Droplet size is an important indicator of emulsion stability. Results of this study show that the encapsulation efficiency of microencapsulates can be improved by decreasing the emulsion droplet size as the larger droplets may be breakdown during atomization leading to the higher surface oil (Tonon et al., 2011). The small oil droplets can be enclosed efficiently within the wall matrix of microcapsules (Jafari et al., 2008). As previously explained, emulsions were produced by using various emulsifiers. Four HLB emulsions with HLB in the range of 8–15 were prepared for the microencapsulation of S. wilsoniana oil. Table 1 presents the droplet mean diameters for each of these emulsions. Droplet mean diameter varied from (4.43 ± 0.017) μm to (29.99 ± 0.026) μm. The emulsions had relatively small droplet sizes except where the emulsifier was soy lecithin; in that case, droplet size was (29.99 ± 0.026) μm. The emulsions prepared by the mixture of Tween 80 and soy lecithin have relatively small droplet size, which may be attributed to the synergy between them: L and T can get stable at the oil-water interface, and the large headgroup of T prevents oil droplets coalescence by steric repulsions.
Emulsifier Emulsion size (D43, μm) Creaming stability (CI) Tween 80 5.67 ± 0.020 (45.27 ± 1.30)% Tween 80/soy lecithin (6/4) 4.43 ± 0.017 (10.89 ± 2.43)% Tween 80/soy lecithin (4/6) 5.01 ± 0.015 0 Soy lecithin 29.99 ± 0.026 0
Table 1. Droplet size distribution and creaming stability of Wilson's dogwood (Swida wilsoniana) oil emulsions with different combinations of emulsifier (Model 1)
Emulsion stability is important in microencapsulation. We found that emulsion stability was markedly affected by the type of emulsifier used, which further influenced the characteristics of the spray dried powder. Emulsion stability was measured using a CI (Table 1). A gradual phase separation occurred after 2 h in both Tween 80 and Tween 80/soy lecithin (6/4). After 24 h, a homogenous original phase fraction of (45.27 ± 1.30)% and (10.89 ± 2.43)% occurred, which were not suitable for spray drying. The other emulsions remained stable (CI = 0) for a 24 h period, which could be due to the formation of network structures in the emulsions (Ye and Singh, 2006). The results showed that neither T nor L was effective emulsifiers for spray-dried S. wilsoniana oil. A combination of T and L, especially at a weight ratio of 4/6, however, did provide an effective oil-in-water (O/W) emulsifier, were capable of forming stable emulsions that resist creaming through the synergy of L and T at the oil-water interface.
In processing, emulsion viscosity is important because it affects encapsulation efficiency and oil retention. The viscosities of the S. wilsoniana oil emulsions prepared with lactose and gum Arabic (as wall materials) in combination with various ratios of T and L (as emulsifiers) are shown in Fig. 2. All emulsions behaved like non-Newtonian fluids when subjected to shear stress below 25 S–1 with a shear-thinning behavior. The viscosity was fairly insensitive at very high shear rates, indicating Newtonian behavior (Bakry et al., 2016). The T/L (6/4) emulsion had the lowest viscosity, which may be due to the differences in chemical composition, and potentially resulting in changes in emulsions properties (e.g., droplet size, interfacial properties) and increased viscosity. A moderate viscosity is desirable for spray drying. Increasing the viscosity to an optimal value can reduce the oscillation and circulation of internal droplets, resulting in improved oil retention (Jafari et al., 2008). Maximum viscosity was reached with the incorporation of T/L (4/6) surfactant.
Moisture content is important in the powder-quality and shelf life of spray-dried products. Excessive moisture in these products causes powder agglomeration, microbial growth, and caking, which result in the release/oxidation of the oil. Table 2 shows that the moisture content of spray-dried microcapsules varied with the wall material from (1.08 ± 0.215)% to (2.89 ± 0.294)%. The moisture content of the sodium caseinate/lactose blend was the highest, while that of the gum Arabic/lactose blend was the lowest. This variation in moisture content between powders is related to the chemical structures of the different wall materials. Compared with other spray-dried microcapsules, the moisture content in our products was relatively low, which may be attributed to the high inlet and outlet air temperatures during spray drying. The relatively low moisture content was expected to minimize oxidation and microbial contamination of the core material.
Wall material Moisture content (%) Particle size distribution (μm) Gum Arabic/lactose 1.08 ± 0.215 8.69 ± 0.023 Sodium caseinate/lactose 2.89 ± 0.294 7.43 ± 0.020 Soy protein isolate/lactose 1.34 ± 0.194 3.21 ± 0.025
Table 2. Moisture content and particle size distribution of microcapsules using different wall materials (Model 2)
Encapsulation efficiency refers to the potential of the wall material to encapsulate the core material inside the microcapsule; Oil loading is one of the most important parameters of microencapsulated powders. Since oil on the microcapsule surface is easily oxidized, the surface and total oil content of the microcapsules were used to evaluate encapsulation efficiency. As shown in Table 3, the use of compound surfactants (T/L, 4/6) markedly improved encapsulation efficiency (91.06 ± 0.96)%, indicating that only a little of oil remained unencapsulated in the Tween 80/soy lecithin (4/6) emulsified microcapsules. Furthermore, the payload of S. wilsoniana oil microcapsules with the Tween 80/soy lecithin (4/6) was (35.53 ± 2.21)%. This suggests that nearly all the oil is encapsulated within the shell matrix. The combinations (Tween 80/soy lecithin) are more stable because T/L blends have a synergistic effect, that is, in combination, they offer an effective emulsifier, which may positively influence encapsulation efficiency of the microencapsulated powders. The overall result was that the composition of emulsifier significantly influenced encapsulation efficiency and other measured data such as emulsion size. In addition, microencapsulated powders obtained from emulsions stabilized by (T/L, 4/6) had efficient encapsulation, with most of the S. wilsoniana oil being retained in the particles.
Emulsifier MEE (%) Oil loading (%) Tween 80 91.63 ± 0.93 33.81 ± 1.60 Tween 80/soy lecithin (6/4) 87.90 ± 1.02 29.00 ± 2.00 Tween 80/soy lecithin (4/6) 91.06 ± 0.96 35.53 ± 2.21 Soy lecithin 51.04 ± 0.76 27.70 ± 1.25
Table 3. Microencapsulation efficiency (MEE (%)) and payload of S. wilsoniana oil with different surfactants (Model 1)
As shown in Table 4, the MEE of encapsulated S. wilsoniana oil ranged from (85.25 ± 1.50)% to (95.20 ± 1.85)%. Sodium caseinate blends and gum Arabic blends had higher MEE (95.20 ± 1.85)% and (91.06 ± 1.06)%, which meant that both blends were effective in maintaining the S. wilsoniana oil inside the microparticles during spray drying, thus limiting oxidation during processing and storage. The sodium caseinate blends had higher MEE than gum Arabic and soy protein isolate blends, may due to the reaction between the amino groups of sodium caseinate and the carbonyl groups of lactose. These groups can form conjugates through the Maillard reaction. The Maillard reaction stabilizes oil microcapsules by changing the physical properties of the wall and contributing to wall formation (Ferreira et al., 2016). Such changes in the wall retain and protect the core material, thus increasing the MEE. Another reason for the higher MEE in the sodium caseinate blends is the susceptibility of the soy protein isolate to heat denaturation during spray drying (Rodea-González et al., 2012). The payload of gum Arabic blends was the most (35.53 ± 2.21)%, which meant the microparticles produced using gum Arabi had more total oil and less surface oil than those made with sodium caseinate or soy protein isolate. The surface oil was the most important factor that affected the microparticles stability since the oil droplets on the surface of the powders were easily to be oxidized.
Wall material MEE (%) Oil loading (%) Gum Arabic/lactose 91.06 ± 1.06 35.53 ± 2.21 Sodium caseinate/lactose 95.20 ± 1.85 32.65 ± 2.05 Soy protein isolate/lactose 85.25 ± 1.50 32.00 ± 1.62
Table 4. Microencapsulation efficiency (MEE (%)) and payload S. wilsoniana oil with different wall materials (Model 2)
The sizes of the spray-dried microcapsules are shown in Table 2. Wall material with soy protein isolate had significantly (P < 0.05) smaller particle size than that of the other two wall materials. Particle sizes of spray-dried microcapsules with the three wall materials were in the range of 3.21–8.69 μm. The sodium caseinate/lactose blend with particle size of (7.43 ± 0.020) μm had the highest MEE, while S. wilsoniana oil encapsulated with soy protein isolate/lactose had smaller particles sizes, which may be attributed to the soy protein isolate and lactose exhibited much higher ability to be redispersed in water when reconstituted emulsions (Tang and Li, 2013). Smaller microcapsules possess larger surface area which may cause more oil for migrating to the surface of the microcapsule during the spray drying or even during the storage period (Botrel et al., 2017). On the other hand, the decrease in MEE value of the SL blend maybe also due to heat denaturation of the soy protein isolate during spray drying (Lim et al., 2012).
Through SEM observations (Fig. 3a and b), it is found that microcapsule particles encapsulated with gum Arabic/lactose had smooth surfaces. However, in the soy lecithin emulsified particles, there were other larger, irregularly shaped agglomerates. The large agglomerates may result from the high viscosity of the feed emulsion or from the potential for partial flocculation of the feeding emulsion, which can lead to agglomerate formation during spray drying due to poor emulsion stability (Silva et al., 2016). The SEM images of S. wilsoniana oil microcapsules with various wall materials are shown in Fig. 3b–d. The particles had smooth surfaces with no apparent cracks to permit the entry of gases and solvents, thus increasing the MEE (Rodea-González et al., 2012). Microcapsules obtained from soy protein isolate and lactose were much smaller than those produced using the other two wall materials. Large and small particles were formed, consistent with the dynamic light scattering data.