Preparation of Piper nigrum Microcapsules by Spray Drying and Optimization with Response Surface Methodology

,

compounds between the Piper nigrum products and the Piper nigrum raw material, and the extracted components are often insufficiency. Meanwhile, the extracts are easily oxidized and deteriorated under conditions of high light, temperature and the like 6 .
The microcapsules consist of one or several different types ingredients as core material, and are coated with another materials. The coating materials is called shell material, wall material or encapsulant. Maltodextrin MD is a kind of carbohydrates that can wrap the core material to form a cover, encapsulate flavor and reduce exposure to oxygen. It is cheap, highly water soluble, bland in flavor and low viscosity at a high solid ratio 7 . Gum, such as xanthan gum XG , is also used for encapsulation due to its good stabilization of emulsion over a wide pH range 8 . Moreover, it is compatible with most starches, carbohydrates, and proteins. Some studies have demonstrated that the use of mixtures of MD and gums appeared to offer a good balance between cost and efficiency of microencapsulation using a spray dryer 9,10 . Starch is a kind of natural macromolecule with hydrophilic properties, which is widely used in the food industry to retain and protect volatile compounds. Starch sodium octenyl succinate SOSS , as a new kind of modified starch, has amphiphilic molecular structure featured with hydrophobic alkenyl groups, which is widely used in food, pharmaceutical and industrial fields 11 .
Microencapsulation by spray drying is the most common and economically viable technology to protect liquids, solids or even gases against the surrounding environment into microscopic particles 12 . To overcome the problems of incomplete flavor components from Piper nigrum extract and the Piper nigrum product easy deterioration in the storage process, enzymatic technology 13,14 and microencapsulation technology 15 can be thought as an advisable alternative to retain the original flavor and stabilize the flavor compounds. Enzymatic technology is used to highly release the active components from Piper nigrum, and microencapsulation is performed to stabilize the flavor compounds of the Piper nigrum products 16 .
Therefore, in order to maintain and stabilize the natural flavor of Piper nigrum, microcapsules of the whole Piper nigrum were prepared by spray-drying combined with enzymatic hydrolysis, in the meantime, the optimization of the microencapsulation process was performed by response surface methodology RSM . Besides, the products were evaluated on volatile aroma components, particle size and surface morphology, etc.

Materials and reagents
Piper nigrum was obtained from Yunnan, China. Starch sodium octenyl succinate SSOS, 99 of food-grade was purchased from Guangzhou Huahui Biological Industry Co., Ltd. Xanthan gum XG, 99 and maltodextrin MD, 99 were food-grade and supplied by Shanghai Yuanpu Biological Technology Co., Ltd. Petroleum ether ≥ 90 , ethanol 99. 8 and methanol 99.8 were analytical grade and purchased from Sigma Chemical Company. Cellulase 50,000 U/g and neutral protease 100,000 U/g were obtained from Henan Yatong Food Raw Material Co., Ltd. Sodium stearoyl lactylate SSL 99.8 was food-grade and obtained from Jinan Lanxin Biological Technology Co., Ltd. Mixed standards of n-alkanes C 5 -C 25 were purchased from sigma Aldrich Beijing, China . Distilled water was used for the preparation of all solutions.

Preparation of core material containing piperine
Our previous work has shown that enzymatic hydrolysis can increase the extraction efficiency of piperine from Piper nigrum 17 . Therefore, enzymatic hydrolysate containing Piper nigrum flavor compounds was as the core material. It was prepared via the following routes. Piper nigrum sample was first ground into a fine powder by using a laboratory grinder mill, and then the powder was sieved less than 10 mesh. Then, a mixture of 10 g of Piper nigrum farina, 0.008 g of cellulose, 0.01 g of neutral protease, 0.04 g of SSL and 50 mL water was stirred thoroughly at 60 for about 4 h. The mixture was then quickly heated to 90 and kept for 5 min in order to inactivate the enzyme, and then quickly cooled to room temperature. After that, the enzymatic hydrolysate was obtained and collected for the next step.

Preparation of microcapsules by spray-drying
SSOS, MD and XG powders, as wall materials the ratio of the three materials was 1:1:0.2 w/w/w , were dispersed in distilled water wall material concentration was 11 w/ v with vigorous agitation, which was continued until the wall materials completely dissolved.
The enzymatic hydrolysate of Piper nigrum was then added into the solution of the wall materials the ratio of core and wall material was set as 1:0.2 w/w with agitation by means of a high-speed blender UltraTurrax T-25, IKA Instruments, Germany .
Microencapsulation was performed by spray drying. The emulsion was fed to a mini spray dryer B290 Büchi, Flawil, Switzerland at the inlet air temperature of 180 . The nozzle cleaner was set to level 4 and the aspiration rate was set to 100 . The feed rate of the peristaltic pump was set to 6.67 mL/min, and the speed of the suction pump was set to 100 . The powders, which got from above conditions, were collected and stored at 4 in a closed container without light 16,18 .

Determination of microencapsulation efficiency
The extraction method of piperine from initial Piper nigrum powder was that 2 g of Piper nigrum powder was dissolved in 20 mL ethanol into a container with ultrasound for about 30 mins 19 . While the measurement of piperine in the microcapsule requires that 2 g of the prepared microcapsules was firstly mixed with moderate amount of petroleum ether, and then was shaken gently by hands for 2 mins at room temperature. After that, the mixture was filtered and the retained sample was rinsed three times with 20 mL petroleum ether. The collected sample was ground, and then 2 g of sample were dispersed in 20 mL of ethanol with ultrasound for 30 mins.
The content of piperine was detected at the wavelength of 343 nm using UV-vis spectrophotometer Shimadzu UV-3600 with Ultraviolet Spectrophotometry 20 . The external standard method was used for quantitative analysis. The standard curve was y 0.2645x 0.0042, R 2 0.9996 n 13 .
The encapsulation efficiency was calculated by the following equation: where Y P was encapsulation efficiency, M P mg/mL was the content of piperine in produced Piper nigrum microcapsules, P P mg/mL was the content of piperine in Piper nigrum powder added in the process of the preparation of microcapsules.

Experimental design
The main factors affecting the microencapsulation efficiency were the ratio of the core and wall material, proportion of wall materials, wall material concentration and inlet air temperature. In order to optimize the microencapsulation process conditions, a single-factor experiment was first performed. Based on the results of the single factor experiment, the ratio of the core and wall material was set of 1:0.2 w/w .
Then, the analysis experiment using Box-Behnken s response surface method was performed, in which the proportion of wall materials x 1 , wall material concentration x 2 and inlet air temperature x 3 were used as the independent variables, and the microencapsulation efficiency y was used as the response value. In order to facilitate multiple regression analysis, the actual variables were coded in Table 1.
In Table 2, experiments 1 12 were factorial experiments, and 13 17 were central experiments. The 17 experimental points were divided into factor points and zero points, where the factor point were three-dimensional vertices composed of independent variables with values of x 1 , x 2 , and x 3 , and the zero point was the center point of the region. The zero-point experiment was repeated five times to estimate the experimental error.
In the regression analysis, the relationship among the independent variables x and dependent variables y was valued by the following second order polynomial equation: where β 0 was constant coefficients of intercept, β 1 , β 2 and β 3 were constant coefficients of linear, β 11 , β 22 and β 33 were constant coefficients of quadratic, β 13 , β 12 and β 23 were constant coefficients of interaction terms.

Fourier transform infrared spectroscopy FTIR analysis
In order to avoid disturbances of other ingredients in the Piper nigrum enzymatic hydrolyzate, Piper nigrum essential oil extracted by steam distillation was used as the infrared detection sample. Pure wall material, Piper nigrum essential oil extracted from Piper nigrum, and microcapsules samples were diluted with potassium bromide powder at a mass ratio of 1:100 respectively. Then they were ground and pressed into a transparent wafer using the specific presser. Finally, the sample wafer was put into the sample cell. The infrared spectra were collected in the range of 400-4000 cm 1 at ambient temperature using a Nicolet iS10 FT-IR spectrometer Thermo Fisher, USA .

Identification of chemical compounds from
Piper nigrum powder and microcapsules by solid-phase microextraction SPME and gas chromatography GCmass spectrometry MS A certain amount of Piper nigrum powder and microcapsules sample approximately 5.0 g were placed in a 15 mL glass vial and sealed with a Teflon septum cap, respectively. Each sample was heated at 60 for 30 mins. After that, the SPME fiber coated with carboxen/polydimethylsiloxane CAR/PDMS, 65 μm, Bellefonte, PA, USA was conditioned in the GC-MS injection port for 30 min at 240 to remove any contaminants, and then was exposed to the vial headspace and inserted into the injection port of the GC system 21 .
The volatile constituents were analyzed with a mass detector model 7000B Agilent Technologies, Palo Alto, CA, USA coupled to a 7890A GC, equipped with a HP-5MS capillary column 30 m 250 μm 0.25 μm from Agilent. The injector temperature was 240 with a splitless injector. The oven temperature program was held at 40 for 5 mins, increased by 5 /min to 240 and was maintained for 4 mins. The carrier gas was helium at a constant flow of 1 mL/min. Mass spectra were recorded at 70 eV in the EI mode. The temperature of the quadrupole mass detector, ion source and transfer line were maintained at 150 , 230 and 250 , respectively. The scanning quality range was m/z 40 400. Compound identification was based on mass spectra matching with the standard NIST 2014 MS library and also on the comparison of retention indexes RI sourced from the NIST Standard Reference Database and references 22 . The calculation of RI was according to references 23 25 . In addition, the relative contents of individual components were calculated using their GC peak areas 22 . All analyses of the volatile compounds were made in triplicate.

Morphology
The morphology of the microcapsules was observed by a scanning electron microscope SEM, Sigma 300, Zeizz under high vacuum conditions at a voltage of 10 kV. The microcapsules were dissolved in the petroleum ether and the mixture was dropped on silicon sheet 10 mm 10 mm . The sample was measured until the solution evaporated completely at room temperature.

Particle size
Particle size of the microcapsules was determined using a particle size analyzer Nano Brook 90PlusZeta; Brookhav-en Instruments Ltd., USA . The suspension sample was prepared by mixing the emulsions with distilled water at 25 .

Results and Discussion
3.1 Microencapsulation efficiency and its condition optimization The experimental results reported in our previous work indicated that surfactant-assisted enzymatic hydrolysis could promote the effective release of piperine in Piper nigrum. For the composition of the emulsion, the content of the extracted piperine markedly increased from 0.14 to 4.42 HPLC , from 0.16 to 4.54 UV spectrophotometric , respectively 17 . With the enhancement of the enzymatic hydrolysis, the content of piperine in the core material enzymolysis solution was greatly increased, so that the active ingredients of Piper nigrum were distributed in a large amount in the emulsion and the prepared microcapsules contain more piperine, which greatly preserved the original ingredients and the biological activities of Piper nigrum. Therefore, enzymatic hydrolysis had a certain effect on the improvement of microencapsulation efficiency.
The polynomial regression model equation, which were calculated for the microencapsulation efficiency, were fitted as follow: y 87.  Table 2. Additionally, the results of variance analysis were given by Design-Expert software, which were shown in Table 3. As can be seen from Table 3, the equation regression was significant R 2 0.9792; R 2 Adj 0.9524 . And furthermore, the results of RSM showed that the proportion of wall materials x 1 and wall material concentration x 2 had a significant effect on the efficiency of microencapsulation p 0.05 . Among the three wall materials, SOSS was the emulsifier, MD was film foaming, and XG had both functions 16 . When the content of MD was high, the components in the system could not be fully emulsified, the emulsion was not stable enough, and the microcapsule film formed on the surface was not dense enough. However, when the content of MD was low, the active ingredients in the system cannot be embedded, resulting in poor film-forming ability and low microencapsulation efficiency of microcapsules. Therefore, the proportion of functional wall material was one of the key factors for the success of microencapsulation 26 . Secondly, the wall material concentration was also one of the factors affecting the efficiency of microencapsulation. When the wall material solid content was high, the film formation speed of the microcapsule coating material and the density of the capsule wall were improved 27 . It would enhance the encapsulation of the core material, reduced the diffusion and migration of the core material to the wall material, and thus improved the embedding rate of the microcapsule. However, when the wall material was too high, the emulsion viscosity

Preparation of Piper nigrum Microcapsules
would also increase. The obtained emulsion was easy to stick and block around the nozzle, which leaded to a long time for the flavor substances in Piper nigrum to be heated. To some extent, it would cause the loss of active ingredients and affect the evaporation of water and the atomization of liquids. Finally, the embedding rate of the microcapsules was reduced 28 . Moreover, the results of RSM also showed that among quadratic and interaction terms of independent variables, x 1 x 3 , x 2 x 3 , x 2 2 , x 3 2 were significant p 0.05 . It indicated that the experimental factor was not a simple linear relationship to the response value. Thence the experimental parameters were further optimized by the Design-Expert software, and the most favorable conditions were the proportion of wall materials of 1:1:0.2 SSOS: MD: XG, w/w/w , wall material concentration of 11 w/v and inlet air temperature of 180 . Under the above conditions, the predicted value of microencapsulation efficiency could reach to 90.62 . Finally, the average microencapsulation efficiency of confirmatory experiment carried out 3 times reached 90.21 , which was only 0.41 difference from the theoretical value. It showed that the experimental value was extremely close to the predicted value and there was no significant difference between them. Meanwhile, it revealed that the RSM was feasible to optimize the microencapsulation conditions, and the obtained microencapsulation conditions had certain practical application value.

FTIR analysis
The microcapsules of Piper nigrum were analyzed by FTIR spectroscopy in order to verify the presence of the piperine in the microcapsules. The FTIR spectrum of pure wall material, which was shown in Fig. 1a, consisted of prominent absorption peaks at 3420 cm 1 for the hydroxyl -OH groups 29 , 1448.56 cm 1 for -CH 2 bending vibration, 1371.36 cm 1 for -CH 3 bending vibration 30 and 1060.86 cm 1 for C-O stretching vibration 29 .
On the other hand, the FTIR spectrum of Piper nigrum essential oil Fig. 1c , showed that the weak absorption band observed at a wavelength of 3648.71 cm 1 was characteristic for hydroxyl -OH groups 31 . The peaks presented with strong intensity at approximately 2927.30 cm 1 and 2858.41 cm 1 belonged to C-H stretching vibration of the methylene groups 32 . The signal at 1635.67 cm 1 corresponded to the C C stretching vibration. Furthermore,  the strong absorption band at the 1367.45 cm 1 and 1456.76 cm 1 was associated to C-H bending vibration 33 . As shown in Fig. 1b, the spectra of the Piper nigrum microcapsules also showed peaks characteristic of Piper nigrum essential oil, and there was no new chemical bond between the microcapsules and the wall materials spectrums, thus confirming that the Piper nigrum essential oil was contained in the wall material and there was no chemical reaction between the core materials and SSOS-MD-XG mixtures. Similar results have been reported by other studies 32, 34 .

Analysis of volatile aroma components
By using GC-MS analysis, twenty-one and fourteen odoractive volatile compounds were detected in Piper nigrum powder and Piper nigrum microcapsules, respectively Fig. 2 . The odor-active volatile compounds identified from Piper nigrum powder and Piper nigrum microcapsules were presented in Table 4. As it can be seen, the quantities of alkenes in Piper nigrum powder and microcapsules were both high. More specifically, the most important volatile compounds for aroma in these two samples were carene, D-limonene, α -phellandrene, and -β -pinene. Furthermore, the highest content of the two samples was D-limonene. Indeed, similar results were obtained in previous studies focusing on Piper nigrum 21,30 .
The recovery percentages of most volatile compounds were relatively low no more than 15 while some main volatile compounds, such as D-limonene and carene, were higher recovery percentages in microcapsules. Meanwhile some volatile components, such as β-caryophyllene and / --δ -elemene, were not detected in the microcapsules. Previous studies reported that the volatile retention quantity of carbohydrate had no relation to the relative content of aroma compounds, but with the volatile molecular weight increasing, it would increase correspondingly. In addition, it was established that some volatile compounds were lost during microencapsulation 35,36 .
However, most of main components from Piper nigrum powder had a little change before and after microencapsulation, which proved that most of the volatile components could be maintained by microencapsulation and the natural flavor of Piper nigrum could be retained to the greatest extent. Figure 3 revealed that the microcapsules had a broad size distribution with an irregular shape Fig. 3A . In general, microcapsules that produced by spray drying generally had a spherical or irregular shape and various sizes 29,30,37 . Those microcapsules produced with SSOS, MD and XG powders as wall materials were characterized by larger deformations and irregular surfaces folds Figs. 3B-3D . These irregular surfaces were mainly represented by many shrinkages and dents on the surface of the microcapsules, which were consistent with some previous reports 38, 39 . The shape irregularity of the microcapsules was closely related to the components of core material including flavor compounds, essential oil and other cellulose and lignin, etc. Besides, the presence of depressions on the surface of the microparticles has been associated to the shrinkage at the early stages of the drying and subsequent cooling 40,41 .

Particle size determination
The particle size distribution curve of Piper nigrum microcapsules emulsions was presented in Fig. 4. As it can be seen in Fig. 4, the particle size distribution of emulsions was uniform. More specifically, the average particle size distributions of the emulsions were 0.15 μm, which were slightly lower than those previous studies on different microcapsules 28,38 . The enormous differences between studies mainly arose from the emulsification methods 28 and the properties of wall and core materials. In addition, recent studies have shown that the average particle size distributions of the microcapsules carried out by the ultrasonic emulsification method were generally smaller 42 , which may be attributed to cavitation and mechanical function of ultrasound 43,44 .

Conclusions
In conclusion, the produced Piper nigrum enzymatic hydrolyzate, the wall material used and the spray drying conditions were effective in keeping the natural chemical compounds in the Piper nigrum microcapsules. The best conditions for the microencapsulation have been determined as the ratio of core and wall material 1:0.2, w/w , proportion of wall materials 1:1:0.2, SSOS : MD : XG, w/w/ w , wall material concentration 11 , w/v and inlet air temperature 180 . Under the optimal conditions determined by RSM, the microencapsulation efficiency of Piper nigrum could achieve 90.21 . In the microscopic observation, the shape of the microcapsules was irregular and the average particle size of the emulsion was 0.15 μm.    Based on the findings, this study could provide a useful guideline for the microencapsulation of Piper nigrum.