Nanoemulsification of soybean oil using ultrasonic microreactor: Process optimization, scale-up and numbering-up in series

Abstract

Ultrasonically-induced nanoemulsions have been widely investigated for the development of functional food, cosmetics, and pharmaceuticals due to ideal droplet sizes (DS), low polydispersity index (PDI), and superior physical stability. However, a series of frequently-used ultrasonic set-ups mainly suffered from a low ultrasonic energy efficiency caused by the large acoustic impedance and energy consumption, subordinately confronted with a low throughput, complicated fabrication with complex structure and weak ultrasonic cavitation. Herein, we employed a typical ultrasonic microreactor (USMR) that ensured the high-efficient energy input and generated intense cavitation behavior for efficient breakage of droplets and continuous production of unified oil-in-water (O/W) nanoemulsions in a single cycle and without any pre-emulsification treatment. The emulsification was optimized by tuning the formula indexes, technological parameters, and numerical analysis using Response Surface Methodology (RSM), followed by a comparison with the emulsification by a traditional ultrasonic probe. The USMR exhibited superior emulsification efficiency and easy scale-up with remarkable uniformity by series mode. In addition, concurrent and uniform nanoemulsions with high throughput could also be achieved by a larger USMR with high ultrasonic power. Based on RSM analysis, uniform DS and PDI of 96.4 nm and 0.195 were observed under the optimal conditions, respectively, well consistent with the predicted values. Impressively, the optimal nanoemulsions have a uniform spherical morphology and exhibited superior stability, which held well in 45 days at 4℃ and 25℃. The results in the present work may provide a typical paradigm for the preparation of functional nanomaterials based on the novel and efficient emulsification tools.

1. Introduction

Nanoemulsions are isotropic, transparent/translucent colloidal systems composed of nanodroplets of oil stably dispersed in aqueous media by using external energy and moderate dosage of surfactant [1]. Because of their good biocompatibility, easy fabrication and cost effectiveness, nanoemulsions have been widely investigated in the development of functional foods, cosmetics, and pharmaceuticals [2], [3], [4], [5]. It is worth mentioning the differences between microemulsions and nanoemulsions as summarized in Table 1. Unexpectedly, the droplet size (DS) of nanoemulsions is slightly larger than that of microemulsions, inconsistent with the dimension defines for nano-/micro- size. The phenomenon can be attributed to the historical reasons that the DS of microemulsions was usually defined to be<100 nm. The traditional preparation of microemulsions with below 100 nm DS requires a large number of surfactants and co-surfactants up to with ca. 15% [6], which brings out some obvious side effects, e.g., severe physiological stimulation and allergic reactions [7]. In order to reduce the use of surfactants and improve the compliance, high-energy technologies (high-pressure homogenization, ultrasound, etc.,) were attempted to prepare microemulsions. However, the DS of emulsions<500 nm are generally achieved and ultimately named as nanoemulsions [8]. Unlike DS discrepancy, the thermodynamic property should be a precise and significant rule to distinguish nanoemulsions from microemulsions, rather than DS of emulsions. The detailed discussion on the differences and similarities have been reported in the previous works [9], [10], [11].

Table 1.

Comparison between nanoemulsions and microemulsions.

Item	Nanoemulsions	Microemulsions
DS (nm)	20–500	10–100
Droplet morphology	Spherical	Spherical, non-spherical
Preparation methods	High/low energy input	Low energy input
Stability	Thermodynamically unstable, kinetically stable	Thermodynamically stable
Surfactant amount	Moderate/high	Very high
Optical properties	Translucent, transparent	Transparent

Nanoemulsions are not thermodynamically stable systems and tend to spontaneously reduce the interface area, manifesting the disappearance of the small droplets and the growth of the large droplets under the influence of Ostwald Ripening [12]. Therefore, the fabrication of nanoemulsions with small DS and low polydispersity index (PDI) is significantly important to reduce Ostwald Ripening and improve the stability of nanoemulsions [1], [12]. However, the small DS and low PDI of emulsions are generally dependent on the preparation methods.

Typically, the preparation methods of nanoemulsions have been classified into low-energy emulsification and high-energy emulsification [13]. In comparison with low-energy emulsification, the high-energy emulsification may avoid the use of high-dosage of surfactant to produce nanoemulsions with high monodispersity and superior stability [14], [15]. In general, a high-pressure homogenizer, ultrasonicator, or microfluidizer are commonly served to high-energy emulsification [13]. Among them, ultrasonic emulsification has prominent advantages, such as high stability, superior emulsification efficiency, and low costs, etc. [16], [17]. Recently, R. Song et al. reported the ultrasonic-assisted preparation of eucalyptus oil nanoemulsions with an average DS of 19 nm and PDI of 0.39 were achieved in conical centrifuge tubes after optimization [18]. Although ultrasonic probes have been widely used for emulsification, the inhomogeneous distribution of ultrasonic field and fine particles of metal contamination severely limit the ultrasonic application in the research and development of emulsification [19]. In addition, the immersion setup of a microchannel into an ultrasonic bath has been recently attempted to intensify the emulsification for continuously producing uniform nanoemulsions without contamination of fine metal particles. S. Manickam et al. prepared the oil-in-water (O/W) nanoemulsions via combining ultrasound and microchannel, where the DS of 262.9 nm and PDI of 0.388 were obtained [20]. The pioneering works have made progresses at the synergistic interaction of coupling microchannels and ultrasound for producing nanoemulsions, but it is still considerable improvement in the ultrasonic energy transmission, energy input and emulsification efficiency due to the high acoustic impedances of soft microchannel and bulk aqueous medium [11], [19]. Recently, Olguin’s group developed a new active ultrasonic/acoustic microfluidic chip to continuously generate O/W or water-in-oil mini-emulsions in total flow rate of 1 mL/h [21]. Although the seamless attachmesnt of the ultrasonic transfer to the microfluidic device can significantly improve the ultrasonic transmission to the microchannels [19], [22], the ultrasonic setup suffers from the very limited throughput, complicated fabrication and structure, and confined ultrasonic cavitation in the production of emulsions [11].

Over the past years, Dong et al. also developed a novel ultrasonic microreactor (USMR), a component integrated with an ultrasonicator and a microreactor of none passive structure based on the mechanism of standing wave resonance, exhibiting uniform ultrasound radiation [23], [24], [25]. The joint USMR could confine the two immiscible fluids in a narrow space and ensure the high-efficient ultrasonic energy into the microreactor, which facilitate to generate intense and controllable ultrasonic cavitation [22], [26]. Such USMR is available in diverse specifications with different throughputs, therefore, multiple benefits, such as wide applicability, good flexibility, and easy fabrication, have been exhibited.

Due to abundant resources, easy absorption and high compatibility, soybean oil has been widely used as oil phase for the preparation of lipid emulsions, including the launched drugs of Aprepitan emulsion and Propofol medium- and long-chain fat emulsion. To date, the emulsification of soybean oil usually repeated homogenized in a high-pressure homogenizer and microfluidizer [27], [28]. Impressively, USMR exhibited superior performance for the emulsification by one single cycle and the pre-emulsification was not required [29]. However, USMR has not been applied for the nanoemulsification of soybean oil to produce uniform nanoemulsions with tunable sizes and monodispersity.

In this study, USMR was used to continuously prepare the monodisperse O/W nanoemulsions of soybean oil with low-dosage surfactants in water. The effects of single factors, such as aqueous phase/oil phase ratio (A/O ratio), surfactant amounts, and ultrasonic parameters, as well as interaction of multifactor with the response surface methodology (RSM) were investigated to obtain uniform nanoemulsions with small DS and low PDI. Furthermore, both numbering-up and scale-up strategies were considered to study the emulsification efficiency in the larger scale and improve the throughput. Finally, the stability of nanoemulsions obtained under the optimal conditions was evaluated in 45 days at different temperatures.

2. Materials and methods

2.1. Materials and setup

Soybean oil and phosphotungstic acid were purchased from Shanghai Aladdin Biochemical Technology Co., Ltd. (Shanghai, China). Tween-80 and Span-80 were purchased from Shanghai Macklin Biochemical Technology Co., Ltd. (Shanghai, China). Deionized (DI) water was used in all experiments.

Two different USMRs (umFlow-D and umFlow-Z) were supplied by MoGe um-Flow Technology Co., Ltd. (Shantou, China). Each USMR system mainly consisted of a microtube, a piezoelectric transducer (20 kHz), a ultrasonic generator, two syringe pumps, etc. (Fig. 1), as reported by Dong’s previous work [30], in which the different inner diameters of the microtube were used (Table S1). Two external fans were used to cool the piezoelectric and microtube device and a non-contact IR thermometer was used to monitor the temperature at the end of the microtube during sonication. The temperature of emulsions increased from the initial temperature of 20℃ to the balance temperature of 32-55℃, which depended on the ultrasonic power. In this device, the USMR was set up by directly coupling a quartz microtube and a piezoelectric transducer via epoxy glue (ENDFEST300, UHU, Germany), as shown in Fig. 1b. The tube with a diameter of 3.5 mm and a length size of 68 mm possesses two inlets of a T-type juncture (Fig. 1c), which is connected to the FEP tubes with an external diameter of 5 mm to deliver the aqueous and oil phases.

Fig. 1.

(a) Schematic of the setup used for the production of O/W nanoemulsions (Conditions: A/O ratio range of 5–19, residence time of 0.5–4 min, and ultrasonic power of 20–50 W). (b) Schematic of the USMR design. (c) Geometric information of the glass microtube.

As a comparison, a 24 kHz ultrasonic probe (BH-800) with adjustable amplitude was also used to prepare the O/W emulsions and provided by Baihang Ultrasonic Technology Co., Ltd. (Jiangsu, China).

The actual ultrasonic powers of ultrasonic devices were measured by the calorimetry method [31]. The ultrasonic probe was inserted in 430 mL of water with an isolated layer and the water temperature was recorded every 30 s during 10 min sonication using a thermometer (YET-610) provided by YOWEXA Measurement Technology Co., Ltd. (Shenzhen, China). The actual ultrasonic power was calculated by Eq. (1):

$$P_{\mathit{us}}=\frac{m\times C_p\times \mathrm{\Delta }T}{\mathrm{\Delta }t}$$ (1)

where P_us is the actual ultrasonic power (J s⁻¹ or W); m is the mass of water (kg); C_p is the specific heat capacity of water (J kg⁻¹ °C⁻¹); ΔT is temperature change (°C); Δt is the change in time (s).

Based on the triple measurement, the average ultrasonic power of umFlow-D was 7.6–35.8 W at 16–44 W of electrical output power, which was slightly higher than the 32.3 W of the 24 kHz ultrasonic probe with 56 W of electrical output power. The electrical output power was conveniently used in terms of factory settings of ultrasonic devices in this study.

2.2. Continuous preparation of nanoemulsions

For the preparation of O/W nanoemulsions, soybean oil was used as the oil phase and DI water containing different proportions of surfactants (Tween-80 and Span-80) was used as the aqueous phase. As shown in Fig. 1, two syringe pumps were used to simultaneously feed the oil phase and the aqueous phase at different ratios of flow rate into a USMR. To obtain nanoemulsions with small DS and low PDI, the critical parameters were investigated and optimized, including emulsion formulation, flow rate ratio, residence time, and ultrasonic power. The DS and PDI of all samples were used to measure one-way analysis of varianc (ANOVA). The significant differences were obtained at the p < 0.05 level [28].

For comparing the emulsification performance of USMRs with the traditional 24 kHz ultrasonic probe, a reasonable volume of 20 mL with the same A/O ratio, pulse mode of 5 s–5 s, and an electrical output power of 56 W was applied in the 24 kHz ultrasonic probe. All emulsion samples were prepared in triplicate. The values and error bars represent the mean and standard deviation, respectively.

2.3. Characterization of nanoemulsions

The DS and PDI of nanoemulsions were determined using dynamic light scattering (DLS) through Nanosizer (Nanolink S900, Linkoptik Instruments Co., Ltd., China). Prior to the measurement, the nanoemulsions were diluted several times with DI water to form translucent media. The morphology of nanoemulsions were analyzed by transmission electron microscope (TEM, Talos F200X G2, Thermo Scientific, USA). Before operating the measurement, the nanoemulsions should be negatively stained with phosphotungstic acid for improving visualized perception. In particle, a few nanoemulsions diluted with DI water were wisely dropped into the copper grid. Subsequently, phosphotungstic acid was added to the copper grid for negative staining and then dried at room temperature for intending the measurement.

2.4. Optimization of emulsifications using RSM and statistical analysis

In the preliminary experiments, A/O ratio, residence time, and ultrasonic power significantly affected the DS and PDI of emulsions prepared with USMR. The experiments were designed according to the Box-Behnken model using Design Expert Software (a trial version 22.0.2) to study the effects of independent variables: A/O ratio (X₁), residence time (X₂), and ultrasonic power (X₃) on DS (Y₁) and PDI (Y₂). The preparation parameters were optimized to reduce the DS and PDI. The independent variables and their levels are given in Table 2.

Table 2.

Independent variables and their corresponding levels for preparing nanoemulsions.

Independent variable	Symbol	Coded levels
		−1	0	1
A/O ratio	X1	7	9	11
Residence time (s)	X2	60	90	120
Ultrasonic power (W)	X3	40	45	50

Independent variables on the response variables. This effect could be expressed as Eq. (2).

$$\mathrm{Y}={\mathrm{\alpha }}_0+{\mathrm{\alpha }}_1{\mathrm{X}}_1+{\mathrm{\alpha }}_2{\mathrm{X}}_2+{\mathrm{\alpha }}_3{\mathrm{X}}_3+{\mathrm{\alpha }}_{11}{\mathrm{X}}_1^2+{\mathrm{\alpha }}_{22}{\mathrm{X}}_2^2+{\mathrm{\alpha }}_{33}{\mathrm{X}}_3^2+{\mathrm{\alpha }}_{12}{\mathrm{X}}_1{\mathrm{X}}_2+{\mathrm{\alpha }}_{13}{\mathrm{X}}_1{\mathrm{X}}_3+{\mathrm{\alpha }}_{23}{\mathrm{X}}_2{\mathrm{X}}_3$$ (2)

Where Y is the responses variable, α₀ is the constant; α_i, α_ii, and α_ij are linear, quadratic, and interactive coefficients, respectively. The coefficients were determined using Design Expert Software.

Statistical analysis for experimental data was performed to fit the second-order polynomial equations for independent variables. ANOVA was used to check the significant difference between these variables. Response surface plots were produced through Design Expert Software to visualize the effect of independent variables on response variables and predict the optimal process.

2.5. Evaluation of the stability of nanoemulsions

Nanoemulsions prepared under optimal conditions were stored at 4℃ and 25℃ for 45 days. In order to evaluate the stability, the DS and PDI of nanoemulsions were determined periodically.

3. Results and discussion

3.1. Effect of hydrophilic–lipophilic balance (HLB) and surfactant concentration on the DS and PDI

The abundant cavitation bubbles with strenuous vibration and periodic collapse can be triggered in the USMRs [32]. The violent bubble collapse causes the local strong turbulence, microjets, and shear force, favorably breaking droplets and forming uniform nanoemulsions [33]. On the other hand, the surfactant is also important for controlling the DS and improving the stability of nanoemulsions [34]. Thus, the influences of different HLB originating from mixed surfactants and corresponding feeding amounts were investigated on the DS and PDI.

As shown in Fig. 2a, the DS and PDI of nanoemulsions decreased with the increase of HLB from 6 to 12, while sharply increasing at the HLB of 14, indicating that the HLB of 12 was suitable for the generation of soybean oil nanoemulsions. Furthermore, the effect of surfactant concentration was shown in Fig. 2b. Both the DS and PDI of nanoemulsions decreased with increasing surfactant concentration from 1% up to 6%. However, the DS didn’t decrease anymore, but PDI remarkably increased as the surfactant concentration increased to 8%. It is speculated that the increased surfactant amount boosts emulsification. On the contrary, the excess surfactants could cause a rearrangement and destabilization of the colloidal system after achievement of the critical micelle con-centration[35]. As a consequence, the optimal formulation was constituted with the HLB of 12 and the surfactant concentration of 6% for minimizing the DS and PDI. One-way ANOVA for HLB and surfactant concentration were given in Table S2. Significant differences were observed between HLB and surfactant concentration (p < 0.05).

Fig. 2.

Influence of HLB (a) and surfactant concentration (b) on the DS and PDI of nanoemulsions prepared by umFlow-D (Other conditions: A/O ratio of 9, residence time of 1 min, ultrasonic frequency of 20 kHz and ultrasonic power of 40 W).

3.2. Effect of A/O ratio, residence time, and ultrasonic power on DS and PDI

Fig. 3a shows the influence of the A/O ratio on the DS and PDI of nanoemulsions. With the increase of the A/O ratio, the DS initially decreased and then remained invariability as the A/O ratio reached 15. In contrast, the PDI exhibited a clear inflection point with a V-type curve, in which the lowest PDI value was achieved at the A/O ratio of 9. It is mainly due to the relatively excess coverage of surfactants on the surface of droplets, which induced grave aggregation under the high A/O ratio. Fig. 3b shows that the DS decreased slowly with the residence time, while PDI decreased significantly as the residence time increased from 0.5 min to 1.0 min. It demonstrates that sonication can break larger oil droplets, and then boost the formation of uniform nanoemulsions in a short residence time [36]. In another word, the PDI had the minimum value in 1.0 min and then fluctuated slightly with the residence time. It is mainly attributed that the increasing temperature causes the droplet coalescence during the continuous sonication [37]. Furthermore, the effects of ultrasonic power on the DS and PDI of nanoemulsions are shown in Fig. 3c. When the ultrasonic power increased, the DS and PDI of nanoemulsions gradually decreased and then kept constant over 40 W of ultrasonic power, indicating that the sufficient emulsification occurred at 40 W of ultrasonic power. As a result, the minimum average DS and PDI reached 110.7 nm and 0.21, respectively. Optical photographs of the above samples were displayed as Fig. 3d, in which the calamine blue could be clearly observed in a wide preparation range, indicating the favorable DS and PDI of nanoemulsions. One-way ANOVA for A/O ratio, residence time, and ultrasonic power were given in Table S3. Significant differences were observed in the A/O ratio, residence time, and ultrasonic power (p < 0.05).

Fig. 3.

Effects of the A/O ratio (a), residence time (b), and ultrasonic power (c) on the DS and PDI of nanoemulsions prepared by umFlow-D. Correspondingly, the optical photographs (d) of comparably dulited nanoemulsions prepared with different conditions. (Other conditions include: (a) residence time of 1 min and ultrasonic power of 40 W; (b) A/O ratio of 9 and ultrasonic power of 40 W; (c) A/O ratio of 9 and residence time of 1 min.

3.3. Comparison of emulsifications with various ultrasonic reactors

Fig. 4a-b shows the effects of the residence time using an ultrasonic probe or a umFlow-D of USMR on the DS and PDI of emulsions. For the ultrasonic probe, there were still a few oil droplets floating on the upper layer of emulsions even though the DS rapidly decreased to 146 nm for 4 min of the residence time (Fig. S1). With the further increase of residence time, the DS of emulsions slightly decreased. The floating oil droplets disappeared at 7.5 min of residence time (Fig. S2) and the average DS of the emulsions reached 131 nm. The findings are ascribed to the rapidly decreasing intensity of ultrasound with the distance to the sonotrode in batch mode, longer times should be required to achieve superior emulsification and homogenization. Correspondingly, the PDI of emulsions decreased from 0.64 to 0.21 when the residence time was extended from 0.5 min to 5 min. Further increasing the residence time, the ultrasonically-induced heat accumulation led to a slight increase in PDI. In contrast, the lower DS (110 nm) and PDI (0.21) of nanoemulsions were achieved by using the umFlow-D for 1.5 min of residence time. It is therefore reasonable to conclude that the umFlow-D, which concentrates ultrasonic energy into microchannels, has highly efficient emulsification and produces superior nanoemulsions compared to conventional ultrasonic probes. These remarkable features can be attributed to the following factors: (1) the highly focused ultrasonic energy into the millimeter-level channel with the favorable and homogeneous transmission; (2) rapid heat transfer reducing the coalescence of nanoemulsions during the continuous flow preparation.

Fig. 4.

Influences of the residence time on DS (a) and PDI (b) of emulsions prepared with different ultrasonic reactors (Conditions: the ultrasonic probe performed with the ultrasonic frequency of 24 kHz, volume of 20 mL, A/O ratio of 9, and ultrasonic power of 56 W (5 s on; 5 s off). umFlow-D performed in continuous operation with the A/O ratio of 9, the ultrasonic frequency of 20 kHz, and the ultrasonic power of 40 W).

As the aforementioned above, the superior emulsification performance with the USMR has been demonstrated, but the throughput needs to be further extended. Therefore, we attempted to employ both the numbering-up and scale-up strategies for increasing the throughput of emulsification. Considering the possibility of uneven flow rate and the ratio of A/O in the case of parallel setup, the numbering-up in series was adopted to ensure the constant flow rate and the ratio of A/O. The numbering-up with different numbers of umFlow-D correspondingly increased the total flow rate while keeping the residence time constant. As shown in Fig. 5a, the DS and PDI of nanoemulsions prepared by two and four umFlow-D in series indicate the excellent magnification effect with the series connection.

Fig. 5.

(a) The DS distributions and (b) the corresponding cumulated volumes of nanoemulsions prepared by umFlow-D for numbering-up and umFlow-Z for scaling-up. (The emulsification conditions are listed in Table 3).

In addition, umFlow-Z with a larger internal volume and ultrasonic power was employed to produce nanoemulsions and tentatively compare the differences with umFlow-D. As shown in Fig. 5a and Table 3, when the A/O ratio of 9:1, the ultrasonic power of 160 W, and the residence time of 0.7 min was adopted, the DS and PDI of the nanoemulsions prepared by umFlow-Z were close to those by using umFlow-D. In addition, the influences of the residence time on DS and PDI of nanoemulsions were shown in Fig. S3 and Table S4, where the superior emulsions could be obtained for umFlow-Z at a residence time of 1.0 min.

Table 3.

DS and PDI of nanoemulsions prepared with different USMRs (A/O ratio: 9/1).

	Residence time (min)	Ultrasonic power (W)	Total flow rate (mL/min)	DS (nm)	PDI
umFlow-D × 1	1	40	0.66	119.8	0.211
umFlow-D × 2	1	40	1.32	121.3	0.239
umFlow-D × 4	1	40	2.64	119.7	0.225
umFlow-Z	0.7	160	20	120.5	0.219

This finding implies that emulsification is mainly influenced by the emulsion formulation with sufficient energy input and adequate shearing of the droplets, which is consistent with the previous studies [38], [39], [40]. Impressively, the total flow rate of umFlow-Z is about 30 times higher than that of single umFlow-D, which can be up to 20 mL/min. As a result, both the numbering-up and the scale-up of USMR are feasible to produce uniform nanoemulsions, but the magnifying differences of umFlow-D and umFlow-Z should be further studied and discussed from multiple perspectives such as energy efficiency and emulsification performance.

3.4. Fitting the model

The effect of A/O ratio (X₁), residence time (X₂), and ultrasonic power (X₃) on DS (Y₁) and PDI (Y₂) are summarized in Table S5. The experimental data was used for the calculation of coefficients of the polynomial equation. Regression equations for different responses and the results of the statistical analysis are shown in Table S6. The quadratic polynomial model can be used to represent experimental data (p < 0.0001). The values of the coefficient of determination (R²) for Y₁ and Y₂ are 0.9794 and 0.9417, respectively, indicating the model is well-fitted. The difference between Adjusted R² and R² reach 0.0265 and 0.075, respectively, which are<0.2, indicative of a strong correlation between observed and predicted values. The lack of fit is non-significant (p ≥ 0.05), which indicates that the equations are relatively well simulated and can be well analyzed [39].

The X₁ model term in the DS equation is significant (p < 0.05), while the rest of the terms are insignificant (p > 0.05). The test results of the PDI show that the correlation coefficients of X₁, X₂, X₁X₃, X₁², X₂², and X₃² are significant (p < 0.05).

3.5. Effect of independent variables on nanoelmusions

3.5.1. Ds

Fig. 6 shows the 3D response surface and contour plots of the effects of the independent variables and their interactions on the DS and PDI of nanoemulsions. For DS, when residence time and ultrasonic power were fixed, the DS of nanoemulsions decreased significantly with the increase of the A/O ratio (Fig. 6a and b). Because of the higher ratio of the aqueous phase, sufficient surfactants can efficiently decrease interfacial tension and then makes it easier to break the droplets. It can be seen in Fig. 6c that residence time and ultrasonic power within the selected range have little effect on the DS of nanoemulsions. During the breakage of the droplets, the ultrasonic energy-induced emulsification is tightly dependent on the residence time and ultrasonic power.

Fig. 6.

3D response surface plots of the combined effects of the A/O ratio, residence time, and ultrasonic power on the DS (a-c) and PDI (d-f) of nanoemulsions prepared by umFlow-D. (Ultrasonic power was held in 45 W for a and d; Residence time was held in 90 s for b and e; A/O ratio was held in 9 for c and f).

Noted that, the ultrasonic energy input should be moderate, which dominated the breakage of droplets rather than induced other side effects such as heat accumulation. Obviously, the high heat accumulation triggered by the high ultrasonic power and long residence time accelerates the collision and aggregation of emulsion droplets. In addition, the polar groups of nonionic surfactants were dehydrated at high temperatures, and the nonionic surfactants without amphiphilic properties were separated from nanoemulsions, resulting in the severe aggregation of oil droplets [41].

3.5.2. Pdi

Generally, the PDI of nanoemulsions is<0.2, revealing the uniformity and high dispersity of the emulsions. Fig. 6d shows the PDI dependence of emulsions on the A/O ratio and residence time. The higher values of the A/O ratio and residence time led to a smaller the PDI of nanoemulsions. The surfactants in the aqueous phase can quickly cover the surface of oil droplets to reduce interfacial tension and the sufficient homogeneization after long ultrasonic residence time, whose contribute to forming nanoemulsions with narrow DS distribution [42], [43]. Furthermore, more aqueous phases can increase the distance between oil droplets. The repulsive force between droplets is conducive to the formation of uniform and stable nanoemulsions [12]. The interaction of the A/O ratio and ultrasonic power is shown in Fig. 6e. Increasing the ultrasonic power could enhance the behavior of bubbles such as the growth and collapse velocity. This may enhance the shock wave and shear force generated when the bubble imploded, promoting the breakage of the oil droplets and then forming nanoemulsions with a low PDI. Meanwhile, sonication can also cause droplet break-up and recoalescence [44]. The mechanical vibration and heating caused by sonication will enhance the collision and coalescence of oil droplets in water [36]. It will raise the PDI of nanoemulsions because of the accelerated separation of oil and water. As shown in Fig. 6f, when the ultrasonic power remained constant, the PDI of nanoemulsions decreased with the residence time. If the residence time in USMR is too short, the oil phase and the aqueous phase cannot be sufficiently emulsified and flowed out from the USMR, leading to the increase of PDI of nanoemulsions.

3.6. Determination and verification of optimal parameters

Based on the target expectations for the DS and PDI of nanoemulsions, the emulsification process was further optimized with RSM. For the formation of uniform nanoemulsions with small DS and narrow granulometric distribution, the optimal parameters on the selected range in umFlow-D were listed as follows: the A/O ratio of 10.1, the residence time of 114 s, and the ultrasonic power of 50 W. The predicted DS and PDI were 94.4 nm and 0.193, respectively. Experimentally, the optimal nanoemulsions with the edge of calamine blue were achieved under the accordant performance of the above parameters (Fig. 7a), exhibiting DS of 96.4 nm and PDI of 0.195 (Fig. 7b and Table 4).

Fig. 7.

Optical photograph of nanoemulsions prepared by umFlow-D under optimal conditions (a). The DS distribution and optical photograph of the nanoemulsions diluted with DI water (b). TEM images of the optimal nanoemulsions with different magnifications (c, d). (Optimal conditions: A/O ratio of 10.10, residence time of 114 s, and ultrasonic power of 50 W).

Table 4.

Predicted optimal and experimental results.

	A/O ratio	Residence time (s)	Ultrasonic power (W)	DS (nm)	PDI
Predicted parameters	10.14	113.93	50	94.4	0.193
Experimental parameters	10.10	114.00	50	96.4	0.195

As listed in Table 4, the experimental results were well in agreement with the predicted values. After dilution, a clear and light blue emulsion is observed for the diluted nanoemulsions (inset of Fig. 7b). Furthermore, as shown in Fig. 7c, TEM images of the optimal nanoemulsions showed high dispersity and spherical morphology. Furthermore, the enlarged image exhibited a uniform particle with a metrical size of ∼ 96 nm, consistent with DLS result (Fig. 7d).

3.7. Stability of nanoemulsions

In this study, the stability of the nanoemulsions prepared by umFlow-D was studied by evaluating the DS and PDI for 45 days. As shown in Fig. 8a, after 45 days of storage in 4℃, the average DS changed from 96.6 nm to 97.1 nm, and PDI changed from 0.189 to 0.164. After 45 days of storage in 25℃, the average DS changed from 96.4 nm to 97.4 nm, and PDI changed from 0.195 to 0.194 (Fig. 8b). The experimental results show that the nanoemulsions prepared by USMR held well and had superior stability at different temperatures.

Fig. 8.

The stability evaluation of nanoemulsions stored at 4℃ (a) and 25℃ (b). (Conditions: A/O ratio of 10.10, residence time of 114 s, and ultrasonic power of 50 W).

In addition, the recently reported DS, PDI, and stability of soybean oil nanoemulsions prepared by using different techniques were summarized in Table 5 [44–52].

Table 5.

Comparison of quality and scale of soybean oil nanoemulsions prepared by various emulsification methods. (HSH: high-speed homogenizer; HPH: high-pressure homogenize; UAE: ultrasound-assisted emulsification; MF: Microfluidizer).

Method	Surfactant and concentration	Scale* (L h−1)	DS (nm)	PDI	Stability (d)	Reference
HSH	Myofibrillar protein (1%)	1.5	2400	–	0.8	[28]
HPH		0.4	700	–	2	[28]
UAE		0.3	150	–	7	[28]
HPH	Tween-20 and Lecithin (4%)	2–2.4	146	0.27	10	[45]
HPH	Gelatin (7%)	1.2	137	0.065	–	[46]
	F127 (8%)	1.2	130	0.081	–	[46]
	F68 (10%)	1.2	136	0.159	–	[46]
MF	Lecithin (1.2%)	5	280	–	–	[27]
MF	Quillaja saponin (0.5%)	2.4	500	–	–	[47]
HPH	Quillaja saponin (1.5%)	0.6	120	0.12	35	[48]
UAE	PC and OL-ECG (2.5%)	0.1–1	290	0.252	–	[49]
UAE	Tween-20 and Tween-80 (10%)	–	216	0.229	28	[50]
HSH	Tween-20 (8%)	0.6	710	–	–	[51]
USMR	Tween-80 and Span-80 (6%)	1.2	96	0.195	45	This work

* Note: Scales were estimated based on the number of homogenization and processing volume depended on the given data of reference or the output parameter of commercial device in the corresponding reference.

Compared with previous nanoemulsions, the USMR-induced soybean oil nanoemulsions had a favorably small DS (<100 nm), good monodispersity (PDI < 0.2), and superior stability for 45 days storage with a relatively low dosage of surfactants. In addition, the scales of different emulsification methods were also compared. As seen in Table 5, the scale of USMR was lower than that of MF, but it was obviously higher than that of frequently used UAE and comparable to HSH and HPH. Comprehensively, USMR exhibited an excellent emulsification effectiveness without pre-emulsification while the scale should be increased by numbering-up. Therefore, the emulsification via USMR is a promising process for the research and development of nanoemulsions in the future.

4. Conclusions

In this work, we employed a typical USMR (umFlow-D) with 660 μL of internal volume in which O/W nanoemulsions were continuously produced in a single cycle and without any pre-emulsification treatment. Compared with ultrasonic probe, the USMR exhibited concentrated energy input, superior emulsification performance, and easy scaleup with remarkable uniformity by series mode. In addition, another USMR (umFlow-Z) with large internal volume and high ultrasonic power was also utilized to produce the concurrent and uniform nanoemulsions, leading to a high throughput in comparison with umFlow-D. In future, more or long snake-formed microchannel should be set on the surface of a piezoelectric transducer for the scale-up. After optimizing the USMR process with RSM, the DS of 96.4 nm and PDI of 0.195 were observed, well consistent with the predicted values. Impressively, the nanoemulsions prepared by the USMR exhibited superior stability and held well for 45 days at 4℃ and 25℃. The present work may provide a typical paradigm using the novel and efficient USMRs as an emulsification tool for the preparation of functional nanomaterials.

CRediT authorship contribution statement

Jiahong Xu: Conceptualization, Methodology, Writing – original draft. Xiaojing Zhu: Conceptualization, Writing – review & editing. Jie Zhang: Investigation. Zhipeng Li: Resources. Wenjiang Kang: Investigation. Haibo He: Resources. Zhilin Wu: Writing – review & editing. Zhengya Dong: Supervision, Funding acquisition.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following are the Supplementary data to this article:

Data availability

Data will be made available on request.