Rapid Determination of Two Triterpenoid Acids in

ANALYTICAL SCIENCES APRIL 2018, VOL. 34 407

Introduction

Oleanoic acid (OA) and ursolic acid (UA) are main triterpenoid acids (Table 1) that exist in many plants as dietary and medicine sources.1 These two triterpenoid acids are abundant in Chaenomelis Fructus,2,3 which is widely used as a traditional Chinese medicine for dysentery, backache and hepatopathy.4 It has been confirmed that OA and UA have a wide range of bioactivities5–7 and several analytical methods have been reported for the analysis of OA and UA, including thin-layer chromatography (TLC), gas chromatography (GC), capillary electrophoresis (CE) and high-performance liquid chromatography (HPLC).8,9 However, most of these methods are carried out with derivatization and a long analysis time with low resolution or sensitivity.9 In addition, the conventional sample pretreatment usually requires a large consumption of hazardous organic solvents and labor, which is not only uneconomical and time-consuming, but also leads to the sample losing or decomposing during sample pretreatment. Therefore, in order to ensure efficient analysis, it is necessary to establish a rapid and convenient method for the quantitative determination of OA and UA in Chaenomelis Fructus.

Supercritical fluid chromatography (SFC) is a powerful

separation technique by using the supercritical fluid (SCF) as a mobile phase. Nowadays, supercritical carbon dioxide (SC-CO2) is used as a common supercritical solvent.10 Due to the low viscosity, high diffusivity and high solvating power of SC-CO2, SFC generally has a higher column efficiency, better selectivity and shorter analysis time than the traditional GC and HPLC.11 Supercritical fluid extraction (SFE) is an environmentally friendly extraction technology that greatly simplifies sample pretreatment. Compared with other extraction methods, SFE requires a less amount of solvents and a shorter extraction time. Moreover, SFE is suitable for thermally and oxidation sensitive compounds, which require mild extraction conditions.12 SFC and SFE have been widely utilized to separate a variety of natural and pharmaceutical products, including lipids, vitamins, alkaloids, saponins and triterpenoids.13–20

In view of the advantages of SFE and SFC in extraction and analysis, SFE coupled with SFC (SFE-SFC) is considered to be an efficient and green alternative compared to conventional analytical technologies.21,22 SFE and SFC can be directly connected and the efficiency extraction and separation can be carried out simultaneously. Unfortunately, even though the coupling SFE-SFC has been studied for more than 20 years, this approach has not been widely used due to the complex operation and the unstable extracts trapping apparatus. Andrea et al.23 reviewed reports of on-line coupling SFE-SFC since the 1990s, yet no on-line SFE-SFC method has been described for the qualitative or quantitative analysis of triterpenoids in medicinal plants.

2018 © The Japan Society for Analytical Chemistry

† To whom correspondence should be addressed.E-mail: [email protected] (S. C.); [email protected] (Z. L.)

Rapid Determination of Two Triterpenoid Acids in Chaenomelis Fructus Using Supercritical Fluid Extraction On-line Coupled with Supercritical Fluid Chromatography

Xiaotian ZHANG,* Feng JI,** Yueqi LI,** Tian HE,* Ya HAN,* Daidong WANG,* Zongtao LIN,***† and Shizhong CHEN*†

* School of Pharmaceutical Sciences, Peking University, Beijing 100191, China ** Analytical Center, Department of Analytical Instruments, Shimadzu (China) Co., Beijing Branch,

Beijing 100020, China *** Department of Pharmaceutical Sciences, University of Tennessee Health Science Center, Memphis,

TN 38163, USA

In this study, an on-line supercritical fluid extraction (SFE) and supercritical fluid chromatography (SFC) method was developed for the rapid determination of oleanoic acid and ursolic acid in Chaenomelis Fructus. After optimization of the conditions, the two triterpenoid acids was obtained by SFE using 20% methanol as a modifier at 35°C in 8 min. They were resolved on a Shim-pack UC-X Diol column (4.6 × 150 mm, 3 μm) in 14 min (0 – 10 min, 5 – 10%; 10 – 14 min, 10% methanol in CO2) with a backpressure of 15 MPa at 40°C. The on-line SFE-SFC method could be completed within 40 min (10.79 mg/g dry plant, Rs = 2.36), while the ultrasound-assisted extraction and HPLC method required at least 90 min (3.55 mg/g dry plant, Rs = 1.92). This on-line SFE-SFC method is powerful to simplify the pre-processing and quantitative analysis of natural products.

Keywords On-line method, Chaenomelis Fructus, pentacyclic triterpenoids, supercritical fluid extraction, supercritical fluid chromatography

(Received August 27, 2017; Accepted October 24, 2017; Published April 10, 2018)

Original Papers

408 ANALYTICAL SCIENCES APRIL 2018, VOL. 34

In this study, we have developed an on-line SFE-SFC system for the rapid determination of OA and UA in Chaenomelis Fructus. We comprehensively investigated the optimal method parameters for the determination of OA and UA. This SFE-SFC method has also been validated and compared with conventional HPLC analysis.

Experimental

Materials and reagentsThe reference compounds of oleanoic acid and ursolic acid

were purchased from Chengdu Must Bio-Technology Co., Ltd. (Chengdu, China). The dried fruits of Chaenomelis Fructus were obtained from Tongrentang Chinese Medicine (Beijing, China) and authenticated by Prof. Hong Wang, Department of Natural Medicines, Peking University. A voucher specimen (ZXT-1611) was deposited at the Herbarium of School of Pharmaceutical Sciences, Peking University (Beijing, China).

Methanol, ethanol, isopropanol, acetonitrile (Merck, Darmstadt, Germany), trimethylamine, formic acid and acetic acid (Sigma-Aldrich, MO, USA) were of HPLC grade. The SFC-grade carbon dioxide (CO2, purity >99.99%) was purchased from Beijing Green-oxy Tiangang Technology Development Co., Ltd. (Beijing, China). Nexera UC extraction dispersant was provided by Shimadzu (Tokyo, Japan). Purified water was prepared using a Milli-Q water purification system (Millipore, MA, USA).

Preparation of sample and standard solutionsThe dried fruits of Chaenomelis Fructus were finely crushed

and passed through different mesh sieves (80, 120, 160 and 200). Approx. 20 mg of two reference compounds were, accurately weighted, dissolved in methanol to produce a mixed standard stock solution (5 mg/mL for each compound). Then, 0.5 g crude drug powder (120 mesh) was, accurately weighted, extracted with 25 mL methanol in an ultrasonic water bath for 20 min, and then filtered to produce sample solution of organic solvent extraction. The sample powder and solutions stored at 4°C. The above solutions were filtered through 0.22 μm membranes before use.

Analytical columnsSeparations were carried out on the following columns (GL

Sciences, Tokyo, Japan): Sil Shim-pack UC-X Sil (2.1 × 150 mm, 3 μm); Diol Shim-pack UC-X Diol (4.6 × 150 mm, 3 μm); NH2 Shim-pack UC-X NH2 (4.6 × 150 mm, 3 μm); C18 Shim-pack UC-X RP (4.6 × 150 mm, 3 μm); CN Inertsil CN-3 (4.6 × 250 mm, 2 μm). The Inertsil ODS-P (C18, 4.6 × 250 mm, 5 μm) was used for RP-HPLC analysis. The above columns were provided by Shimadzu (Kyoto, Japan) and Shimadzu-GL Sciences (Beijing, China).

Apparatus and conditions of the Nexera UC systemAll experiments were performed on a Shimadzu Nexera UC

SFE-SFC system (Kyoto, Japan) consisting of a CBM-20A controller, a DGU-20A5 degasser, an LC-30ADSF CO2 pump, two LC-20ADXR modifier pumps, an SIL-30AC autosampler

Table 1　Structures and physicochemical properties of OA and UAa

Compound Formula Structure M/g mol–1 λmax/nm Density/g cm–3 Melting point/°C Boiling point/°C

OA C30H48O3

456.7 210 1.1 310 553.5

UA C30H48O3

456.7 210 1.09 281 556.9

a. The molar mass, wavelength of maximum absorbance, density, melting point and boiling point of the compounds were obtained from SciFinder.

Fig. 1　Schemes of the on-line SFE-SFC-DAD-MS system.


(with 5 μL sample loop), a CTO-20AC column oven, an SPD-M20A diode-array detector (with high pressure cell) and two SFC-30A backpressure regulations (BPR). As shown in Fig. 1, the Nexera UC system can be switched to SFE, SFC and on-line coupled SFE-SFC modes according to changes of the two BPRs. All data were processed by Shimadzu Labsolution Ver. 5.8.6 software.SFE mode. In SFE mode, BPR-A pressure (40 MPa) was much higher than BPR-B. The extracts were extracted by SC-CO2 in an extraction vessel (static extraction). The fresh mobile phase caused the extracts to enter the fraction collector from BPR-B (dynamic extraction) to complete the entire extraction process.SFC mode. In SFC mode, BPR-B pressure (40 MPa) was much higher than BPR-A. Sample solution was injected by autosampler and separated by the column, while the chromatogram was obtained by DAD detector. After the sample had been discharged from BPR-A, it could also couple with mass spectrometer to obtain more structure information of unknown compounds.On-line SFE-SFC mode. In this on-line mode, two BPR pressures were firstly switched to the SFC mode, and samples were subjected to static extraction. In the following dynamic extraction, the BPR-B pressure was switched to be slightly higher than BPR-A. Most of the extract was discharged from BPR-B, and quantitatively trapped at the head of an analytical column as a narrow band.22 Upon completion of the dynamic extraction, two BPR pressures were switched back to the SFC mode. The trapped analytes were analyzed to complete the on-line SFE-SFC mode.

In all trials, the herbal sample (50 mg, 120 mesh) was well mixed with Nexera UC extraction dispersant, and then filled into a 5-mL extraction vessel. The purpose of mixing was to improve sample-solvent contact, and to avoid channeling when CO2 flowed through the mixture.16 In the static extraction of SFE mode, 20% methanol in SC-CO2 was delivered into extraction vessel in 1 min (5 mL/min) at 35°C. The pressures of BPR-A and BPR-B were 15 and 40 MPa, respectively. The static stage was to obtain sufficient extract in 5 min. In the following dynamic extraction, the extract was trapped on the column head by fresh mobile phase (5% methanol in SC-CO2) at 5 mL/min for 2 min, and the pressures of BPR-A and BPR-B were 14.8 and 15 MPa, respectively. Under these system conditions, the ratio of the extracts that distributed to the column was 2.36%.

After the SFE mode, the SFC mode was carried out on a Shim-pack UC-X Diol column (0 – 10 min, 5 – 10%; 10 – 14 min, 10%; 14 – 16 min, 10 – 40%, 16 – 20 min, 40% methanol in CO2, v/v) at 40°C. The flow rate was 2 mL/min, the detection wavelength of UV was set at 210 nm, and the pressure of BPR-A and BPR-B were 15 and 40 MPa, respectively.

Optimization of SFE conditionsIn this on-line SFE-SFC system, static extraction was the

main process affecting the yield, while the dynamic extraction was to enrich the extract. In order to obtain the optimal SFE extraction efficiency of OA and UA from Chaenomelis Fructus, we investigated the effects of four operating parameters, including the static extraction time (A), temperature (B), modifier concentration (C) and mean particle size (D). A L9 (34) orthogonal array design was applied to optimize these parameters. The experimental factors and levels are shown in Table 2. All trials were determined by on-line SFE-SFC and performed in triplicate (n = 3). The data were expressed as “mean ± standard deviation”, and SPSS 20.0 variance analysis was used to compare the differences between groups.

Validation of optimized methodThe LOD and LOQ were determined by analysis of dilutions

of mix standard solution until the signal-to-noise ratio (S/N) was equal to 3 and 10, respectively. The precision of the method was evaluated by the relative standard deviation (RSD) of the peak areas when analyzing the same mixed standard solution for 6 consecutive times. The repeatability were evaluated by the RSD of peak areas when analyzing six parallel samples. The accuracy was evaluated by spiking experiments; three portions of sample were added of different concentrations (high, 4 mg/mL; medium, 3.5 mg/mL; low, 3.0 mg/mL) of the mix standard solution, and then analyzed under above on-line conditions. The estimated concentrations of the mix standard solution were compared with the theoretical concentrations, and the accuracy was expressed as the recovery rate.

Results and Discussion

Optimization of SFC modeThe SFC conditions, including the stationary phase, mobile

phase and system parameters, were optimized. First, five columns (Sil, Diol, NH2, C18, CN) were selected to separate mix standards of OA and UA under the above-mentioned conditions in the SFC mode (0 – 10 min; 0 – 20% methanol in CO2, v/v). We found that OA showed a stronger retention than UA. The retention times of OA in different columns followed the order of Sil > C18 > CN > Diol > NH2, and the order of the chromatographic resolution (Rs) was Diol > C18 > NH2 > Sil > CN. Mixed standards were eluted in 9 min on all five types of columns, and their resolutions on C18 and Diol columns were 1.61 and 1.76, respectively, to meet the baseline separation requirements (Fig. 2). Since the resolution of sample on the C18 column was sharply declined in the SFE-SFC mode, which may be due to weak retention, the Diol column was chosen to optimize the conditions in the SFC mode.

Secondly, four organic modifiers were compared for the separation of OA and UA, including methanol, ethanol, isopropanol and acetonitrile. The results showed that methanol, ethanol and isopropanol had similar selectivity, and elution ability on the SFC chromatographic behavior. The resolution between OA and UA followed the order of methanol (1.85) > ethanol (1.80) > isopropanol (1.78). However, the compounds could not be eluted within 20 min when acetonitrile was used as a modifier. As we know, modifiers with hydrogen-bond capability character could weaken the hydrogen-bond interaction between the analytes and stationary phase, and thus resulted in weaker retention of the analytes.24 Methanol was chosen as a modifier, which dissolved standard compounds as well. In addition, we also tried to add 0.1% formic acid into methanol as an additive, but it did not significantly enhance the resolution.

Table 2　Factors and levels of L9 (34) SFE orthogonal test in this study

Level

Factor

A (static time/

min)

B (temperature/

°C)

C (modifier ratio,%)

D (sample particle

size/mesh)

1 1 35 10 802 3 40 15 1203 5 45 20 160


Finally, the factors including the flow rate (1.0, 1.5, 2.0, 2.5, 3.0 mL/min), BPR-A pressure (10, 15, 20, 25 MPa) and column temperature (35, 40, 45, 50°C) were optimized (Fig. 3). For OA and UA, the significantly decreased retention times, capacity factor and chromatographic resolutions were observed as the flow rate and BPR pressure increased. A higher flow rate and backpressure would enhance the density of supercritical fluid (SCF), and thus enhance its eluotropic strength.25 The higher column temperature could extend the retention times, and generally improve the chromatographic resolution of OA and UA.20,24

Optimization of SFE modeThe effects of the static time (1, 3, 5, 7, 9 min) were

investigated to optimize the SFE process. As shown in Fig. 4A,

the extraction yields of OA and UA increased significantly (P <0.05) when the static time increased from 1 to 9 min, as expected. The highest yields of OA and UA were 5.34 and 0.93 mg/g (in dry material). This is because SCF could more sufficiently permeate the material with a longer static time, and to extract target compounds easily. However, a shorter static time facilitates the analysis speed and energy efficiency.16 Taking them together, the final selection of the extraction time is 5 min.

The influence of the extraction temperature on the yields of OA and UA was evaluated under fixed conditions: static time of 3 min, modifier concentration of 10% and mean particle size of 120 mesh. In Fig. 4B, the extraction temperatures were operated at 35, 40, 45, 50 and 55°C, and the yields of OA and UA were temperature-dependent and significantly increased in the range

Fig. 2　Separation of OA and UA on different stationary phase columns. A, NH2 column; B, Diol column; C, CN column; D, C18 column; E, Sil column.

Fig. 3　Effect of the column temperature, flow rate and BPR pressure (B) on the retention time and chromatographic resolution of OA and UA. BPR pressure 15 MPa (A); flow rate 2 mL/min (B); Shim-pack UC-X Diol (4.6 × 150 mm, 3 μm) column; (0 – 10 min, 5 – 10%; 10 – 25 min, 10%) methanol in CO2 as modifier; column temperature 40°C; UV 210 nm.


of 35 – 45°C (P <0.05), while the yield were not further improved when the temperature further increased from 45 to 50°C. There were several concurrent factors affecting the extraction efficiency if the temperature rose. The increasing temperature led to a decreased density of SCF, thus subduing the solvent power to dissolve the substances. On the other hand, elevated temperature could increase the SCF diffusivity and the vapor pressure of the solute, which enhanced the solubility and the interaction between extracts and SCF. However, in previous reports, sometimes the SFE extraction yield was consistently correlated with the temperature positively or negatively,26,27 and sometimes there was an optimum extraction temperature in SFE,28 which might be caused by the solubility and the interaction between different solutes and SCF. The optimal column temperature was set at 45°C, since the extraction temperature close to column temperature could reduce the sample solvent effect and minimize the noise signals.

To determine the effect of the modifier concentration on the extraction of OA and UA, the modifier concentrations were studied at 5, 10, 15, 20 and 25% of methanol in SC-CO2 (Fig. 4C). It was found that the addition of methanol slightly enhanced the yields of OA and UA, and the maximum extraction yield was achieved at 25% methanol. Since SC-CO2 had a polarity similar to n-hexane, pure SC-CO2 was not efficient to extract polar triterpenic acids. The addition of organic modifiers significantly improved the solvation power of pure SC-CO2 to obtain compounds with higher polarity and larger molecular weight. However, the addition of modifiers to SC-CO2 would decrease the density and diffusivity of SCF, which would decrease the interaction between the sample matrix and SCF. The results indicated that the addition of modifiers could increase the extraction efficiency of OA and UA. In view of the resolution on SFC was decreased significantly when excessive proportion modifier used, 15% was selected as the optimum

ratio of the modifiers in SFE.The effect of the material particle size on the yields of OA and

UA was investigated at different mean particle sizes (80, 120, 160 and 200 mesh). It was found that the particle size remarkably influenced the extraction efficiency of OA and UA (Fig. 4D). The extraction yield incipiently increased (P <0.05) when the material particle size was changed from 80 to 160 mesh, but distinctly decreased (P <0.05) when the particle size decreased to 200 mesh. Similar results were given in a previous report, in which the particle size had obvious effects on SFE yield.29 The small particle size could decrease the diffusion pathway and increase the extraction interfacial area between SCF and the sample matrix. On the contrary, the smaller particle size led to a stronger re-adsorption effect of the solutes, which resulted in a decreased extraction efficiency of triterpenic acids. The optimal particle size was selected at 160 mesh.

Orthogonal screening and analysisAn orthogonal screening test was designed and tested on each

factor. The data of nine trials are summarized in Table 3, and analyzed by SPSS 20.0 to evaluate the effect of each factor on SFE process. The total yield values of factors in different levels were calculated for the range analysis (Table 3). It was found that the effect of the factors on the total yields of OA and UA is in the order of A > D > B > C based on the range (R) values. These results indicated that the static extraction time and the particle size were the significant determinants of the OA and UA yields, and the influence of extraction temperature and modifier concentration on the extraction yield was remarkable. The optimal levels of four factors for the OA and UA extraction by SFE were 5 min, 120 mesh, 35°C and 20% modifiers in SC-CO2.

Table 4 shows the analysis of variance (ANOVA) of the orthogonal experiments. All four factors had a statistically

Fig. 4　Effect of the static extraction time (A), extraction temperature (B), modifier ratio (C) and mean particle size (D) on the SFE yield of OA and UA in Chaenomelis Fructus. The experimental conditions are described in the relevant sections.


significant effect on the SFE process with 95% confidence, and the order of the influence factors was consistent with the results of an intuitive analysis. According to the above results of an ANOVA analysis, the optimal SFE method was conducted at a static extraction time of 5 min, an extraction temperature of 35°C, a modifier concentration of 20% and a mean particle size of 120 mesh; 50 mg of sample was extracted with SFE, and the total yield of OA and UA were obtained at 8.67 and 2.12 mg/g, respectively.

Validation of optimal on-line methodThe results of the validation are given in Table 5. LOD and

LOQ were determined to be less than or equal to 0.5 and 1.0 μg, respectively. The calibration curves were constructed by plotting the absorption peak area of OA (Y1) and UA (Y2) versus the corresponding injection concentrations (X, μg/mL). The calibration curves were linear (R2 = 0.9999) in the range of 1.0 – 500.0 μg/mL for both OA and UA. For precision and reproducibility, the relative standard deviations (RSDs) of OA and UA peak areas were 0.89 and 0.85% (precision), 2.07 and 2.07% (reproducibility), respectively. The recovery rates of three levels (low, medium and high) of spiked samples were between 96.0 and 104.2%. The data indicated that the established method was sensitive and reliable for the on-line analysis of OA and UA.

Comparison of solvent extraction-RPHPLC and on-line SFE-SFC in determination of OA and UA

The extract was quantified by solvent extraction (SE)-RPHPLC to compare with the results from on-line SFE-SFC method (Fig. 5 and Table 5). An aliquot of 5 μL sample was

analyzed on the same Nexera UC instruments.4 An Inertsil ODS-P column was used for sample separation with an isocratic elution (methanol/water/acetic acid/trimethylamine = 260/40/ 0.1/0.05). The flow rate was 1.0 mL/min, the column temperature was 17°C and the UV detection wavelength was set at 210 nm. The organic solvent extraction gave yields of 2.87 and 0.68 mg/g for OA and UA, respectively, and separated in 35 min with a resolution of 1.92. While the on-line SFE-SFC method gave extraction yields of 8.67 and 2.12 mg/g for OA and UA in 7 min, and the separation was completed in 20 min with a higher resolution of 2.36 than HPLC. In addition, the LOD and linearity range for OA and UA are achieved at the level of 0.5 and 1.0 – 500.0 μg/mL by SFE-SFC, which is superior to the SE-RPHPLC results (4.17 μg/mL for LOD and 20.0 – 180.0 μg/mL for linearity range). The RSDs of precision and reproducibility determined by SFE-SFC is higher than SE-RPHPLC. These results may due to the sampling error of the SFE extracted sample in each trials. Therefore, this on-line SFE-SFC method is preferable to simplify the sample pretreatment and analysis of OA and UA from Chaenomelis Fructus, rather than SE-RPHPLC.

Conclusions

In this study, an efficient method for the simultaneously separation and determination of OA and UA in Chaenomelis Fructus by using on-line SFE-SFC was established for the first time. Under the optimized on-line method, the high yields of 8.67 and 2.12 mg/g for OA and UA were determined within 20 min with a favorable resolution of 2.36. The on-line SFE-SFC validation data is also superior to the SE-RPHPLC results. These results showed that this on-line SFE-SFC method provided a promising and efficient tool to analyze the triterpenoid acids in Chaenomelis Fructus. This analytical system is useful for the rapid quality testing of the main ingredients in herbal medicines and the application to further investigation in complex natural products is extensively.

Table 3　Results of L9 (34) SFE orthogonal array test and range analysis

LevelFactor Overall yield of

OA and UA/mg g–1A B C D

1 1 1 1 1 4.482 ± 0.0862 1 2 2 2 6.024 ± 0.0603 1 3 3 3 6.107 ± 0.0994 2 1 2 3 7.558 ± 0.0525 2 2 3 1 5.328 ± 0.1226 2 3 1 2 7.396 ± 0.0557 3 1 3 2 10.787 ± 0.0498 3 2 1 3 10.148 ± 0.0259 3 3 2 1 6.153 ± 0.024

K1 5.538 7.609 7.342 5.321The optimal conditions: A3B1C3D2

The influence on the extraction yield: A > D > B > C

K2 6.761 7.167 6.578 8.069K3 9.029 6.552 7.407 7.938R 3.491 1.057 0.829 2.748

Table 4　Analysis of variance for the experimental results of orthogonal array design

Source of variance

Sum of square

Degree of freedom

Mean square

F-value F0.05

Static time 56.489 2 28.244 1844.259 19.247Temperature 5.074 2 2.537 165.666 19.247Modifier ratio 3.825 2 1.913 124.895 19.247Particle size 43.242 2 21.621 1411.783 19.247Pooled error 0.276 18 0.015Total variance 1473.515 27

Table 5　Comparison of validation data of OA and UA between solvent extraction (SE)-RPHPLC and on-line SFE-SFC method

Method CompoundYielda/mg g–1

Linearity/μg mL–1

Regression equation

R2 LOD/μg mL–1

LOQ/μg mL–1

Precision, % (n = 6)

Reproducibility, % (n = 6)

Accuracy, %

Low Medium High

SE-RPHPLC OA 2.87 20.0 – 180.0 y = 521469.4x 0.9999 4.1 12.5 0.77 0.85 — — —UA 0.68 20.0 – 180.0 y = 441218.3x 0.9999 4.1 12.5 0.54 0.95 — — —

SFE-SFC OA 8.67 1.0 – 500.0 y = 2488.3x – 4029.4 0.9999 0.5 1.0 0.89 2.07 99.0 103.6 99.6UA 2.12 1.0 – 500.0 y = 2039.0x – 2237.1 0.9999 0.5 1.0 0.85 2.07 100.4 104.2 96.0

a. SFE yield of OA and UA in dry plant.


Acknowledgements

This work was supported by Department of Analytical Instruments of Shimadzu Analytical Center (Beijing, China). We thank Shimadzu-GL Sciences (Shanghai, China) laboratory supplies Co., for providing the laboratory supplies in this research.

References

1. S. Fontanay, M. Grare, J. Mayer, C. Finance, and R. E. Duval, J. Ethnopharmacol., 2008, 120, 272.

2. L. Sun and Y. Hong, J. Chin. Pharm. Sci., 2000, 9, 6. 3. S. Yu, L. Zhang, M. Shan, Y. Qian, and A. Ding, Chin.

Tradit. Herb. Drugs, 2016, 47, 2465. 4. C. P. Commission, “Chinese Pharmacopoeia”, 2015, China

Medical Science Press, Beijing, China. 5. J. Liu, J. Ethnopharmacol., 1995, 49, 57. 6. J. H. Park, H. Y. Kwon, E. J. Sohn, K. A. Kim, B. Kim, S.

J. Jeong, J. H. Song, J. S. Koo, and S. H. Kim, Pharmacol. Rep., 2013, 65, 1366.

7. T. Akihisa, S. Kamo, T. Uchiyama, H. Akazawa, N. Banno,

Y. Taguchi, and K. Yasukawa, J. Nat. Med., 2006, 60, 331. 8. C. Zhou, K. Chen, C. Sun, Q. Chen, W. Zhang, and X. Li,

Biomed. Chromatogr., 2007, 21, 755. 9. M. Martelanc, I. Vovk, and B. Simonovska, J. Chromatogr.

A, 2009, 1216, 6662. 10. C. Chu, S. Zhang, S. Tong, X. Li, and J. Yan, J. Sep. Sci.,

2013, 36, 3958. 11. E. Lesellier, J. Sep. Sci., 2008, 31, 1238. 12. L. Wang and C. L. Weller, Trends Food Sci. Technol., 2006,

17, 300. 13. Z. Han, X. Wang, M. Xu, Y. Wang, L. Yang, and M. Han,

Chin. Herb. Med., 2016, 8, 314. 14. K. Zaghdoudi, X. Framboisier, C. Frochot, R. Vanderesse,

D. Barth, J. KalthoumCherif, F. Blanchard, and Y. Guiavarc’h, Food Chem., 2016, 208, 209.

15. M. Gallo, A. Formato, D. Ianniello, A. Andolfi, E. Conte, M. Ciaravolo, V. Varchetta, and D. Naviglio, J. Supercrit. Fluids, 2017, 119, 104.

16. Y. Yang and M. Wei, Ind. Crops Prod., 2016, 79, 7. 17. Z. Wang, H. Zhang, O. Liu, and B. Donovan, J. Chromatogr.

A, 2011, 1218, 2311. 18. E. Lesellier, E. Destandau, C. Grigoras, L. Fougere, and C.

Elfakir, J. Chromatogr. A, 2012, 1268, 157. 19. Q. Fu, Z. Li, C. Sun, H. Xin, Y. Ke, Y. Jin, and X. Liang, J.

Supercrit. Fluids, 2015, 104, 85. 20. I. Pfeifer, A. Murauer, and M. Ganzera, J. Pharm. Biomed.

Anal., 2016, 129, 246. 21. D. Okamato and Y. Hirata, Anal. Sci., 2006, 22, 1437. 22. K. Suto, Y. Ito, K. Sagara, and H. Itokawa, J. Chromatogr.

A, 1997, 786, 366. 23. A. D. Sanchez-Camargo, F. Parada-Alfonso, E. Ibanez, and

A. Cifuentes, J. Sep. Sci., 2017, 40, 213. 24. X. Qiao, R. An, Y. Huang, S. Ji, L. Li, Y. M. Tzeng, D. A.

Guo, and M. Ye, J. Chromatogr. A, 2014, 1358, 252. 25. R. Samimi, W. Z. Xu, Q. Alsharari, and P. A. Charpentier,

J. Supercrit. Fluids, 2014, 86, 115. 26. Y. Gao, B. Nagy, X. Liu, B. Simándi, and Q. Wang, J.

Supercrit. Fluids, 2009, 49, 345. 27. J. A. Egydio, Â. M. Moraes, and P. T. V. Rosa, J. Supercrit.

Fluids, 2010, 54, 154. 28. B. P. Nobre, A. F. Palavra, F. L. P. Pessoa, and R. L.

Mendes, Food Chem., 2009, 116, 680. 29. G. Sodeifian, S. A. Sajadian, and N. Saadati Ardestani, J.

Supercrit. Fluids, 2016, 116, 45.

Fig. 5　Separations of OA and UA by RP-HPLC (A) and on-line SFE-SFC (B). The conditions of RP-HPLC and the on-line SFE-SFC conditions are described in the relevant sections (the previous 8 min in on-line SFE-SFC is the SFE process).

Documents

Rapid Determination of Two Triterpenoid Acids in