Cold-pressed flaxseed oil and coconut oil were obtained as pure (crude) oils from the oil extraction unit at the National Research Centre (NRC), Dokki, Cairo, Egypt. Tween 20 (Polysorbate; Polyoxyethylene (20) sorbitan monolaurate) and sodium dodecyl sulfate (SDS) were purchased from Sigma Chemical Company. Pure soybean lecithin with 70% phosphatidylcholine (Lipoid S 75) was purchased from Lipoid GmbH, Ludwigshafen-Germany. Rosemary (Rosmarinus officinalis L.) leaves were purchased from a local herb store in Cairo, Egypt. All solvents and reagents were of the analytical grades needed for each application.

Table 1 shows the fatty acid compositions of flaxseed and coconut oils and their oil blend. Saturated fatty acids were found to be the major fatty acids in coconut oil 91%, where lauric acid constituted 48%, caprylic 8.5%, capric 7.5%, myristic 16% and palmitic acid 9%. Unsaturated fatty acids constituted only 9% with 7% oleic acid and 2% linoleic acid. On the other hand, unsaturated fatty acids were predominant in flaxseed oil, where linolenic acid (ω-3) was the major fatty acid at 48.2%, with oleic at 22%, linoleic at 18.3% and palmitic acid 7.1%. As a result of blending, large variations were noticed in the contents of lauric, oleic, linoleic and linolenic acids of the starting oils. Mixing flaxseed oil with coconut oil at a ratio of 1:1 wt/wt caused a significant (p < 0.05) increase in the total content of saturated acids (from 11.5% to 50.73%) and a significant (p < 0.05) reduction in the total content of unsaturated fatty acids (from 88.5% to 49.27%) as observed in Table 1. The results indicated that total saturated acids were increased by more than 440%, whereas total unsaturated acids decreased by more than 55%. Such significant changes in total saturated and unsaturated acids could enhance the oxidative stability of the formulated nanoemulsion (Hamed and Abo-Elwafa, 2013; Bhardwaj et al., 2015). Bhardwaj et al., (2015) concluded that blending flaxseed oil with oils rich in saturated fatty acids greatly enhanced its resistance against oxidation by decreasing the content in polyunsaturated fatty acids (which are more sensitive toward oxidation than saturated fatty acids).

Formulation, stability, as well as physicochemical characteristics depend on the emulsifier as a corner stone of the whole emulsion system. Food and nutrition sectors use different types of emulsifiers. Phospholipids, proteins, and polysaccharides are used extensively as emulsifiers for nutraceuticals and food industries (Benichou et al., 2004; McClements, 2015). Therefore, the effect of type and concentration of emulsifier on nanoemulsion formation using high-pressure homogenization is an important target in optimizing the proper methodology.

In our present work we examined the effect of three different types of emulsifiers on the droplet size: Zwitterion surfactant (Lecithin), a non-ionic surfactant (Tween20) and an anionic surfactant (sodium dodecyl sulfate, SDS). 5 wt% (flaxseed:coconut, 1:1 wt/wt) oil/water nanoemulsions with different emulsifier concentrations (2–8%) of the above-mentioned emulsifiers (Lecithin, Tween 20, or SDS) were formulated with the high-pressure homogenizer. Homogenization was carried out under high pressure (1000 bar) for 6 successive cycles. The droplet sizes were then measured.

The results (Table 2 and Figure 1) show the effect of emulsifier type and concentration on the droplet sizes. As shown, at 2% emulsifier concentration, the droplet sizes were 401.1, 179.9 and 110 nm for lecithin, Tween 20 and SDS, respectively. Variations in droplet size may be returned to variations in the molecular weight of the emulsifier used (Lecithin, 643.887g/mol; Tween 20, 522.676 g/mol and SDS, 288.38 g/mol). Our results agreed with previous researchers (Karbstein and Schubert, 1995; Qian and McClements, 2011) who reported that by using fixed concentration and homogenization parameters, smaller molecular weight emulsifiers (e.g. SDS and Tween 20) showed superior efficiency to produce smaller droplets than larger molecular weight emulsifiers. Such efficiency may be due to the faster adsorbability of small molecules on the droplet surface.

Regarding the influence of emulsifier concentration, Table 2 shows that, for Tween 20, an increase emulsifier concentration from 2 wt% caused an insignificant decrease in the droplet diameter (p < 0.05). As for SDS the lowest droplet size (110 nm) was produced at 2% concentration, while as the concentration increased a significant increase of more than 4% in droplet size occurred. On the other hand, lecithin-containing nanoemulsion showed a decreasing order of droplet size as the concentration of the emulsifier increased until the concentration reached 6% when a sudden and noticeable increase was registered thereafter (Table 2).

Although it was theoretically expected that the droplet size would be decreased significantly for all the emulsifiers with increasing surfactant concentration so as to cover all oil droplets produced during homogenization, this did not actually happen, as shown in Table 2. This unexpected behavior may be interpreted either by proposing that not all emulsifier molecules are sufficiently adsorbed onto oil droplets or there was an insufficient disruptive force of the homogenizer at the operating pressure (Qian and McClements, 2011).

From a practical point of view, both SDS and lecithin were excluded for two reasons. First, although SDS showed the smallest droplet size (110 nm) it was excluded due to some health hazards for using it in the food sector (Anderson et al., 1986; Lindberg et al., 1991; Asonye et al., 2015). Second, if lecithin were used as the emulsifier it’s concentration would have to be increased to reach a desirable droplet size which means more cost for producers who tend to use smaller emulsifier concentration due to regulatory and economic reasons. Therefore, using Tween 20 at 2% concentration satisfied the previous desired characteristics and at the same time was effective in generating nanoemulsions with acceptable droplet size (179–174 nm). Tween 20 was then used throughout the study.

Physicochemical properties and emulsion stability are greatly affected by emulsification parameters such as homogenization cycles and the pressure used inside the homogenization chamber (Yuan et al., 2008). Therefore, these parameters were investigated to assess their influence on the physicochemical properties of the ω-3 based oil-in-water nanoemulsions.

Figure 2 illustrates that as the homogenization pressure or number of homogenization cycles increased the droplet went to a smaller diameter. There was a dramatic decrease in mean droplet diameter when the pre-emulsion passed the HPH at all tested pressures (Figure 2). Diameters decreased from 928.8 nm (Control Sample, data not shown on figure) to 214.8, 238.2, and 199.7 nm after one cycle at homogenization pressures of 500, 1000 and 1500 bar, respectively. This result is considered clear proof of the effective power of HPH to formulate nanoemulsions. Our results agreed with those previously reported (Qian and Mc Clements, 2011; Tan and Nakajima, 2005).

As for the effect of number of passes there was a pronounced decrease in droplet diameter with increasing passes from a single pass until the pass number six for all used pressures. Increasing homogenization more than six passes resulted in a moderate decrease in droplet diameter (Figure 2). As seen in Figure 2, there was no noticeable difference between the mean droplet diameter after six passes and emulsions prepared at 1000 bar and at 1500 bar. Also, Figure 2 shows that increasing the number of passes from six to eight at 1500 bar resulted in a very slight decrease in droplet diameter. Therefore, six homogenization cycles (passes) at 1000 bar was the method of choice to be proceeded with throughout the entire study.

The polydispersity index (PDI) is a dimensionless (see section 2.2.5) parameter and is represented in such a way that values smaller than 0.05 are rarely seen and values greater than 0.7 have a broad-size distribution and are not suitable to be analyzed by the dynamic light scattering technique (DLS). Algorithms with different size distributions work with those data that fall between these two extreme PDI values (i.e., 0.05–0.7). The calculations used for the determination of size and PDI parameters are defined in the ISO standard documents 13321:1996 E and ISO 22412:2008 (Malvern Instruments Worldwide, 2011).

The droplet size distribution of the prepared O/W nanoemulsion is shown in Figure 3 and Table 3. This distribution was uni-modal with accompanied polydispersity indices ranging from 0.123 to 0.334. The results from the PDI show that all tested homogenizer pressures (500–1500 bar) and number of homogenization cycles (1 - 8 cycles) gave mono-modal dispersions of the produced droplets.

The influence of the different processing conditions (homogenization pressure and homogenization cycles) on Zp was also evaluated. The zeta potential is a stability indicator of colloidal dispersion systems. The magnitude of the zeta potential indicates the degree of electrostatic repulsion between adjacent, similarly charged particles in dispersion. So, for particles and molecules that are very small, the higher the magnitude of zeta potential the more stable is their colloidal system (i.e., the solution or dispersion will resist aggregation and flocculation). When Zp is small, attractive forces will exceed this repulsion and the dispersion may break and flocculate. As presented in Table 3, Zp values ranged from -29.05 mV to -58.49 mV for nanoemulsions produced with Tween 20 under the different homogenization conditions tested. A zeta potential of ±30 mV is believed to be sufficient for ensuring the physical stability of nanoemulsions (Gurpreetand Singh, 2018). Such high Zp values obtained for the prepared nanoemulsions using Tween 20 as the emulsifier may be attributed to the composition of its backbone oil (free fatty acids and salts). McClements (2005) reported that even at equilibrium, there was a continuous exchange of molecules between the interior of the droplets, the continuous phase and the interface.

The influence of the continuous phase (aqueous phase) viscosity on both droplet diameter and PDI of the produced nanoemulsions was investigated at fixed conditions of homogenization (1000 bar, 6 passes) and 2% Tween 20 as emulsifier. The oil phase was kept as is (flaxseed oil: coconut oil, 1:1wt/wt, at 5 wt% of total emulsion) but the aqueous phase was formulated to contain varying amounts of glycerol (0 to 50% of the aqueous phase) to change its viscosity before homogenization. As the glycerol content increased a parallel increase in the viscosity of the whole nanoemulsion was noticed (Table 4). As shown in Table 4, the nanoemulsion produced without glycerol (100% water) showed the smallest droplet size (172.6 nm) as well as the narrowest range of size distribution (PDI = 0.123). As the viscosity of the continuous phase increased by increasing glycerol concentration, fluctuation (irregular increase and decrease) in droplet size was observed (Table 4). The results of the present study disagreed with the results of Ramisetty and Shyamsunder (2011) who reported that droplet size increased by increasing the viscosity of the continuous phase. On the other hand, Qian and McClements (2011) revealed that using b-lactoglobulin as the emulsifier produced a slight change in droplet size while using small molecules (e.g. SDS) as surfactant caused a pronounced decrease in the size of the droplets with increasing aqueous-phase viscosity. Such variations in results may be due to differences in emulsifier type or homogenization equipment used.

During the oxidation of lipids there is a continuous formation of hydroperoxides (primary oxidation products) that in turn break down to a variety of nonvolatile and volatile secondary products. The formation rate of hydroperoxides exceeds their rate of decomposition during the initial stage of oxidation, and this becomes reversed at later stages (Shahidi and Zhong, 2005). Therefore, peroxide value (PV) is a very important indicator of oxidative stability. Also, it has been well established that primary oxidation products (e.g. hydroperoxide formation) increases with increasing unsaturation. For example, linoleic acid (C18:2) has been estimated to be 10~40 times more susceptible to oxidation than oleic (C18:1) and linolenic acids (C18:3, n-3) and is far more readily oxidized than linoleic acid (Yun and Surh, 2012; McClements and Decker, 2008).

The oxidative stability of bulk oils is affected by its fatty acid composition. In our present work blending flaxseed oil with coconut oil (1:1, wt/wt) was first carried out to tailor the fatty acid composition in such a way as to decrease the total polyunsaturated acids and increase total saturated acids. This was the first step to enhaninge oxidative stability (see section 3.1.). As a second step, rosemary extract (60 mg) was added to give more protection against oxidation because of its highly potent phenolic antioxidants (Kayashima et al., 2012; Nieto et al., 2011; Nieto et al., 2018). Finally, the effect of the nanoemulsification technique on the oxidative stability of the substrate oil was evaluated. As shown in Figure 4 the levels of lipid hydroperoxides formed were periodically measured for the bulk oils and the nanoemulsion stored at 25±1 °C over the entire storage period of 11 weeks. The results indicated that during the first two weeks of storage there was a slow and comparable increase in the PV of both the bulk oil and the nanoemulsion. Beginning from the 3rd week the difference between bulk oil and the nanoemulsion PV became more pronounced and the bulk oil PV (62.4 meq O₂/Kg oil) became about 3 times higher than that of the nanoemulsion (22.6 meq O₂/Kg oil) at the end of storage period. These results could be interpreted in light of the interfacial properties of the prepared emulsion. While for bulk oil the site of oxidation is usually done at interface between the air and the oil surface, the site of oxidation in the case of emulsion is the interface between the oil and water. So, the invasion of oxygen onto the unsaturated sites on the fatty acids in case of bulk oil is far easier than its invasion into those inside the emulsion. Also, enhancement in oxidation stability may be caused by the shielding effect of the emulsifier which formed interfacial barriers between the oil droplets and the invading oxygen.