Extraction and purification of isochlorogenic acid C from Chrysanthemum morifolium using ionic liquid‐based ultrasound‐assisted extraction and aqueous two‐phase system

Abstract Ionic liquid‐based ultrasonic‐assisted extraction (IL‐UAE) was developed to extract and separate the isochlorogenic acid C (ICGA) from a cultivar of Chrysanthemum morifolium (Chrysanthemummorifolium Ra Tnat.). The influencing parameters, including IL concentration, liquid‐to‐solid ratio, and ultrasonic time, were optimized using response surface methodology. Of the ILs studied, 1‐butyl‐3‐methylimidazolium bromide [(Bmim)Br] exhibited the best extraction ability. The optimized conditions included liquid‐to‐solid ratio of 23.44:1, ultrasonic time of 48.99 min, and IL concentration of 0.65 mol/L. Under the optimal conditions, the extraction yield of ICGA could reach to 4.20 mg/g. An aqueous two‐phase system was applied for purification and separation of ICGA. The maximum extraction efficiency of 98.18% was obtained under the conditions of (NH 4)2 SO 4 of 4.5 g, pH of 3.0, and a temperature of 20°C at aqueous solution. Furthermore, the thermodynamic parameters showed that the purification of ICGA from salt‐rich phase to IL‐rich phase was a spontaneous and exothermic process. The results indicated that the proposed system is simple, rapid, and effective to serve as a viable and sustainable platform for the extraction and purification of ICGA from Chrysanthemum morifolium flowers.

extraction (UAE) (Mazvimba, Ying, Cui, & Zhang, 2012), and pressurized liquid extraction (Wianowska, Typek, & Dawidowicz, 2015). However, there were very few researches aiming at extraction of ICGA. Meanwhile, traditional methods have suffered from many disadvantages such as high costs of instrumental consumption, operation, and maintenance. Moreover, the extraction and separation of these traditional methods need to be divided into two parts, which prolonged the reaction time.
Aqueous two-phase system (ATPS), which is usually composed of polymer/polymer, polymer/salt, or salt/salt, has been recognized as an economical and efficient processing method (Atefi, Joshi, Mann, & Tavana, 2015). It provides a mild operating environment. Special structures like vesicles presented in an ATPS can also facilitate the separation process (Lu et al., 2015). Therefore, the utilization of ATPS is considered as a promising separation technique (Dong et al., 2015). In addition, ionic liquids (ILs) have emerged as alternatives for conventional organic solvents due to their characteristics such as negligible vapor pressure and volatility, low toxicity, adjustable polarity, general nonflammability, ease of recycling and manipulation, and high thermal and chemical stability (Lv, Jiang, Li, & Ren, 2012).
Moreover, ILs can be designed by different combinations of cations, anions, and functional groups to allow for different purification processes. These designed ILs can be called task specific or functionalized ILs (Yang, Tan, Li, & Li, 2016). Thus, ILs have the properties of high extractability for both organic and inorganic compounds.
Ionic liquid-based aqueous two-phase system (IL-ATPS), which is developed by Gutowski et al. (2003), is usually composed of hydrophilic ILs and inorganic salts, and IL-ATPS has been widely applied as a more efficient and greener way for extraction and purification in one single procedure of various compounds, such as food additives and veterinary pesticides of agricultural products (Fan et al., 2014), pharmaceutical biomolecules (Tan, Li, Xu, & Xing, 2012), biochemical esterase (Lee, Khoiroh, Ling, & Show, 2017), heavy metal ions (Zheng, Tong, Wang, Zhang, & Yang, 2015), and protein (Tan et al., 2012).
In this work, an optimized single-step procedure of IL-UAE-ATPS was studied to extract ICGA from a cultivar of Chrysanthemum morifolium. The studied IL was (Bmim)Br (1-butyl-3-methylimidazolium bromide), the chemical structure of which is presented in Figure 1.
The conditions (IL concentration, liquid-to-solid ratio, and ultrasonic time) were optimized using the response surface methodology (RSM).
The key parameters including the salt type and amount, pH, and temperature were investigated in terms of their effect on the purification of ICGA. Moreover, the thermodynamics was explored simultaneously in the extraction procedure for the first time. The whole flowchart of extraction and purification of ICGA is shown in Figure 2.

| Ionic liquid-based ultrasonicassisted extraction
Dried flowers were pulverized into a homogeneous size by a grinder, sieved through a 60 mesh sieve, and dried at 45°C in the oven for 6 hr. Powders and IL solutions were mixed with a certain ratio in a centrifuge tube. The blend was placed in an ultrasonic bath at 25°C for some time to ensure the dissolution of the active ingredients.
The extract solution was centrifuged at 12,000 g for 10 min, and the supernatant was filtrated through a 0.45μm microporous membrane and then injected into HPLC for analysis. The extraction yield of ICGA by ionic liquid-based ultrasonic-assisted extraction (IL-UAE) was calculated according to Equation 1: where the m is the mass (mg) of ICGA determined and M is the mass (g) of dried sample powder.

| Ionic liquid-based aqueous two-phase system
Ten milliliters extraction solution and a given amount of salt were added to a 25-ml centrifuge tube. The mixture was vortexed for 3 min to ensure the complete dissolution of the salt. Due to the incompatibility of IL solution and salt solution, and the low viscosity in the system, the phase separation can be achieved in a few seconds. The top IL-rich phase was mainly composed of IL and ICGA (1) Extraction yield = m∕M F I G U R E 1 Chemical structures of (Bmim)Br 1-butyl-3methylimidazolium bromide (a) and IGCA (b) occupying a small volume, and the bottom phase was salt-phase solution containing the aqueous impurities occupying a large volume.
The volume of each phase was noted down.

| Analysis of ICGA
The two phases were separately withdrawn using pipettes for determining the concentration of ICGA by HPLC. The residues accumulated at the interface between two phases were discarded. All experiments were carried out at 25°C. The parameters are defined as follows.
The phase ratio (R), partition coefficient (K), and extraction efficiency (E) were evaluated. The parameters were defined as follows: where the V t and V b are the volumes of IL-rich phase and salt-rich phase, respectively, and C t and C b denote the measured ICGA concentration (mg/ml) in the IL-rich phase and salt-rich phase, respectively.

| HPLC analysis
Chromatographic analyses were performed using an Agilent 1260 series HPLC system equipped with an Agilent Extend-C 18 column (4.6 mm × 250 mm, 5 μm). The injection volume was 5.0 μl, the temperature of the column oven was controlled at 30°C, and the flow rate of the mobile phase was maintained at 1 ml/min. The mobile phase consisted of 0.1% formic acid water (A) and 0.2% formic acid ACN (B) using a gradient elution of 10%-19% A at 0-16 min, 19%-20% A at 16-24 min, 10%-24% A at 24-48 min, 24%-50% A at 48-55 min, and 50%-10% A at 55-60 min. The UV wavelength was set at 384 nm. The data acquisition and analysis were performed by Agilent ChemStation software.

| Comparison between IL-UAE and other conventional extraction methods
In order to further compare the extraction yield of conventional heat reflux extraction (HRE) and UAE, the extractions of ICGA based on different solvents ([Bmim]Br aqueous solution, methanol, ethanol, and water) were performed. All the experiments were performed at the same conditions. Briefly, 0.42 g sample powder was mixed with 10 ml methanol, ethanol, or water ultrasonic treatment for 48.99 min. The extraction time for HRE was 2 hr at 80°C. The subsequent steps were the same as those in IL-UAE.

| Single-factor variation
Ionic liquid is usually composed of an organic cation in combination with an inorganic anion. The structure of IL has a significant impact on the extraction yield, and different structures may lead to a different extraction yield of analytes. According to Yang et al. (2013), the effect of changing the type of anion and the alkyl chain length of the cation of IL on the extraction yield was studied, and it was found that [Bmim]Br had the best extraction yield compared with other ILs. Therefore, [Bmim]Br was chosen for the subsequent study.
Structurally, the π-bond of the heterocyclic structure in ILs is rich in numerous electrons, which can produce intermolecular binding through n-π or π-π stacking and promote the dissolution of effective substances in the samples (Chen, Cao, Gao, Qi, & Li, 2013).
At the same time, ILs can help expansion and fragmentation of cell To some extent, the liquid-to-solid ratio and ultrasonic time play significant roles in promoting the extraction yield of the targeted compounds during the extraction. In general, a higher volume of solvent leads to a higher extraction yield. However, excessive solvent may dilute the extracts, which results in waste of solvents.
Conversely, insufficient solvent may cause an incomplete extraction.
As for ultrasound time, moderate sonication time is used to get maximum extraction yield. As shown in Figure 3, a liquid-to-solid ratio of 20:1 and ultrasonic time of 50 min were selected in the subsequent experiments.

| Optimization of IL-UAE conditions by RSM
To further study the interactions between the various factors, a Box-Behnken design approach and desirability function were used to optimize the IL concentration, liquid-to-solid ratio, and ultrasonic time in the IL-UAE system. The Design-Expert software was used. The experimental model defined the amount of different factors to be used to extract ICGA. The ranges for the extraction conditions were as follows: (A) liquid-to-solid ratio, 10:1-30:1; (B) ultrasonic time, 40-60 min; and (C) IL concentration, 0.25-0.75 mol/L. Table 1 shows the conditions of the 17 runs and the corresponding values of the test responses.
As shown in Table 2 concentration was found to be the most significant factor to affect the extraction yield of ICGA, followed by solid-to-liquid ratio and ultrasonic time. The quadratic terms (A 2 and C 2 ) were highly significant (p < 0.01), and B 2 (p < 0.05) was also significant. Among the interaction terms, the interaction between AB (p < 0.01) was very significant with regard to the extraction yield of ICGA. The others were F I G U R E 3 Single factor affecting the extraction yield of IGCA, including IL concentration (a), liquid-to-solid ratio (b), and ultrasonic time (c) not significant (p > 0.05). The "lack-of-fit F-value" of 31.21 implied that the "lack of fit" was significantly relative to the pure error and that there was a 0.31% chance that a "lack-of-fit F-value" this large

| Comparison of IL-UAE approach with conventional extraction methods
To further demonstrate the extraction yield of the IL (Bmim[Br]), a comparison was made between the proposed IL-UAE and conventional UAE and HRE methods. The results are summarized in Figure 5.

| Extraction and purification of ICGA by ATPS
Following the extraction of ICGA by IL-UAE, it was further purified by ATPS. The type and amount of inorganic salt, temperature, and pH can be considered as the prominent factors influencing the purification efficiency of ICGA by ATPS.

| Effect of salt type and amount
In order to choose an optimal inorganic salt for the enrichment To optimize the amount of (NH 4 ) 2 SO 4 added to form Bmim[Br]/ (NH 4 ) 2 SO 4 , the range of (NH 4 ) 2 SO 4 (from 4.5 to 6.5 g) was investigated. When the amount of (NH 4 ) 2 SO 4 was too low, ATPS could not be formed and the system was a homogeneous solution. With the amount of salt gradually increased, the solution was cloudy. Until the amount reached to 4.5 g, phase separation was formed. The upper phase was formed by the more hydrophobic IL solution, and the F I G U R E 4 3D response surface plots showing the effects of variables on the average extraction yield of isochlorogenic acid C: (a) interaction between [Bmim]Br concentration (mol/L) and solid-to-liquid ratio (ml/g); (b) interaction between [Bmim]Br concentration (mol/L) and ultrasonic time (min); and (c) interaction between ultrasonic time (min) and solid-to-liquid ratio (ml/g) lower phase was consisted of the more hydrophilic salt solution. The amount of salt also directly affected the volume of the upper phase, thereby affecting the purification effect. As shown in Figure 6a, with the amount of (NH 4 ) 2 SO 4 increased, there was a slight decrease in the extraction efficiency of ICGA. This phenomenon may be due to the enhancement of salting-out effect as the concentration increased in the bottom phase, and more free water molecules enter into the bottom phase, which leads to the increased viscosity and decreased contact area of the upper phase. Thus, the entry of target compound into the IL-rich phase becomes more difficult, and the extraction efficiency decreased. Based on economic factors and extraction efficiency, 4.5 g of (NH 4 ) 2 SO 4 was selected to be the suitable amount.

| Effect of pH
According to Tan et al. (2014), the pH of the aqueous solution was

| Effect of temperature
Temperature plays a crucial role in the mass transfer and solubility. A relative high temperature can enhance the mass transfer of target compounds and is also beneficial to the dispersion of IL.
The results in Figure

| Thermodynamic parameters in extraction procedure
The parameters of thermodynamics during purification process were investigated to study the transfer process of ICGA from saltrich phase to IL phase. The Gibbs energy of transfer (∆G), from one phase to another, may be associated with its partition coefficients by Equation 6, where R is the universal gas constant and thermodynamic temperature (T) and distribution coefficient (K) were known.
As shown in Equation 7, a linear correlation relation curve between ln K and 1/T was found within the temperature range of 20-40°C.  Table 3. All the ∆G values were negative (∆G < 0), which indicated that the reaction of purification proceeded spontaneously from salt-rich phase to IL-rich phase.
Partitioning of ICGA in IL-ATPS was marked by negative values for ∆H and TΔS, in which the ∆H could have greater value than the TΔS. Hence, the purification of ICGA in IL-ATPS from salt-rich phase to IL-rich phase was a spontaneous and exothermic process (Tan et al., 2014). The signatures of thermodynamics were agreement with the experimental results.

| CON CLUS ION
In this work, IL-UAE coupled with the IL-ATPS could be effec-

ACK N OWLED G M ENTS
The research was supported by Shanghai Science and Technology Innovation Action Plan Project (No. 16391904400).

CO N FLI C T O F I NTE R E S T
The authors declare that they do not have any conflict of interest.

E TH I C A L R E V I E W
This study does not involve any human or animal testing.