Separation of the main flavonoids and essential oil from seabuckthorn leaves by ultrasonic/microwave-assisted simultaneous distillation extraction

Volatile essential oils (EOs), non-volatile rutin (RU), quercetin (QU), kaempferol (KA) and isorhamnetin (IS) were effectively extracted and isolated from seabuckthorn (Hippophae rhamnoides L.) leaves by ionic liquid-based ultrasound/microwave-assisted simultaneous distillation extraction (ILUMASDE). After optimization by response surface methodology, EOs, RU, QU, KA and IS were separated under the following optimum conditions: an ionic liquid of 1.0 M 1-butyl-3-methyl imidazole bromine salt ([C4mim]), liquid/solid ratio of 12 ml g−1, extraction time of 34 min, microwave power of 540 W and a fixed ultrasonic power of 50 W. Under the optimized conditions of ILUMASDE, the extraction yields of RU, QU, KA, IS and EOs were 9.18 ± 0.35, 5.52 ± 0.23, 3.03 ± 0.11, 5.64 ± 0.24 mg g−1 and 0.095 ± 0.004%, respectively. The yield of EOs obtained using ILUMASDE was 1.07-fold higher than that obtained by conventional hydrodistillation extraction (HDE). In addition, the components of the EOs obtained using ILUMASDE and HDE were similar. The extraction yields of RU, QU, KA, IS obtained by ILUMASDE were 1.03–1.35-fold higher than that obtained by the ethanol ultrasonic-assisted extraction (EUAE), ionic liquid-based ultrasonic-assisted extraction (ILUAE) and ionic liquid-based microwave-assisted extraction (ILMAE). And the extraction time used by ILUMASDE was 34 min, which is 14.17%, 56.67%, 56.67% and 85.00% less than those used by HDE, EUAE, ILUAE and ILMAE, respectively. Therefore, ILUMASDE can be considered a rapid and efficient method for extracting flavonoids and EO from seabuckthorn (Hippophae rhamnoids L.) leaves.

Formic acid, methanol and acetonitrile of HPLC grade were purchased from J&K Chemical Ltd (Beijing, China). Deionized water was acquired by a Milli-Q system (Millipore, Billerica, MA, USA). All solutions prepared for HPLC analysis were filtered through a 0.45 µm nylon membrane (Guangfu Chemical Reagents Co.,Tianjin, China) and degassed by ultrasonication in advance. Seabuckthorn (Hippophae rhamnoides L.) leaves were obtained from Heilongjiang Academy of Agricultural Sciences (Heilongjiang, China). They were dried in the shade at room temperature, comminuted using a disintegrator (HX-200A, Yongkang Hardware and Medical Instrument Plant, Yongkang, China), passed through a stainless-steel sieve (80 mesh), and stored in closed desiccators at room temperature before use.

ILUMASDE
In the ILUMASDE device (figure 2) the Likens-Nickerson SDE unit (Chaoyue Laboratory Instrument Works Co., Shanghai, China) is connected to the UMAE system (CW-2000, Shanghai Xintuo Analytical, China Shanghai Instrument Technology Co., Ltd). First, the seabuckthorn leaves powder 20.0 g were passed through an 80 mesh sieve and then added to the reaction flask (I) together with the ionic liquid, mixed by a stirring bar, and 50 ml of dichloromethane was put in a 100 ml flask (II). Flask I was subjected to ultrasonic/microwave treatment in the oven of the ILUMASDE apparatus and flask II was heated in a water bath at 45°C. After the extraction, the mixture in flask I was rapidly cooled to room temperature using a cold bath, filtered through a filter paper and then the filtrate was filtered through a 0.45 µm microporous membrane for subsequent HPLC analysis. The organic solvent in flask II was collected and dried using anhydrous sodium sulfate. EOs were obtained by removing solvent under reduced pressure, then were sealed and stored at 4°C for GC-MS analysis.

Conventional extraction procedure 2.3.1. HDE for EOs
In total, 20.0 g of seabuckthorn leaves powders and 200 ml of water were placed in the reaction flask, then the reaction flask was connected to the hydrodistillation unit. The reactor was heated for 4 h using an electric heating set, EOs were separated from the water, dried with anhydrous sodium sulfate, and then sealed at 4°C. Three operation replicates were performed.

Ethanol ultrasonic-assisted extraction (EUAE) for the main flavonoids
In total, 20.0 g seabuckthorn leaves powder were mixed with 200 ml of 60% (V:V) ethanol and then RU, QU, KA and IS were extracted by UAE method. The extracts were filtered through a filter paper after 60 min of extraction time. The filtrate was filtered through a 0.45 µm microporous membrane for subsequent HPLC analysis. Three operation replicates were performed.

Ionic liquid-based ultrasonic-assisted extraction (ILUAE) for the main flavonoids
In total, 20.0 g seabuckthorn leaves powder and appropriate ionic liquid solution 200 ml were mixed, and then UAE method was used to extract RU, QU, KA and IS with 200 W of ultrasonic power, 60°C extraction temperature, 60 min of extraction time. The filtrate was filtered through a 0.45 µm microporous membrane for subsequent HPLC analysis. Three operation replicates were performed.

Ionic liquid-based microwave-assisted extraction (ILMAE) for the main flavonoids
Seabuckthorn leaves powder 20.0 g and appropriate ionic liquid solution 200 ml were mixed, RU, QU, KA and IS were extracted by MAE method with 500 W of microwave power, 10 ml g −1 of liquid/solid ratio and 40 min of extraction time. The filtrate was filtered through a 0.45 µm microporous membrane for subsequent of HPLC analysis. Three operation replicates were performed.

Experimental design 2.4.1. ILUMASDE optimization using RSM
In order to obtain the optimal extraction effect of ILUMASDE, the main parameters of the extraction experiment were optimized under the condition of the optimum ionic liquid. Three main influencing factors were selected from the single-factor experiments as independent variables, i.e. liquid/solid ratio (X 1 ), reaction time (X 2 ) and microwave power (X 3 ). The effects of three factors on the extraction yields of four flavonoids and volatile oil in seabuckthorn leaves were studied using a three-factor-three-level BBD followed by RSM analysis. The second-order multivariate regression equation of the extraction yields of four flavonoids and EO in seabuckthorn leaves is as follows: Where X i and X j represent independent variables, Y is the response variable, β 0 , β i, β ii and β ij are the constants, regression coefficients of one term, quadratic terms and interaction terms, respectively. The actual and coded levels of the independent variables used in the experimental design are shown

GC-MS analysis of EOs
The EOs was analysed by a gas chromatography-mass spectrometry system (Agilent 7890A-Agilent 7000B, Agilent, Santa Clara, CA) with HP-5MS column (   2.6. HPLC determination of RU, QU, KA and IS The chromatographic system (Jasco, Tokyo, Japan) consisted of a HiQ Sil-C18 reversed-phase column (4.6 mm × 250 mm, 5 µm), a PU 980 pump and 1575 UV detector. The HPLC conditions were as follows: the mobile phase was methanol-acetonitrile-water (40 : 15 : 45, v/v/v) containing 1% formic acid, the column temperature was maintained at 30°C, the detection wavelength was 368 nm, the flow rate was 1.0 ml min −1 , the injection volume was 10 µl, and the run time was 15 min. HPLC chromatograms of standards and sample are shown in figure 3.

Scanning electron microscopy observation
The effect of the different extraction method on the microstructure of the plant material was observed using scanning electron microscopy (SEM). The dried extraction samples obtained after treatments by different methods were observed using a scanning electron microscope (Quanta-200 SEM, FEI Co., Hillsboro, OR, USA). The samples were fixed on aluminium stubs using adhesive tape then sputtered with gold using a sputter coater. All the samples were scanned under high vacuum conditions at an accelerating voltage of 12.5 kV (500× magnification).

Choosing an appropriate IL
The anion is considered to be the significant factor influencing the properties of ILs [35]. 1-Alkyl-3methylimidazolium-based ILs are widely used in sample preparation and compounds extraction from plant materials, thus, 1-Butyl-3-methylimidazolium-based ILs with the same concentrations but six different anions (Br − , Cl − , BF4 − , ClO 4 − , HSO 4 − and NO 3 − ) were used in UMASDE to identify the best IL anion [43]. The results (figure 4a) show that the yields of both EOs and the four flavonoids obtained using [C 4 mim]Br was higher than those obtained using other ILs. The results reflect the higher ability of the [C 4 mim]Br to dissolve cellulose in plant cells, which involves having the inter-and intra-molecular hydrogen bonds dissociated and new hydrogen bonds between carbohydrate hydroxyl protons and the IL anions formed [44,45]. Br − was therefore chosen as the IL anion for extracting RU, QU, KA, IS and EOs from seabuckthorn leaves. The length of IL alkyl chain affects water miscibility, and thus affects the extraction efficiency of compounds [46]. 1-Alkyl-3-methylimidazolium ILs containing the same anion, i.e. Br − , but different lengths (ethyl to decyl) of alkyl chains in the cation, were used in UMASDE to evaluate the effect of the alkyl chain length. The results (figure 4b) show that the cation alkyl chain length significantly affects the yields of the target compounds and the highest yields of all the five target compounds were obtained using [C 4 mim]Br. This may be due to the difference of nano-structuring of alkyl domains within the IL, which affected the intermolecular forces between IL and the target compounds. Based on the above analysis, [C 4 mim]Br was considered to be the best IL for use in further experiments. This may be due to the hydrogen bonds formed by [C 4 mim]Br and the target compounds, which affected van der Waals force between them.

Model building and statistical analysis
To optimize the interactions between the three variables (liquid/solid ratio X 1 , reaction time X 2 and microwave power X 3 ) and the yields of RU, QU, KA, IS and EOs, 17 experiments were performed. The experimental design and the results are shown in tables 1 and 2. The results for each dependent variable and their coefficients of determination (R 2 ) indicated that the proposed models were adequate, there was no significant lack of fit, and the R 2 for the yields were satisfactory (table 3). A model is well fitted to the experimental data when the model F-value is significant but the lack of fit is non-significant [47].
and 4 and Y 5 are yields of RU, QU, KA, IS and EOs, respectively. X 1 is the liquid/solid ratio (ml g −1 ), X 2 is the extraction time (min) and X 3 is the microwave power (W).

Analysis of the response contour
The effects of the independent variables and their interactions on the yields of RU, QU, KA, IS and EOs are shown in figure 5. It can be seen from figure 5 that the yield of target compounds increased first and then decreased with the increase of the liquid/solid ratio, microwave power and extraction time.
With the increase of liquid/solid ratio, the concentration difference of the target compounds in the liquid and solid phase increased, thus improving the mass transfer of materials to solvents, yields of the target compounds increased.
However, the liquid/solid ratio increases further, which decreases the rate of temperature rise. At lower temperature, it is unfavourable to dissolve and diffuse the target component in the liquid phase.                 Thereby, the yield of target compounds increased first and then decreased with the increase of the liquid/solid ratio. From figure 5, different microwave power resulted in the different yields of target compounds. This increase in the yields was attributed to the fact that microwave irradiation energy can enhance the extent to which the solvent penetrates into the solid matrix of the powdered sample [48]. However, too high microwave power caused an excessively high temperature inside the plant material, which destroyed the composition of the target compounds and led to the loss of volatile oil in the experimental process. Therefore, the yield of target compounds increased first and then decreased gradually with the microwave power.
Extraction time is a crucial parameter in solvent extraction for natural active ingredients. From figure 5, the yield of target compounds increased first and then decreased with the increase of extraction time, which is mainly due to the fact that, within a certain period of time, the longer extraction time made the target compounds into solvents completely. But the longer time led to decomposition of the target compounds, thus the extraction yield of target compounds decreased [49].
The optimum conditions of RU, QU, KA, IS and EOs for the maximum predicted yields by software were, respectively: liquid/solid ratio of 10.5 ml g −1 for RU, 10.8 ml g −1 for QU, 13.4 ml g −1 for KA, 11.2 ml g −1 for IS and 11.7 ml g −1 for EOs; extraction time of 32.7 min for RU, 34.0 min for QU, 32.7 min for KA, 37.5 min for IS; 9.09 mg g −1 for RU, 5.42 mg g −1 for QU, 3.14 mg g −1 for KA, 5.72 mg g −1 for IS and 0.095% for EOs.  microwave power 540 W. The suitability of the model equations for predicting the response values was confirmed by performing a verification experiment under these optimized conditions. The actual yields of RU, QU, KA, IS and EOs were 9.18 ± 0.35 mg g −1 , 5.52 ± 0.23 mg g −1 , 3.03 ± 0.11 mg g −1 , 5.64 ± 0.24 mg g −1 and 0.095 ± 0.004%, respectively. The RU, QU, KA, IS and EOs yields obtained using ILUMASDE were close to the predicted values and showed low deviations (less than 1.2%), demonstrating the reliability of the RSM models.

Effect of different extraction method on the yield of RU, QU, KA and IS
A comparison of ILUMASDE with other methods (ethanol ultrasonic-assisted extraction (EUAE), ionic liquid-based ultrasonic-assisted extraction (ILUAE) and ionic liquid-based microwave-assisted extraction (ILMAE)) was performed based on the yields of RU, QU, KA and IS ( figure 6). The yield of the four flavonoids obtained by ILUMASDE was much higher than those obtained by the other methods. It may be the IL dissolves the cellulose in the plant cell walls and the target compounds are highly soluble in the IL. In addition, the ultrasonic/microwave combination breaks the plant cells, which also accelerates the release of RU, QU, KA and IS from the matrix and it also increases the molecular motion of the extraction solvent, which increases mass transfer, leading to higher yields of the target compounds.

Effect of different extraction method on yields and components of EOs
Experimental results show that the EOs yields were 0.095 ± 0.004% by ILUMASDE and 0.089 ± 0.003% by HDE, respectively. The yield of EOs obtained by ILUMASDE was much higher than that obtained by HDE method. The relative contents of individual volatile components are expressed as percentages for their peak areas relative to the total peak area. The results of GC-MS analysis showed that the

Conclusion
ILUMASDE was proposed to isolate and quantify EOs and non-volatile RU, QU, KA and IS from the leaves of seabuckthorn. Satisfactory yields for the five target components were obtained through optimization by RSM with a BBD. Compared with traditional methods, ILUMASDE gives higher yields of target components while consuming a shorter extraction time. And it is worth mentioning that ILUMASDE accelerates the isolation of EOs without causing major changes in the EOs composition. Volatile and non-volatile active compounds co-exist in many plants. Simultaneous extraction of nonvolatile and volatile compounds from plants can reduce the operation steps of the extraction process. Therefore, the proposed effective ILUMASDE method is a promising technique for the simultaneous extraction of non-volatile and volatile compounds from other plants.