Microencapsulation of Hibiscus sabdariffa (Roselle) Extracts by Spray Drying Using Maltodextrin and Gum Arabic as Carriers

Microencapsulation by spray drying is one of the most common methods used to obtain food material powders. In this study, different gums (maltodextrin [MD], gum arabic [GA], and mixtures of MD:GA [60:40] at various concentrations [0–10% w/w]) were used to microencapsulate Hibiscus sabdariffa (Roselle) extracts by spray drying. The yield, physicochemical properties, and antioxidant characteristics (total monomeric anthocyanins [TMAs], total phenolic compounds [TPCs], and antioxidant capacity [AC]) of the microencapsulated Roselle powders (RP) were evaluated. The highest RP yield (73.3 ± 3.3%) was obtained with the 3% MD:GA blend. The red color (a*) average for all powders (39.9 ± 2.0) decreased as the gum concentration increased. The 3% MD:GA RP showed the highest amount of TMAs (539.19 ± 13.27 mg cyaniding-3-glucoside equivalents/100 g) and TPCs (3,801.6 ± 125.9 mg of gallic acid equivalents/100 g of powder). The highest AC was observed with a 5% GA RP (1498.5 ± 44.0 mg of Trolox equivalents/100 g of powder).


Introduction
Hibiscus sabdariffa is a plant in the Malvaceae family. It grows in tropical and subtropical regions and can have green or red calyces (Cissé et al., 2009). The red color of calyces reflects the high anthocyanin, mainly delphinidin-3-sambubioside (71.4%) and cyanidin-3-sambubioside (26.6%) content (Peng-Kong et al., 2002). These compounds are highly unstable and degrade easily, producing compounds with an undesirable color (browning). One of the main attributes of food quality is color and consumer acceptance depends greatly on color. Anthocyanins are colorful pigments from vegetable products; the stability of anthocyanins depends on various factors such as temperature, pH, oxygen, light, enzymes, and metallic ions (Idham et al., 2012;Ersus et al., 2007). aluminum foil, and stored a 4°C until spray drying. RE and RE-gum concentrates were analyzed for physicochemical (TSSs, density, viscosity) and antioxidant (TMAs, TPCs, and AC) characteristics.

Spray Drying
A mini spray dryer (Büchi B-290, Switzerland) with a two-fluid nozzle with an orifice 0.7 mm in diameter (particle diameter of 1-25 microns) was used for spray drying. The inlet and outlet air temperatures were 180.01 ± 0.25 and 105.16 ± 3.52°C, respectively. Blends of RE-gum concentrates were fed into the dryer at a flow rate of 10 mL/min (Andrade and Flores, 2004;Ochoa-Velasco et al., 2017). The aspirator power of the drying system was 100% (equivalent to an airflow of 35 m 3 /h), and the spray drying airflow was maintained at 55 mm (equivalent to 670 L/h with a pressure of 1.05 bar). Calibration curves were constructed (10-100% pump power) to determine the percentage equivalent to 10 mL/min for each mixture (38-41%). RPs were weighed and placed in 100-mL amber glass bottles, and these bottles were capped and stored at room temperature (22 ± 2°C) in a desiccator containing silica. Yield, productivity, physical characteristics (moisture content, water activity [aw], average diameter, bulk and tapped densities, and color), and antioxidant characteristics (TMAs, TPCs, and AC) of RPs were determined.

Physical Properties of Extracts
Total soluble solids (TSSs). TSSs were measured according to the 932.14C AOAC (1995) method using a handheld refractometer (Atago Co. LTD, Tokyo, Japan). To correct values for 20°C, a standard set of tables were used (AOAC, 1995).
Density. Density was determined according to the 945.06 AOAC (1995) method and expressed in g per mL. Empty (W 1 ), filled with distilled water (W 2 ), and filled with sample (W 3 ) pycnometer weights were determined, and the density at 25°C was calculated according to Eq. (1): (1) where (g/mL) is the density of water at 25°C.
Absolute viscosity (µ). A Cannon-Fenske capillary viscometer (Cannon Instrument Co., State College, PA, USA) was used to determine absolute viscosity. Kinematic viscosity was calculated by multiplying the time (s) of 6.6 mL of extract at 40°C flowing through the viscometer per the constant of the apparatus (0.4754 mm 2 /s 2 ) at the same temperature. To obtain absolute viscosity, kinematic viscosity was multiplied by the density of the extract according to Eq. (2) (Cannon Instrument Co., 2000): (2) where µ is the absolute viscosity (cP = mPa . s),  s is the density (g/mL), and  c is the kinematic viscosity (mm 2 /s = cSt) of an extract. The absolute viscosity at 25°C was calculated using Eq. (3) (Cannon Instrument Co., 2014): ( where C (0.4754 mm 2 /s 2 ) is the constant of the apparatus at 40°C, C o (mm 2 /s 2 ) is the viscometer constant at the filling temperature, B (79×10 −6 /°C) is the calibration temperature factor obtained from the calibration certificate for the viscometer, T t is the working temperature (40°C), and T f is the filling temperature. Using the equation above, the constant C o was calculated and then, using the same equation, the constant C was calculated at 25°C.

Antioxidant Properties
Total monomeric anthocyanins (TMAs). TMAs were determined according to the method described by Lee et al. (2005). First, 0.5 mL or 100 mg of extract or powder, respectively, were diluted with distilled water to reach 10 mL in a volumetric flask. The mixture was stirred for 5 min using a vortex at 2900-3000 rpm. One milliliter of each solution was diluted with buffer pH 1.0 or pH 4.5 to reach 5 mL in test tubes wrapped with aluminum foil. The blends were left for 30 min at room temperature (23 ± 2º C) in the dark. Then, absorbances in 4-mL glass cells were measured at 520 and 700 nm using a Cary 100 UV-Vis spectrophotometer (Varian Inc., Palo Alto, CA, USA). A blank with distilled water was used to correct these absorbances. Results were calculated as mg of cyanidin-3-glucoside equivalents per 100 mL of RE or per 100 g of powder using Eq. (4): where TMA is the concentration of anthocyanins (mg/100 mL or mg/100 g), A = (A 520nm -A 700nm ) pH=1.0 -(A 520nm -A 700nm ) pH=4.5 , MW is the molecular weight of cyanidin-3-glucoside (449.2 g/mole), DF is the dilution factor, L is the cell width (1 cm),  is the coefficient of molar extinction for cyanidin-3-glucoside (26,900 L/mole-cm), and 100 is the conversion factor for obtaining mg/100 mL of RE or mg/100 g of RP.
Total phenolic compounds (TPCs). TPCs were determined using the Phenol Folin and Ciocalteu method (Singleton and Rossi, 1965) with some modifications. Three milliliters of distilled water, 150 L of extract solution, or 100 L of powder solution (the same solutions prepared to determine TMAs), and 250 L of Folin and Ciocalteu reagent were placed in test tubes that were then covered with aluminum foil. Mixtures were stirred and left for a maximum of 8 min in the dark, and then 750 L of 20% Na 2 CO 3 was added and thoroughly mixed. Distilled water was added (850 or 900 L) to reach 5 mL, mixed thoroughly, and left for 2 h at room temperature (23 ± 2°C) in the dark. Absorbances were measured at 765 nm using a Cary 100 UV-visible spectrophotometer (Varian Inc. where TPC is the total phenolic compounds content (mg/100 mL or mg/100 g), A is the absorbance of the sample, b is the intercept, m is the slope, and DF is the dilution factor for the sample.
( 100 ) = ( − ) * * 100 where AC is the antioxidant capacity (mg/100 mL or mg/100 g), A is the absorbance of the sample, b is the intercept, m is the slope, and DF is the dilution factor of the sample.

Physicochemical Properties of Powders
Yield (Y). Yield was calculated based on the amount of TSSs in the encapsulated extract and the amount of powder obtained (Fazaeli et al., 2012)  Water activity (a w ). Water activity was measured using an AQUA-LAB hygrometer model 3TE (Decagon Devices Inc., Pullman, WA, USA). The temperature at the time of measurement was 25.10 ± 0.06°C.
Bulk density. Bulk density was measured according to the method described by Jumah et al. (2000). One gram of powder was weighed in a 10-mL graduated cylinder. The cylinder was gently tapped 10 times on a polystyrene mat from a height of 15 cm. Bulk density ( a ) was calculated according to Eq. (9): where W is the weight of powder (g) and Va is the apparent volume (mL) occupied by the powder in the cylinder after tapping.
Tapped density. Tapped density was measured according to the Mexican Official Norm number NOM-104- STPS-2001(NOM, 2001 with some modifications. One gram of powder was weighed in a 10-mL graduated cylinder with a rubber stopper. The cylinder was subjected to a manual vibration process so that the sample were shaken from bottom to top for 8 min (estimated time at maximum volume). The tapped density ( c ) was calculated according to Eq. (10): where W is the weight of powder (g) and Vc is the compacted volume (mL) occupied by the powder in the cylinder after tapping.
Color. A Colorgard system 05 colorimeter (BYK-Gardner Inc., Silver Spring, MD, USA) was used to determine the color of powders and solutions. For powders, the color parameters were obtained in reflectance mode. A plate with a light gap 1.9 cm in diameter and external diameter of 2.65 cm was used. Samples were placed in weighing bottles for color determination. For solutions, a solution of 10 mg of powder/mL of distilled water was prepared, and color parameters were determined in transmittance mode using a 3-mL quartz cell (Konica Minolta Sensing, Inc., Kyoto, Japan) (Ochoa-Velasco et al., 2017; Silva et al., 2013). Color parameters of powders and solutions were obtained using the CIEL*a*b* scale: L* (lightness, 0-100), a* (green to red) and b* (blue to yellow). From these data, purity (color saturation, C = [a* 2 + b* 2 ] 1/2 ) and hue (H = tan -1 [b*/a*]) were calculated.

Statistical Analysis
Data were subjected to analysis of variance (ANOVA) testing using MINITAB software version 14.1. Multivariate analysis and Tukey's multiple comparison tests were used to compare differences between means. Values shown are average values. A value of 0.05 was considered significant for differences between means of treatments.

Physical Properties
TSSs (°Bx). Significant differences (p < 0.05) were observed between average TSS contents (Table 3) for REs with GA, MD, or MD:GA added. The TSS content in REs was 15.04 ± 0.80. An increase of gums in REs increased TSS content (Table 3). Density. Densities of RE-gum concentrates (Table 3) showed significant differences (p < 0.05) among types and concentrations of gum. Extracts with the MD:GA blend added were denser (1.06 ± 0.01 g/mL) than extracts without gum (RE) or with GA and MD (1.03 ± 0.01, 1.04 ± 0.02, and 1.03 ± 0.01 g/mL, respectively). Densities of RE-gum concentrates with the three concentrations of gums also showed significant differences (p < 0.05).
Viscosity. Viscosities of RE-gum concentrates are shown in Table 3. The viscosity of RE-gum concentrates increased as gum concentration increased (1.66 ± 0.01, 2.26 ± 0.40, 277 ± 0.69, and 5.27 ± 2.45 mPa . s, for 0, 3, 5, and 10% gums, respectively). Extracts with 10% GA showed the highest viscosity, perhaps because GA has the ability to form gels due to its protein contents (Lopez et al., 2009). Significant differences (p < 0.05) in RE-gum concentrate viscosities were observed among concentrations and types of gum. Viscosity and TSS content are important for spray drying because low viscosities along with high TSS content results in better flow during atomization and higher yields (Lopez et al., 2009). Therefore, a positive correlation between viscosity and TSS content was observed with each treatment: GA (R 2 = 0.938), MD (R 2 = 0.988), and MD:GA (R 2 = 0.980).

Physicochemical Characteristics
Moisture content. In general, RPs with the highest (p > 0.05) moisture contents were the following: 3.34 ± 0.30 and 3.29 ± 0.27% for 3 and 5% MD:GA, respectively; 3.29 ± 0.27 for RE; and 3.09 ± 0.24 and 2.29 ± 0.15% for 10 and 3% GA, respectively (Table 4). No significant differences (p > 0.05) were observed in moisture contents with gum concentrations of 0, 3, 5, and 10% (3.29 ± 0.27, 2.68 ± 0.57, 2.48 ± 0.68, and 2.60 ± 0.41%, respectively). Gonzales-Palomares et al. (2009) reported a moisture content of 4% in spray-dried powders using inlet and outlet temperatures of 180 and 80°C, respectively, for control REs, which is similar to values obtained in this study (3.29 ± 0.27%). Likewise, Ochoa-Velasco et al. (2017) reported an average moisture content in spray-dried microencapsulated powders from REs using mesquite gum as an encapsulating agent at different concentrations (1, 2, 3, 4, and 5% w/v) of 2.29 ± 0.45%. This value is similar to the average moisture content (2.59 ± 0.55%) in RPs with different gums obtained in this study. Water activity (a w ). The stability of many foods depends on water activity (Fennema, 1985). High a w indicates high free water content and thus low food stability. Table 4 shows low a ws for all RPs. Average a ws for RPs with MD, GA, and MD:GA were 0.163 ± 0.014, 0.208 ± 0.036, and 0.223 ± 0.022, respectively. Significant differences (p < 0.05) were observed among types and concentrations of gums. The highest values were observed with 10% GA.
Tapped density. Table 4 shows the tapped densities of RPs. Tapped densities showed trends that were similar to those of bulk densities; however, because the RPs were tapped, all densities were higher.

Color of Powders
Lightness (L*). Significant differences were observed between RPs based on type and concentration of gums (Table 5). RE RPs (41.15 ± 1.00) were darker than those with GA (55.76 ± 1.90), MD (57.14 ± 3.97), and MD:GA (52.42 ± 5.22). Lightness of RPs increased with increasing gum concentration, with lightness measures of 41.15 ± 1.00, 51.51 ± 3.18, 54.36 ± 3.03, and 59.45 ± 2.07 for 0, 3, 5, and 10% gums, respectively. Ochoa-Velasco et al. (2017) reported an average lightness value of 40.3 ± 0.71 for microencapsulated RPs obtained by spray drying using mesquite gum at different concentrations (1, 2, 3, 4, and 5% w/v); however, no significant differences were observed with the different microencapsulated powders. The authors concluded that gum concentration did not have a significant effect on all color properties. Idham et al. (2012) reported color parameters of RPs with the same gums used in this study, but they purified anthocyanins before they were mixed with the carrier for spray drying. They obtained L*, a*, and b* values of 44.9, 30.3, and − 6.3 for RPs with GA; 39.3, 43.1, and −0.8 for RPs with MD; and 45.9, 34.8, and −4.3 for RPs with MD:GA. Gums were added to the extracts to reach a concentration of 20%. The mixtures were fed into the spray dryer at a flow rate of 9.5% with inlet and outlet temperatures of 150 and 110°C, respectively. Green-red color (a*). Green-red color values (a*) decreased significantly, as gum concertation increased, with averages of 39.72 ± 1.07, 38.94 ± 2.01, 39.94 ± 2.23 for RPs with GA, MD, and MD:GA, respectively. RE RPs had the highest red color value (42.68 ± 0.26). The average green-red color values by gum concentration were 42. 68 ± 0.26, 41.20 ± 0.55, 39.97 ± 0.94, and 37.43 ± 1.13 for 0, 3, 5, and 10% gums.
Purity (C). The purity (chroma) of all RPs showed trends that were similar to those of hue. Significant differences were observed based on type of gum added, with purities of RPs with GA, MD, and MD:GA of 40.91 ± 1.38, 40.32 ± 2.43, and 41.72 ± 2.90, respectively. RE RPs were purer (45.69 ± 0.17) than RPs with GA, MD, or MD:GA. The purity of powders were found to decrease as gum concentration increased, with purity values of 45.69 ± 0.17, 43.17 ± 0.99, 41.49 ± 1.18, and 38.29 ± 1.07 with gum concentrations of 0, 3, 5, and 10%, respectively. RE RPs showed the highest purity (p < 0.05). Purity values specify the position of colors between gray and a pure hue (saturation). Therefore, the purity or chroma of a color is proportional to the amount of color it has (McLaren, 1986). Figure 4 illustrates the correlation between a* and purity values for all treatments; an increase in the a* value was associated with an increase hue purity.  3.4.6 Antioxidant Characteristics Table 6 shows TMAs, TPCs, and AC for the different RPs.

Conclusions
Based on the drying conditions used in this study, the microencapsulated RPs obtained with a mixture of MD and GA (60:40) as a carrier were the preferred powders because of its higher yields and better antioxidant and color characteristics. However, the red color (a*) average for all powders decreased as the gum concentration increased which is due to the gum concentration. In addition, the 3% MD:GA RP showed the highest amount of TMAs (cyaniding-3-glucoside equivalents/100 g) and TPCs; however, TMAs and TPCs were well maintained in all MD:GA RPs. These results indicate that microencapsulated powders can be used successfully to produce attractive functional foods as well as imparting flavor characteristics to foods. However, a stability study should be conducted with these RPs to evaluate their carrier efficiency. A study of MD:GA mixtures at different ratios than those used in this work should also be conducted to optimize yields and physicochemical properties of RPs obtained. Therefore, more studies about stability of color, solubility, moisture sorption characteristics, and maintenance of antioxidant properties are required.