and interaction effects of key variables on the emulsion SI.

SOV	DF	Seq SS	Adj SS	Adj MS	F	P
NPs concentration	1	0.355835	0.101586	0.101586	58.70	0.001
Surfactant concentration	1	0.043293	0.021569	0.021569	12.46	0.017
W/O ratio	1	0.009038	0.038920	0.038920	22.49	0.005
NPs concentration * NPs concentration	1	0.015948	0.024859	0.024859	14.37	0.013
Surfactant concentration * Surfactant concentration	1	0.012832	0.018757	0.018757	10.84	0.022
W/O ratio * W/O ratio	1	0.097438	0.097438	0.097438	56.31	0.001
NPs concentration * Surfactant concentration	1	0.012201	0.012201	0.012201	7.05	0.045
NPs concentration * W/O ratio	1	0.001678	0.001678	0.001678	0.97	0.370
Surfactant concentration * W/O ratio	1	0.017638	0.017638	0.017638	10.19	0.024
Residual error	5	0.008652
Total	14	0.574553
R2=98.49% R2pred=76.58% R2adj=95.78%

SOV: Source of Variation, DF: Degree of Freedom, SS: Sum of Squares, MS: Mean Square.

In addition to the linear effects, the quadratic terms were also found to be significant. In particular, the squared term of the W/O ratio showed a strong nonlinear relationship (p < 0.01), indicating the existence of an optimum value for this parameter. Moreover, the interaction between NPs and surfactant concentrations was statistically significant (p < 0.05), suggesting that the performance of these two factors is mutually dependent. Specifically, in the presence of an adequate amount of surfactant, an increase in NPs concentration exerts a more pronounced effect on emulsion stability.

For statistical analysis of the experimental data, a second-order regression model based on response surface methodology (RSM) was fitted to the results. The regression and analysis of variance (ANOVA) confirmed the high accuracy of the model, with coefficients of determination of R² = 0.9849 and R²_pred = 0.7658. These values indicate that variations in the emulsion SI are adequately captured by the selected independent variables and their interactions. The fitted response surface model for SI is given in Eq. 3:

3

Where CNP_s denotes the NPs concentration (wt%), CS represents the surfactant concentration (mM), and R refers to the W/O ratio.

Although the high R² value indicates excellent agreement with the experimental data, the comparatively lower R²_pred can be attributed to the inherent nonlinearity of the emulsion system and the limited number of experimental runs typical of Box–Behnken designs. Since the primary objective of the model is to elucidate variable interactions and identify optimal operating conditions within the design space rather than to achieve global prediction, the statistical indicators confirm that the model remains robust and suitable for optimization and mechanistic interpretation, without dominant overfitting.

The two-dimensional contour plots (Fig. 5) illustrate the combined effects of two variables on the emulsion SI, while the third variable was held constant. As shown in Fig. 5a, a simultaneous increase in NPs and surfactant concentrations led to a pronounced improvement in stability, with SI values exceeding 0.9 observed in the region where both factors were at high levels. However, at low NPs concentrations, even an increase in surfactant content failed to achieve satisfactory stability.

Fig. 5.

Two-dimensional contour plots showing the combined effects of: (a) NPs concentration and surfactant concentration, (b) NPs concentration and W/O ratio, and (c) surfactant concentration and W/O ratio on emulsion stability.

Figure 5b demonstrates that increasing the NPs concentration consistently enhanced stability, whereas the W/O ratio exhibited a nonlinear trend, with maximum stability observed at intermediate ratios (approximately 60/40). Similarly, Fig. 5c reveals that higher surfactant concentrations, when combined with moderate W/O ratios, created optimal conditions for forming a stable emulsion. This finding highlights the synergistic interaction between these two factors. It was also observed that the beneficial effect of surfactant concentration reached a plateau beyond a certain threshold, beyond which no significant improvement in stability occurred.

To examine the effect of temperature on emulsion stability, experiments were conducted under fixed conditions—NPs concentration of 1 wt%, surfactant concentration of 1 mM, and a W/O ratio of 75/25—across a temperature range of 25 °C to 120 °C over a one-week period. The results showed that at all tested temperatures, the emulsions maintained a water-in-oil structure (Fig. 6a). However, the emulsion SI exhibited a decreasing trend with increasing temperature, dropping from 0.9595 at 25 °C to 0.6382 at 120 °C. The trend indicates that while the decline in stability was gradual up to approximately 80 °C, a more pronounced reduction occurred beyond 100 °C, suggesting a higher sensitivity of the emulsion structure to elevated temperatures (Fig. 6b).

Fig. 6.

(a) Evaluation of emulsion stability and type at different temperatures (40, 60, 80, 100, and 120 °C); (b) variation of the emulsion SI as a function of temperature.

Microscopic images and droplet size distribution plots obtained at various temperatures (25–120 °C) revealed that increasing temperature had a direct impact on the microstructure and droplet size distribution of the emulsions (Fig. 7). The average droplet diameter increased from 11.79 μm at 25 °C to 13.97, 30.98, 46.81, 148.62, and 174.10 μm at 40, 60, 80, 100, and 120 °C, respectively. This trend indicates that droplet growth occurred gradually up to about 80 °C, whereas a sharp and substantial increase was observed beyond 100 °C, consistent with the pronounced decline in the SI at higher temperatures.

Fig. 7.

Effect of temperature on the average droplet size and droplet size distribution of the emulsions.

Analysis of the droplet size frequency distributions further confirmed that from 25 °C to 80 °C, the distributions progressively shifted toward larger sizes, indicating droplet coalescence and the formation of coarser droplets. This shift was accompanied by a reduction in the total droplet count, as smaller droplets merged to form larger ones. At higher temperatures (100 °C and 120 °C), in addition to coalescence, Ostwald ripening became more pronounced, further accelerating droplet growth and leading to enhanced emulsion destabilization.

Rheological test results

Viscosity and rheological behavior are among the most critical physical properties of emulsions, as they directly influence both their stability and performance in practical applications. Based on the results of the stability experiments, emulsions were prepared using fixed concentrations of 1 wt% NPs, 1 mM surfactant, and various W/O ratio for rheological testing. The variation of shear stress with respect to shear rate is presented in Fig. 8. The rheological analysis revealed that the emulsions exhibited non-Newtonian flow behavior, which was well described by the power-law model (Eq. 4)⁵⁷.

4

Fig. 8.

Rheological behavior of emulsions with different W/O ratios at 25 °C, showing the variation of shear stress with shear rate and indicating a shear-thinning trend that becomes more pronounced as the water fraction increases.

In the power-law model, the shear stress (τ) is expressed in pascals (Pa), the shear rate (γ̇) in reciprocal seconds (s⁻¹), the consistency index (K) in Pa.sⁿ, and the flow behavior index (n) is dimensionless. Table 6 presents the parameters obtained from the power-law fitting, along with the apparent viscosity measured at a shear rate of 1022 s⁻¹, for emulsions prepared with different W/O ratios at 25 °C.

Table 6.

Apparent viscosity at a shear rate of 1022 s^− 1 and power-law model parameters for emulsions prepared at different W/O ratios at 25 °C.

W/O ratio	45/55	55/45	65/35	75/25	85/15
AV (mPa.s)	30.4518	39.9352	49.4174	54.4029	56.8943
K (Pa.sn)	0.6878	0.9278	1.1773	1.4266	1.5654
n	0.5646	0.5611	0.5575	0.5512	0.5508
R2	0.9888	0.9873	0.9879	0.9706	0.9493

As observed, increasing the water fraction led to higher values of K and apparent viscosity, while n decreased. This indicates that emulsions became more viscous yet more shear-thinning as the water content increased. The variation in apparent viscosity was consistent with the stability results—at higher W/O ratios, emulsions exhibited lower stability despite higher K values, primarily due to the decrease in n and the formation of weaker, more fragile structures. The enhanced shear-thinning behavior observed at higher water fractions can be attributed to the increase in average droplet size, which promotes emulsion coalescence and, consequently, reduces overall stability. At elevated W/O ratios, the apparent viscosity increase is mainly governed by droplet crowding and transient droplet–droplet interactions rather than by the formation of a mechanically robust interfacial network. As nanoparticle surface coverage becomes insufficient, the interfacial films weaken, leading to reduced resistance to coalescence despite higher apparent viscosity.

The variation of shear stress with shear rate is illustrated in Fig. 9, while Table 7 summarizes the apparent viscosity and power-law parameters of the emulsions at different temperatures under a fixed W/O ratio of 75/25. The results indicate that as temperature increased, both K and n decreased, reflecting a more pronounced shear-thinning behavior at elevated temperatures. The observed decline in viscosity with increasing temperature at 1022 s^− 1, confirms this trend. These findings are consistent with the stability analysis, as higher temperatures promote droplet growth, intensify coalescence and Ostwald ripening, and consequently lead to a marked reduction in emulsion stability.

Fig. 9.

Rheological behavior of emulsions at a fixed W/O ratio of 75/25 under different temperatures (25–120 °C), showing the variation of shear stress with shear rate and the decline in viscosity with increasing temperature.

Table 7.

Apparent viscosity at a shear rate of 1022 s^− 1 and power-law model parameters for emulsions prepared at different temperatures with a fixed W/O ratio of 75/25.

Temperature	25	40	60	80	100	120
AV (mPa.s)	54.4029	51.4157	45.4218	40.4364	31.9427	15.4739
K (Pa.sn)	1.4266	1.3754	1.2648	1.1398	0.9194	0.4452
n	0.5512	0.5470	0.5434	0.5428	0.5373	0.5345
R2	0.9706	0.9733	0.9657	0.9620	0.9694	0.9771

Glass micromodel injection test results

A dual-permeability micromodel, consisting of interconnected low- and high-permeability regions, was employed to evaluate the ability of emulsions to restrict water flow through high-permeability channels. Experiments were conducted using 1 wt% NPs and 1 mM surfactant, while varying the W/O ratio and salinity conditions (SW and FW) to assess their effect on the blocking efficiency of the emulsions.

After the micromodel was fully saturated with crude oil, water flooding was initiated using SW and FW as displacing fluids. The results indicated that, due to the permeability contrast, water preferentially flowed through the high-permeability region, appearing prematurely at the outlet (Fig. 10). This behavior, accompanied by flow fingering, significantly reduced the displacement efficiency, resulting in a low oil recovery factor of approximately 6–10%. Such findings clearly demonstrate the limitations of conventional water flooding in heterogeneous media and highlight the necessity of employing enhanced oil recovery (EOR) techniques to improve sweep efficiency.

Fig. 10.

(a) SW and (b) FW injection into the oil-saturated micromodel, illustrating the occurrence of flow fingering through the high-permeability region.

In the next stage, the primary objective was to evaluate the ability of emulsions to block water flow within the high-permeability region of the micromodel. Initially, SW flooding was performed at different W/O ratios (50/50, 65/35, and 75/25). Overall, the observations revealed that the emulsions effectively restricted flow through high-permeability channels, thereby increasing flow resistance and diverting part of the injected fluid toward the low-permeability zones.

This improvement was primarily attributed to two key mechanisms: (i) the enhanced viscosity of the emulsions at higher water fractions, which facilitated better mobility control and reduced the effective permeability of the high-permeability paths; and (ii) the jamming phenomenon, caused by droplet entrapment within the observable pore throats. Figure 11 presents the evolution of the SW injection front at different time intervals.

Fig. 11.

Fluid front movement in the micromodel during SW injection at different W/O ratios (50/50, 65/35, and 75/25).

Although the 50/50 emulsion achieved a noticeably higher oil recovery compared with conventional brine flooding, a large fraction of the injected fluid still preferentially flowed through the high-permeability region. Increasing the water fraction to 65% yielded an additional ~ 6% improvement in recovery, primarily due to the enhanced viscosity of the emulsion, which improved mobility control and reduced water channeling. A further increase to 75% water resulted in a more substantial enhancement of approximately 15%, suggesting that at higher water contents, mechanisms beyond viscosity effects—such as droplet entrapment and selective flow diversion—became increasingly dominant (Fig. 12).

Fig. 12.

Oil recovery factor during SW injection into the micromodel at different W/O ratios (50/50, 65/35, and 75/25).

Two primary mechanisms can explain this trend: first, the increase in emulsion viscosity, which improves flow front control, and second, the growth of dispersed-phase droplets. Since viscosity enhancement occurred in both transitions (from 50/50 to 65/35 and from 65/35 to 75/25), it appears that the major contribution to the improvement observed in the second stage originated from droplet enlargement. Larger droplets in the 75/25 emulsion had a higher likelihood of becoming trapped within pore throats, thereby more effectively blocking high-permeability channels. Figure 13 presents microscopic images of emulsions with different W/O ratios captured within the pore constrictions of the micromodel. These observations suggest that although increasing the water fraction led to droplet growth in both cases, only at higher water ratios (65/35 to 75/25) did the droplets reach a sufficient size to exert a significant blocking effect on the porous medium.

Fig. 13.

Microscopic images of emulsions with different W/O ratios (a) 50/50, (b) 65/35, and (c) 75/25), each showing three representative regions of the micromodel to illustrate droplet distribution and entrapment within pore constrictions.

A review of previous investigations reveals that the mechanisms responsible for water mobility control by emulsions are influenced primarily by the emulsion type and its internal structure⁷⁵. In several studies, the enhancement of emulsion viscosity has been identified as a key factor in improving mobility control, as higher viscosity increases flow resistance, stabilizes the displacement front, and suppresses viscous fingering within porous media^76–79. Such behavior is most pronounced in inverse emulsions (water-in-oil), where the continuous oil phase provides a viscous matrix that effectively reduces the mobility ratio. These systems exhibit an optimal water content, at which both viscosity and injection pressure reach their maximum values, resulting in piston-like displacement and selective blockage of high-permeability channels. Beyond this range, structural instability reduces their efficiency, indicating that viscosity-driven resistance dominates only within a specific compositional window⁷⁷.

In contrast, studies emphasizing the capillary-controlled behavior of emulsions attribute the reduction in water mobility mainly to droplet entrapment and local jamming within pore constrictions^80–85. Under low capillary number conditions or when the droplet size approaches or exceeds the pore throat radius, capillary forces surpass viscous effects, causing intermittent plugging of conductive pathways and diversion of flow to less permeable regions. This mechanism is particularly relevant for oil-in-water systems, which generally possess lower viscosity; here, flow control arises predominantly from physical droplet blocking rather than rheological resistance⁸⁶. Recent work on Pickering emulsions has bridged these two perspectives, demonstrating that NPs-stabilized systems combine the advantages of high viscosity and interfacial rigidity, achieving selective and reversible water blockage without significantly impairing oil permeability^87–92. Overall, the collective evidence suggests that viscosity-driven control governs inverse and Pickering emulsions, whereas jamming and capillary entrapment dominate in classical oil-in-water systems, reconciling the apparent inconsistencies reported in earlier studies.

In the subsequent stage, FW—with a higher salinity than SW—was used as the displacing fluid in the micromodel to examine the effect of salinity on emulsion performance. The results revealed that increased salinity had a markedly negative impact on emulsion stability. Under these conditions, the injected water front predominantly advanced through the high-permeability channels, and unlike the SW case, no significant flow diversion toward low-permeability zones was observed.

Despite this, the oil recovery factor ranged between 27% and 34%, which, although lower than that achieved with SW flooding, indicates that the emulsions still provided a degree of flow resistance even under high-salinity conditions (Fig. 14). These findings demonstrate that excessive salinity substantially reduces emulsion stability and thereby diminishes the blocking efficiency of emulsions in high-permeability regions, ultimately limiting the overall process performance.

Fig. 14.

Oil recovery factor during FW injection into the micromodel at different W/O ratios (50/50, 65/35, and 75/25).

Conclusions

This study provided a comprehensive experimental and statistical investigation into the formulation, stability, rheology, and flow behavior of SiO₂ NPs–CTAB-stabilized Pickering emulsions for water mobility control in heterogeneous porous media. The results demonstrated that the composition of the emulsion formulation strongly governs its overall performance. Increasing the NPs concentration markedly improved emulsion stability due to the formation of a denser interfacial layer that restricted droplet coalescence, while the surfactant concentration exhibited a secondary yet synergistic influence by enhancing particle dispersion and surface activity. The W/O ratio showed a nonlinear effect, with the highest stability achieved at an intermediate ratio near 60/40, where interfacial forces between the dispersed and continuous phases reached equilibrium.

The response surface model derived from the BBD successfully correlated the effects of formulation variables with the emulsion SI, providing strong predictive capability (R² = 0.9849). This approach enabled quantitative optimization of the emulsion system with minimal experimental effort and can serve as a valuable tool for practical formulation design. Thermal analyses revealed that the emulsions maintained excellent stability up to 80 °C; however, above this threshold, droplet enlargement caused by coalescence and Ostwald ripening resulted in a significant decline in stability and homogeneity.

Rheological investigations confirmed that the emulsions exhibited non-Newtonian, shear-thinning behavior consistent with the power-law model. Increasing the water fraction led to higher apparent viscosity and more pronounced pseudoplasticity, reflecting the formation of a more interconnected droplet network. Conversely, increasing temperature reduced viscosity and consistency, indicating thermal weakening of interfacial films and disruption of droplet structures.

Micromodel flooding experiments further validated the potential of these emulsions for flow control applications. Increasing the water fraction to 65% raised oil recovery by ~ 6%, mainly due to higher viscosity and better mobility control, while a further rise to 75% enhanced recovery by ~ 15%, driven by droplet entrapment and selective flow diversion. Emulsions with higher water fractions effectively diverted the injected brine from high- to low-permeability zones, improving sweep efficiency and increasing oil recovery to 68% at 75/25 under SW flooding. In contrast, elevated salinity adversely affected emulsion stability and reduced recovery to approximately 27–34%, underscoring the sensitivity of NPs–surfactant systems to ionic strength.

Author contributions

B.A.: Writing – original draft, Visualization, Validation, Methodology, Investigation, Formal analysis, Conceptualization.E.S.: Writing – review & editing, Supervision, Methodology, Investigation, Conceptualization.All authors read and approved the final manuscript.

Data availability

The datasets used and/or analysed during the current study available from the corresponding author on reasonable request.

Declarations

Competing interests

The authors declare no competing interests.