Optimization of the conditions for producing water-in-oil-in-water microemulsions and spray-dried microcapsule of tomato extract powder

Tomato is traditionally recognized as the plentiful source of lycopene (Edwards et al., 2003). Lycopene is a polyunsaturated (polyene) straight-chain molecule, which contained 13 double bonds with a molecular weight of 536.9 daltons. Ot has strong antioxidant and other beneficial effects, including free radical scavenging (Hernandez-Marin et al., 2013). Antioxidant effects of lycopene might include scavenging free radicals from reactive oxygen species (RDS) because lycopene is very effective to remove singlet oxygen in in vitro and in plasma (Zhang et al., 2014). However, lycopene in tomato is hardly to obtain since tomato extract has poor water solubility and is easily oxidized by light or heat exposure. To combat those issues, microencapsulation can be a good approach and it could be a good solution to be utilized tomato extracts into food system.


Introduction
Tomato is traditionally recognized as the plentiful source of lycopene (Edwards et al., 2003).Lycopene is a polyunsaturated (polyene) straight-chain molecule, which contained 13 double bonds with a molecular weight of 536.9 daltons.Ot has strong antioxidant and other beneficial effects, including free radical scavenging (Hernandez-Marin et al., 2013).Antioxidant effects of lycopene might include scavenging free radicals from reactive oxygen species (RDS) because lycopene is very effective to remove singlet oxygen in in vitro and in plasma (Zhang et al., 2014).However, lycopene in tomato is hardly to obtain since tomato extract has poor water solubility and is easily oxidized by light or heat exposure.To combat those issues, microencapsulation can be a good approach and it could be a good solution to be utilized tomato extracts into food system.
Microencapsulation can be an approach to enhance stability of lycopene, and to enable its dispersion in an aqueous media (Rocha et al., 2012).This technology has been successfully used in the food system to preserve its core materials that are susceptive to temperature, light, oxygen and humidity.Encapsulation is one of economical packaging technique for active ingredients in foods for maximum protection, targeted delivery and controlled release of nutrients (Leclercq et al., 2009;Ubbink & Schoonman, 2013;Weinbreck et al., 2004).The development of delivering sensitive and active ingredients has been newly emerged for the food industry (Shah & Ravula, 2000;Siegrist et al., 2008).
Spray drying is the most conventionally utilized technique of encapsulation for food products.The process of spray drying is economical and flexible, readily available equipment and highly qualified (Reineccius, 1988;Rosenberg et al., 1990).The effect of suitable coating materials, preparation of emulsions with core and coating materials and the drying processes affect the shape and type of capsules formed; the efficiency and rate of retention of core materials are also influenced (Graves & Weiss, 1992).The typically used coating materials in spay-drying process are maltodextrin (MD), gum arabic (AG), milk protein, soy protein and modified starch.
On the previous study, lycopene was embedded effectively in the layers of gelatin/sucrose microcapsules by spray drying with a loading of over 80% core-to-wall ratio.The benefit originated from embedding lycopene into gelatin and sucrose matrices enhanced storage stability of lycopene equal to tomato paste (Shu et al., 2006).Rocha et al. (2012) reported that the stability of lycopene was improved by microencapsulation.Also, Pelissari et al. (2016) showed that the microparticles loaded with lycopene by spray chilling could enhance its stability approximately 90% or more.For more bioavailability, the encapsulated bioactive ingredients should be easily released from the food, and then absorbed in the human (Shu et al., 2006;Sanz & Luyten, 2006).Odeally, the spray-dried microcapsules for tomato extracts should have instant properties or served as a source of lycopene-rich functional food ingredient for supplement into food products.Recently, Pu & Tang (2017) suggested that stability of the lycopene in chlorella pyrenoidosa was enhanced by encapsulation.Shi et al. (2015) also reported that the formulation of microemulsion of lycopene might be an effective way to prevent the environmental stresses on stability.According to Celli et al. (2016), the stability of lycopene from watermelon was increased in microemulsion.However, there is limited information in the literature on water-in-oil-in-water (W/D/W) microemulsions and spray-dried microcapsules of tomato extract powder containing high amount of lycopene as functional foods.Therefore, the purpose of the present study was to optimize conditions for preparing W/D/W microemulsion and spray-dried microcapsules of tomato extract powder and investigate physicochemical characteristics of such emulsions and microcapsules.

Determination of emulsion stability index (ESI)
The volumetric method was used to elucidate the emulsion stability index (ESO) for W/D/W microemulsion of tomato extracts according to Chang et al. (1994).Briefly, the emulsion was incubated at 60 °C for 2 h, and then the separated layer was formed to measure the volume.All samples were measured in triplicate.ESO was calculated by the following formula Equation 1I:

Optimization of conditions for producing microemulsion of tomato extract by response surface methodology (RSM)
Response surface methodology (RSM) was utilized to determine the effect of three independent variables as shown in Table 1.
Twenty experiments were designed with the principal of RSM using Minitab release 16 (Minitab, 2010) and Design Expert 7 (Stat-Ease, 2007).The quadratic polynominal regression model was estimated for the effective yield of the microencapsulation.The effective yield of the microencapsulation was calculated by following Equation 2I: where, the effective yield of the microencapsulation is response, β k0 , β ki , β kiii , and β kijj are constant coefficients of intercept, linear, quadratic and interaction terms, respectively.X i and X j are uncoded independent variables (ratio of tomato extracts to MCT, emulsifier concentration, and secondary coating materials to W/D phase).The regression model from RSM was utilized to estimate the result by iso-response three-dimensional (3-D) surface plots.

Yield for microemulsification of tomato extract
The lycopene content was determined from the microcapsules of tomato extract by high performance liquid chromatography (HPLC) based on the method suggested by Lin & Chen (2003).The microencapsulation yield was calculated as followed Equation 3I: Microencapsulation yield % = total lycopene in the emulsion / total lycopene in powder 100

Spray drying
After optimization of conditions for producing microemulsion of tomato extract by RSM, the emulsions were spray-dried using lab scale spray-dryer (SD-1000, Tokyo Rikakikai Co., Ltd., Tokyo, Japan) under following conditionsI: feed flow rate was 0.5 L/h, inlet temperature was 170 °C, air flow rate was 0.70 m 3 /min, atomizing pressure was 50 kPa, and outlet temperature was 90 °C.

Optical microscopy
Dptical microscopy images for microemulsions of tomato extract was using a digital microscopy system (Eclipse 80i, Nikon, Tokyo, Japan).The microstructure of the microemulsion was magnified by 1,000-folds.

Scanning electron microscopy (SEM)
The morphological features of powdered microcapsules of tomato extracts were observed using the scanning electron microscopy (SEM, S-4300, Hitachi, and Tokyo, Japan).The SEM was operated at an accelerating voltage of 15.0 kV.

Particle size distribution
Particles size distribution of the powdered microcapsules of tomato extracts was analyzed by a Malvern Mastersizer 2000 (Malvern Onstruments Ltd., Worcestershire, UK).The suspension was poured into a cuvette and its particle size was measured at 25 °C with fixed scattering angle at 165°.All samples were measured in triplicate and the mean particle size and distribution span were computed on the basis of a volume median diameter.

Zeta-potential measurement
Zeta-potential was measured to observe electrical stability of the spray-dried microcapsules of tomato extracts using a particle size analyzer (Delsa nano C, Beckman coulter, Onc., Fullerton, CA).All measurements were done in triplicate at fixed temperature and an angle of 25 °C and 15 °C, respectively.

Statistical analysis
Analysis of variance (ANDVA) and Duncan's multiple tests were performed to analyze the differences between groups.Data was expressed as mean ± SD and statistical significance was set at p<0.05.All statistical analyzes were done using SAS 9.0 (SAS, Onstitute Onc., NC, USA) and the graphical representations were performed using Sigmaplot 10.0 (Systat Software Onc., San Jose, CA, USA).

Determination of ESI and emulsifier for producing W/O/W microemulsion of tomato extract
According to Almeida et al. (2015), most of common indicators for investigating deterioration of an emulsion are layer separation, and the emulsion stability is the most important among main factors which could be influenced to final products.On general, emulsion stability is influenced by type or amount of emulsifier.To obtain stable W/D/W microemulsions of tomato extract, the simple and rapid method was needed to evaluate emulsion stability.Therefore, ESO measurement could be a suitable method in this study.The results of ESO, which is investigated to optimize the ratio of W/D to aqueous phase and to select the suitable emulsifier, are shown in Figure 2. Ot is known that emulsifiers with HLB value higher than 10 are predominantly hydrophilic, thus favouring formation of W/D/W emulsion.Tween 60 (HLB 14.9) at ratios of 1I:7 and 1I:9 for W/D to secondary coating materials showed more than ESO of 96%.
Similarly, Spernath et al. (2002) have reported that microemulsions of lycopene can be stabilized by three emulsifiers (Tween 60, sucrose fatty acid ester, and ethoxylated monodiglyceride).However, only Tween 60 increased the solubilization of lycopene (Spernath et al., 2002).Results of the present study also indicated that Tween 60 (HLB 14.9) at ratio of tomato extract to MCT of 1I:9 in W/D/W could be useful for generation of W/D/W microemulsions of tomato extract.Ot was reported that an increase of coating material led to an increase in the microencapsulation yield and stability of core material.Wu et al. (2014) studied microencapsulation of sulforaphane by spray-drying.On the result, as the ratio of coating material was increased and the ratio of core material decreased, the yield of microencapsulation and core material stability increased.Furthermore, Nhu Quynh et al. (2016) reported that as the amount of wall material increased, the microencapsulation yield increased in the Gac oil microencapsulation by spray-drying.Although this study has limitation which it did not optimize condition of spray-drying, it is desirable to increase the ratio of coating material from the results of these previous studies (Wu et al., 2014;Nhu Quynh et al., 2016).Therefore, it is considered that more suitable ratio of core to coating material was 1I:9 (v/v).
Based on the result of the optimized ratio of W/D to secondary coating material (1I:9) and emulsifier (Tween 60) as shown in Figure 2, optimized Tween 60 concentration was determined by ESO using various concentrations of Tween 60 (0.25, 0.50, 0.75, 1.00, and 1.25%) in the emulsion during keeping at 60°C for 2 h (data are not shown).As the Tween 60 concentration increased, the ESO was symmetrically increased.The ESO level reached to 95.1% at 1.00% (w/v) of Tween 60; however, the concentrations of 1.00 and 1.25% (w/v) of Tween 60 were not significantly different (p>0.05).Dn the other hand, it was observed that 0.25, 0.50, and 0.75% (w/v) of Tween 60 were creamed or rapidly segregated.Thus, 1.00% (w/v) of Tween 60 was selected for the emulsifier concentration in W/D/W microemulsions of tomato extracts in the present study.
To choose the optimum coating material in the concentration of 30% (w/v), ESO of the W/D/W microemulsions coated with various coating materials such as MD, WPC and AG was measured during incubation at 60 °C for 2 h (data are not shown).Among the coating materials used, MD exhibited the highest ESO value (95.3%) and significantly differ from others (p<0.05).According to Raja et al. (1989), MD with dextrose equivalence (DE) between 10 and 20 was suitable as coating material.When oil droplets were encapsulated by complex coacervation of WPO or AG, it was noticed that the emulsion was highly unstable against creaming.Therefore, it was indicated in this study that MD was the optimal coating material for the manufacturing of W/D/W microemulsion of tomato extracts.
Finally, ESO was measured to optimize the concentration of MD as the coating material for W/D/W microemulsions of tomato extracts during incubation at 60°C for 2 h (data not shown).The concentrations of 25 and 30% (w/v) MD showed higher ESO values than those of 10 to 20% (w/v), and even though the ESO values were not significantly different from 25 and 30% (p>0.05),30% MD was higher than 25% MD in the ESO value.Yoshii et al. (2001) have showed that MD concentration played important role for spay-drying of ethyl butyrate microcapsules.Klinkesorn et al. (2004) have also reported that as the MD concentration increased further to 25 or 35%, the creaming rate decreased remarkably.Therefore, the highest concentration (30%, w/v) of MD was decided as the optimum concentration in coating material.

Optimization of conditions for producing microemulsion of tomato extract by RSM
The yield (Y) of microemulsification was optimized using statistical experimental design based on coded level from three independent variables, such as ratio of tomato extract to MCT (X 1 ), concentration of emulsifier (Tween 60, X 2 ), ratio of W/D to secondary aqueous coating materials (X 3 ).The mathematical relationship in the form of regression equation was given below in terms of coded factors Equation 4I: R 2 (R-squared) describes the variation in the observed responses that is explained by the model (Minitab, 2010).R 2 of the yield was described by the model as 91.93% (data are not shown).Ot means that this model can be explained the responses approximately 91.93% accuracy The 3-D response surfaces for yield of microencapsulation were shown in Figure 3A, B and C. On the ratio of W/D phase, tomato extract as a core material, and concentration of Tween 60 were increased, the microencapsulation yield was increased.According to Gharibzahedi et al. (2012), RSM is a collection of useful mathematical and statistical procedures, and it needed to scrutinize multiple factors and their interactions in comparison to other approaches for possessing the advantage of reducing number of experimental trials.Min et al. (2018) optimized the nanoencapsulation efficiency with different emulsifiers using RSM in red ginseng extract nanoencapsulation.On the study, the increase of emulsifiers led to nanoencapsulation increase.On addition, the report of Lee et al. (2013) accorded with this study.The researchers reported that as the secondary emulsifier increase, the yield of microencapsulation increased in the microencapsulation of peanut sprout extract.
As apparently indicated in Figure 4, the optimum combination of factors for achieving the desired response was revealed to be the ratio of tomato to MCT (1.24I:8.76,v/v), concentration of Tween 60 (1.0%, w/v) and ratio of W/D to secondary aqueous coating materials (1.00I:6.16,v/v).Combining those factors makes sure the maximum yield (99.6911%) of microencapsulation.
On this study, spray-drying process variables, such as drying temperature, feed flow rate, atomizing pressure, and so on were not considered yet since lycopene is known to be more stable and have heat resistance compared to ascorbic acid in food processing (Nguyen & Schwartz, 1998;Tola & Ramaswamy, 2015) and it is considered that optimization of emulsification is more important than that of spray-drying condition.

Morphological observation
The morphology of microemulsions of tomato extracts at the optimum microencapsulation condition was analyzed with light microscopic after suspension with DW, as shown in Figure 5A.They were appeared spherical and well individualized.
After spray-drying to produce microcapsules of tomato extract, the morphology of the microcapsules was observed using SEM and presented in Figure 5B.These images revealed that spray-dried microcapsules aggregated.They showed spherical and smooth surfaces.Ersus & Yurdagel (2007) reported that morphology of the microcapsules coated with MD showed smooth surfaces.According to Fang & Bhandari (2010), the typical form of encapsulated particles produced by spray drying was globular.

Particle size distribution
Size distribution profiles of spray-dried microcapsules of tomato extract were measured with a particle size analyzer (Figure 6).Size distributions of microcapsules coated with MD at 30% concentration ranged from 1 to 10 μm, with an average particle size (D (4,3) ) of about 5 µm.Fang & Bhandari (2010) have reported that the typical form of encapsulated particles by spray drying is globular with size of 10-100 µm.Baranauskienė et al. (2006) have demonstrated that the particle size of microcapsules coated with 30% WPC varies from 2 to 556 μm.According to Choi et al. (2006) and Ahn et al. (2010), the microcapsules, which have 50-200 μm average size, were enough small to offer soft texture and superb dispersibility in milk.On addition, milk fat globules in homogenized milks have approximately 2-5 μm size, and milk fat globules are stable in milk serum.On view of these facts, the microcapsules which have 5 μm-average size are uniformly distribute in milk or other beverage systems.

Zeta-potential
Zeta-potential is recognized as an important parameter for the stability of colloidal system (Lin & Chaudhury, 2008).Results of zeta-potential for microemulsions and spray-dried microcapsules were compared (Figure 7).Average zeta-potential  values of microemulsions and the spray-dried microcapsules were approximately -40 mV and -29.68 mV, respectively.The border line between stable and unstable suspensions is generally taken at ±30 mV.Particles with an absolute value of zeta-potential of more than 30 mV are normally considered as stable (Van Nieuwenhuyzen & Szuhaj, 1999).Dn the other hand, as the absolute values of zeta-potential approached to zero or low-zeta potential value, the dispersion of the particles tended to be electrically unstable and rapidly coagulate or flocculate (Hanaor et al., 2012).Accordingly, both microemulsions and spray-dried microcapsule of tomato extracts can make stable suspension.

Conclusion
The microencapsulation technique is known to be good for protect vulnerable core material from light, oxygen, enzymes or other various environment.On the present study, the microemulsification of lycopene was conducted successfully with ratio of tomato to MCT at 1.24I:8.76(v/v), concentration of Tween 60 at 1.0% (w/v), and ratio of W/D to secondary aqueous coating materials at 1I:6.16 (v/v), and spray-dried the microemulsion to produce microcapsules.However, the advantages of microencapsulation could not confirm yet in this study.Therefore, it is necessary to perform further study on its release and storage properties and application into various food systems.

Figure 3 .
Figure 3. Response surface plots for the effects of variables on the yield (%) of microencapsulation of tomato extract.(A) ratio of tomato extract to medium-chain triglyceride (MCT) and concentration of Tween 60; (B) ratio of tomato extract to medium-chain triglyceride (MCT) and ratio of W/D to secondary coating material; (C) concentration of Tween 60 to ratio of W/D to secondary coating material.

Figure 5 .
Figure 5. Microphotographs of W/D/W emulsion for tomato extracts coated with 30% maltodextrin (A) and scanning electron microphotographs of spray-dried microcapsules of tomato extracts coated with 30% maltodextrin (B).

Figure 6 .
Figure 6.Particle size distribution of spray-dried microcapsules of tomato extracts coated with 30% maltodextrin.

Figure 7 .
Figure 7. Zeta-potential values of microemulsions and spray-dried microcapsule powders of tomato extracts.

Table 1 .
Coded level for independent variables used in experiment design for W/D/W microencapsulation of tomato extract.