an interesting two-phase solvent system and its use in preparative isolation of aconitines from...

1304 J. Sep. Sci. 2013, 36, 1304–1310

Quan-Bin Han1

Wai-Lun Tang1

Cai-Xia Dong1

Hong-Xi Xu2

Zhi-Hong Jiang3∗

1School of Chinese Medicine,Hong Kong Baptist University,Hong Kong, China

2School of Pharmacy, ShanghaiUniversity of TraditionalChinese Medicine, Shanghai,China

3State Key Lab of QualityResearch in Chinese Medicine,Macau Institute for AppliedResearch in Medicine andHealth, Macau University ofScience and Technology, Taipa,Macau, China

Received November 23, 2012Revised December 15, 2012Accepted January 3, 2013

Research Article

An interesting two-phase solvent systemand its use in preparative isolation ofaconitines from aconite roots bycounter-current chromatography

Two-phase solvent system plays crucial role in successful separation of organic compoundsusing counter-current chromatography (CCC). An interesting two-phase solvent system,composed of chloroform/ethyl acetate/methanol/water, is reported here, in which bothphases contain sufficient organic solvents to balance their dissolving capacities. Adjustingthe solvent system to get satisfactory partition coefficients (K values) for target compoundsbecomes relatively simple. This solvent system succeeded in sample preparation of aconitine(8.07 mg, 93.69%), hypaconitine (7.74 mg, 93.17%), mesaconitine (1.95 mg, 94.52%) fromraw aconite roots (102.24 mg, crude extract), benzoylmesaconine (34.79 mg, 98.67%) fromprocessed aconite roots (400.01 mg, crude extract), and yunaconitine (253.59 mg, 98.65%)from a crude extract of Aconitum forrestii (326.69 mg, crude extract).

Keywords: Aconitine / Counter-current chromatography / Two-phase solventsystemDOI 10.1002/jssc.201201079

1 Introduction

High-speed counter-current chromatography (HSCCC) hasbeen popularly used to prepare high-purity natural com-pounds [1, 2]. HSCCC is a continuous liquid–liquid sol-vent partition technique in which the target compoundsare competitively distributed between the two-phase solventsdue to their different partition coefficients (K values). Thetwo-phase solvent system plays a crucial role in HSCCCseparation.

Many different two-phase solvent systems have beendeveloped in order to separate the highly diverse naturalcompounds [1–3]. Aqueous two-phase solvent systems havebeen designed for macromolecules [4–9]. Recently, a newtwo-phase solvent system composed of 1-butanol–ethanol–saturated ammonium sulfate solution–water was report-edthat can be used to separate extremely polar compounds[10]. In the separation of small molecules, two-phase sol-vent systems composed of organic solvents and water arecommonly used, e.g. hexane/ethyl acetate/methanol/water[3]. The lower phase is always aqueous. There is an excep-tion, namely chloroform/methanol/water, in which the lowerphase is mainly the organic solvent chloroform that is denserthan water.

Optimization of the two-phase solvent system for thetarget compound(s) is the most important and also time-

Correspondence: Professor Quan-Bin Han, School of ChineseMedicine, Hong Kong Baptist University, Kowloon Town, HongKong SAR, ChinaFax: +852-3411-2461E-mail: [email protected]

consuming step in CCC separation. The selected solvent sys-tem should: (i) not make the target compound(s) decomposeor denature; (ii) make the target compound(s) soluble; (iii)yield satisfactory retention of the stationary phase in the col-umn; and (iv) generate a suitable partition coefficient for thetarget compound [1].

This paper reports an interesting two-phase solventsystem, chloroform/ethyl acetate/methanol/water, in whichboth phases can contain sufficient organic solvents to balancetheir dissolving capacities, respectively. The integrity of thetwo phases is maintained not only by the water-oil differencebut also by differences in density. It succeeded in prepara-tion of conitine, hypaconitine, and mesaconitine from rawaconite roots (Fig. 1) [11], benzoylmesaconine from the pro-cessed aconite roots [12,13], and yunaconitine from Aconitumforrestii [14]. By contrast, the commonly used system of hex-ane/ethyl acetate/methanol/water failed in these separations.

2 Experimental

2.1 Materials and reagents

Hexane, methanol, ethyl acetate, and ACN were of HPLCgrade and were bought from Tedia, USA. Chloroform was ofanalytical grade and was purchased from VWR International,EC. Ammonia solution (35%), 37% fume hydrochloric acid,and 70% perchloric were purchased from BDH Laboratory

∗Additional Correspondence: Professor Zhi-Hong Jiang,E-mail: [email protected]

C© 2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.jss-journal.com

J. Sep. Sci. 2013, 36, 1304–1310 Sample Preparation 1305

Figure 1. The chemical struc-tures of aconitine, hypaconitine,mesaconitine, benzoylmesaco-nine, and yunaconitine.

Supplies, England. Pure water was prepared by MILLI-Q SPwater system (Nihon Millipore Kogyo K.K., Japan) and wasdistilled twice before use.

The aconite roots (Aconitum carmichaeli, CW; Aconitumforrestii, HCW; and processed aconite roots, ZCW) were pur-chased in Qingping Market, Guangzhou City, China, andwere authenticated by Dr. Chunfeng Qiao. The crude extractof ZCW was prepared as follows: 500 g of the ZCW samplewas ground into powder, and then percolated with 2 L 0.3 Mhydrochloric acid. The acidic extraction was basified with 35%ammonium solution to pH 10 and then extracted with chlo-roform three times (3 L each time). The chloroform layer wascollected and the solvent was removed under reduced pres-sure, resulting in a yellowish-brown and amorphous powder(0.5 g). The crude extracts of CW and HCW were prepared inthe same way [11, 12]. The reference standards of aconitine,mesaconitine, hypaconitine, yunaconitine, and benzoylme-saconine were provided by the National Institutes for Foodand Drug Control (NIFDC, China).

2.2 Apparatus

The HSCCC instrument employed in the present study was aTBE-300A high-speed counter-current chromatograph (TautoBiotechnique Company, Shanghai, China) with three multi-layer coil separation columns connected in series (id of thetubing = 1.6 mm, total volume = 300 mL). The revolutionradius was 5 cm, and the β values of the multilayer coil variedfrom 0.5 at the internal terminal to 0.8 at the external termi-nal. The revolution speed of the apparatus can be regulatedwith a speed controller in the range between 0 and 1000 rpm.A HX 1050 constant-temperature circulating implement (Bei-jing Boyikang Lab Instrument Co., Beijing, China) was usedto control the separation temperature. As a chromatographiccolumn, the HSCCC system was equipped with a preparativeAgilent 1100 HPLC Series (G1361A Prep Pump and G1365BMWD), and a 3725i-038 injector (Rheodyne, USA).

An analytical Agilent 1100 HPLC series was used to exam-ine the chemical profile of crude extracts, reference standards,and obtained isolates. High-resolution ESIMS was performedon a Waters Q-TOF Premier (Micromass MS Technologies,Manchester, UK) mass spectrometer in [M+H]+ mode. Thenebulization gas was set to 400 L/h at a temperature of 350�C,and the source temperature was set to 102�C. The capillaryvoltage and cone voltage were set to 2900 and 35 V, respec-tively. The Q-TOF Premier spectrometer’s acquisition ratewas set to 0.2 s, with a 0.01 s interscan delay. Argon was em-ployed as the collision gas at a pressure of 5.3 × 10–5 Torr. Themolecular masses of the molecular ions and of the productions were accurately determined with reference compoundLeucine–enkephalin in the LockSpray mode (m/z 556.2771)at a concentration of 50 pg/�L and an infusion flow rate of 10�L/min. The CID energies were set at 45 eV. The pH valueswere obtained on a Sartorius PB-20 pH meter.

2.3 Measurement of partition coefficient

First, HPLC examination of the reference standardsand these three crude extracts was conducted on an Ag-ilent 1100 series. A ZORBAX R© SB-C18 column (5 �m,4.6 × 250 mm) was used for the examination of CW, aconi-tine, hypaconitine, and mesaconitine. The mobile phase ofmethanol/ACN/0.2% acetic acid (containing 0.5% triethy-lamine) (20:40:40) was set at a flow rate of 1.0 mL/min. UVdetection was set at 240 nm. For HCW and ZCW, the HPLCexamination was conducted on a Sunfire C18 column (5 �m,4.6 × 250 mm). For HCW, the mobile phase was ACN/0.1%perchloric acid (35:65) at a flow rate of 1.0 mL/min. UV de-tection was set at 267 nm. For ZCW, the mobile phase wascomposed of 0.5% v/v ammonium solution and ACN in a gra-dient program as follows: 0–20 min: 35% ACN; 20–50 min:35% ACN → 60% ACN; 50–60 min: 60% ACN → 95% ACN,at a flow rate of 1.0 mL/min. UV detection was set at 230 nm.


1306 Q.-B. Han et al. J. Sep. Sci. 2013, 36, 1304–1310

Table 1. The partition coefficients (K value) of aconitines in the two-phase solvent systems of chloroform/ethyl acetate/methanol/water(0.3% HCl)

Volume ratio (v/v/v/v) K value

Mesaconitine Aconitine Hypaconitine Benzoylmesaconine Yunaconitine

4.00:0.00:1.5:2 0.173 0.339 0.149 1.456 02.00:2.00:1.5:2 – – – 6.632 5.6673.00:0.50:1.5:2 0.351 0.856 0.141 2.739 0.3763.00:0.75:1.5:2 0.673 1.696 0.261 – –3.00:1.00:1.5:2 0.975 2.615 0.355 4.411 0.9722.75:1.00:1.5:2 1.183 3.359 0.493 – –2.25:1.00:1.5:2 1.811 4.856 0.692 – –2.00:1.00:1.5:2 2.005 5.374 0.837 – –1.75:1.00:1.5:2 2.281 6.473 0.871 – –1.50:1.00:1.5:2 2.462 6.744 1.024 – –

“-”, not measured.

Figure 2. The influences of chlo-roform and ethyl acetate in thetested two-phase solvent sys-tems on the generated K valuesof aconitine analogues.

For all these three crude extracts, approximately 2 mg ofeach sample was weighed in a 10 mL test tube to which4.0 mL of each phase of the equilibrated two-phase sol-vent system was added. The tube was shaken vigorously for2 min to equilibrate the sample thoroughly. Then the upperand lower phases were analyzed by HPLC. For each targetcomponent, the partition coefficient (K) was expressed as thepeak area of target components in the upper phase dividedby that in the lower phase.

2.4 Selection of solvent ratios

As listed in Table 1, chloroform/ethyl acetate/methanol/water, and chloroform/methanol/water, were investigated ina varied composition to generate an expected K value (around1) for individual aconitine analogues.

The two-phase solvent system was prepared by addingthe solvents to a separation funnel according to the selected

volume ratios and then shaking repeatedly at room tem-perature to equilibrate them. The upper and lower phaseswere separated and degassed by sonication for 30 minbefore use.

2.5 CCC separation procedure

The separation procedure was carried out as follows: first,the coil column was entirely filled with the upper phase ofthe solvent system using the HPLC pump at a flow rate of20 mL/min. Then the apparatus was rotated at 800 rpm,while the lower phase was pumped into the column at a flowrate of 1.5 mL/min. After the mobile phase front emergedand hydrodynamic equilibrium was established in the col-umn, sample solution was injected into the column throughthe injector. The separation temperature was controlled at20�C. The effluent from the outlet of the column was contin-uously monitored by an Agilent 1100 HPLC UV-Vis detector



Table 2. The individual solvent composition (%), pH values of the upper phase and lower phase and their volume ratio in the tested solventsystems of chloroform/ethyl acetate/methanol/water (0.3% HCl)

Volume ratio (v/v/v/v) Upper phase Lower phase

U/LCHCl3 EtOAc MeOH Water pH CHCl3 EtOAc MeOH Water pH (v/v)

4.00:0.00:1.5:2 4.7 0.0 32.2 63.0 1.68 88.7 0.0 11.4 1.0 1.64 35:402.00:2.00:1.5:2 3.6 8.4 32.3 55.5 1.67 46.9 42.2 9.5 1.9 1.85 33:42

3.00:0.50:1.5:2 4.8 1.6 33.5 60.0 1.65 76.9 11.9 10.9 1.2 1.85 35:353.00:0.75:1.5:2 4.6 2.4 33.2 59.8 1.67 72.3 16.8 10.5 1.2 1.84 35:383.00:1.00:1.5:2 4.4 3.1 32.8 59.6 1.67 68.3 21.2 10.1 1.3 1.84 32:43

2.75:1.00:1.5:2 4.4 3.5 33.2 58.8 1.67 66.8 22.5 10.2 1.3 1.85 31:422.25:1.00:1.5:2 4.6 4.3 33.8 57.2 1.67 63.4 25.5 10.4 1.5 1.84 33:362.00:1.00:1.5:2 4.7 4.9 34.1 56.3 1.65 61.4 27.4 10.4 1.5 1.84 33:321.75:1.00:1.5:2 4.7 5.6 34.4 55.3 1.65 59.0 29.6 10.5 1.6 1.85 32:301.50:1.00:1.5:2 4.7 6.4 34.6 54.2 1.67 56.3 32.1 10.5 1.7 1.85 34:26

and ChemStation. Fractions were collected every 10 min andevaporated under reduced pressure. The residue was dis-solved in methanol for subsequent HPLC analysis.

2.6 HPLC analysis and identification of CCC fractions

CCC fractions were examined using the HPLC methods de-scribed previously. Their purities were calculated using areanormalization. Their structures were identified by compari-son of the retention times with reference standards and con-firmed by high-resolution ESIMS analysis.

3 Results and discussion

3.1 Selection of two-phase solvent system

The quaternary two-phase solvent system of hexane/ethylacetate/methanol/water is the most popularly used in theseparation of various natural products [1, 2]. An acid is of-ten added with the solvents in order to shorten the peaktailing in separation of alkaloids. Therefore, hexane/ethylacetate/methanol/0.3% HCl in four different compositions(5:5:5:5, 4:5:4:5, 3:5:3:5, and 2:5:2:5) was used here first, withthe result that the alkaloids were all found in the lower phaseand the K values were zero. This is unsatisfactory becausethe alkaloids were mainly kept in acidic water of the lowerphase. Another system, namely, chloroform/methanol/0.3%HCl (4:1.5:2) was then tested. The resulting K values of threeaconitine analogues were still very small (Table 1), althoughthe upper phase had already changed to aqueous. Chloroformseemed to be more attractive to these alkaloids.

These two solvent systems showed big difference in dis-solving aconitines. Combination of them might generate bet-ter results. Therefore, hexane was replaced with chloroform inorder to get a balance between the organic solvent and acidicwater. Here, the two phases formed not only because of thewater-oil difference but also due to the difference in density.

Both phases contained organic solvents, but the upper phase,containing ethyl acetate and methanol, was much less densethan the lower phase, comprised mainly of chloroform.

As shown in Table 1, when the ratio of chloro-form/methanol/water was fixed at 3:1.5:2, the K values ofthree aconitine analogues increased as the ratio of ethyl ac-etate increased from 0.5 to 1.0; at the same time, when theratio of ethyl acetate/methanol/water was fixed at 1:1.5:2, theK values decreased along with the increase of chloroformfrom 1.0 to 3.0. These results showed that chloroform andethyl acetate had opposite effects on the K values of thesealkaloids. It will be easier to adjust the solvent system and getsatisfactory K values, as illustrated in Fig. 2.

The solvent system was adjusted to get satisfactory K val-ues for mesaconitine, aconitine, hypaconitine, benzoylme-saconine, and yunaconitine. The solvent ratio of 2.75:1:1.5:2was found to be optimal for the separation of mesaconitine,aconitine, and hypaconitine; while 3:1:1.5:2 was optimal foryunaconitine. Benzoylmesaconine required a more radicaladjustment of the system. Its K value stayed much higherthan 2 when the K value of other alkaloids approached 0. Inorder to lower its K value, ethyl acetate was removed from thesystem and the ratio of chloroform was increased to 4; thisresulted in a satisfactory K value of 1.456. All tested systemsgenerated a stable volume ratio between two phases around1:1.

3.2 Solvent composition in the solvent systems

The relative ratio of chloroform to ethyl acetate in the solventsystems is directly responsible for the K value variation. Ac-cording to the calculated content (%) of chloroform and ethylacetate using a model of solvent selection software (VersionI, Tauto Biotech), as demonstrated in Table 2, it was foundthat chloroform’s content increased in the lower phases anddecreased in the upper phases when its ratio to ethyl acetatewas increased from 1.75 to 3.0; at the same time, water’s con-tent decreased in the lower phases and increased in the upper



Figure 3. Application of thetwo-phase solvent system inseparation of aconitine and itsanalogues from the crude ex-tract of aconite roots. (A) Chem-ical profile of the crude aconiteextract by HPLC; (B) HSCCCchromatogram of the crudeaconite extract; (C) HSCCCfraction I’s (210–250 min) HPLCchromatogram (C1) and itsHRMS spectra (C2); (D) HSCCCfraction II’s (350–390 min) HPLCchromatogram (D1) and itsHRMS spectra (D2); (E) HSCCCfraction III’s (540–570 min) HPLCchromatogram (E1) and itsHRMS spectra (E2). The HPLCexamination was conductedon an Agilent 1100 seriesequipped with a ZORBAX R©

SB-C18 column (5 �m, 4.6 ×250 mm). The mobile phase wasmethanol/ACN/0.2% acetic acid(containing 0.5% triethylamine)(20:40:40) at a flow rate of1.0 mL/min. UV detection wasset at 240 nm.

phases. Furthermore, the organic solvents’ contents in lowerphase are more sensitive to the ratio of chloroform/ethyl ac-etate; the content of chloroform in upper phase changed littlewhen the ratio was increased from 1.75 to 3.0.

3.3 Effects caused by the pH values of the solvent

containing acid

It was presumed that pH values might affect the K valuessomewhat, since all the solvents contained acid and the iso-lates were alkaloids. Therefore, the pH values of the separatetwo phases of all the tested systems were determined usinga pHmeter. As shown in Table 2, the upper phase possesseda stable pH value of around 1.66. Similarly, the lower phaseshowed a stable pH value of 1.84. This similarity could beexplained by the calculated composition of all tested solventsystems (as listed in Table 1) in which the ratios of acid/water

were stable in both upper phase (around 50%) and lowerphase (around 1.5%). The above-mentioned data indicate thatthe variations in K values were not related to pH.

3.4 CCC separation and identification of CCC

fractions

As shown in Fig. 3A, the crude extract of CW mainly containsthree alkaloids, which is consistent with published reports[13–15]. A total of 102.24 mg of CW crude extract was sepa-rated using the two-phase solvent system of chloroform/ethylacetate/methanol/0.3%HCl 2.75:1:1.5:2 in a 300-mL CCCcolumn. As a result, these three alkaloids, bearing significantK value differences (1.183, 3.359, and 0.493), were elutedat 210–250 min, 350–380 min, and 540–570 min, respec-tively. These three fractions were identified by HPLC andHigh-Resolution MS analysis as, respectively, hypaconitine



Figure 4. Application of thetwo-phase solvent system inseparation of yunaconitinefrom the crude extract ofAconitum forresti. (A) Chemicalprofile of the crude aconiteextract by HPLC; (B) HSCCCchromatogram of the crudeaconite extract; (C): HPLCchromatogram of the obtainedHSCCC fraction (120–160 min)and its HRMS spectra; (D)The HPLC examination wasconducted on an Agilent1100 series equipped with aSunfire C18 column (5 �m,4.6 × 250 mm). The mobilephase was ACN/0.1% perchloricacid (35:65) at a flow rate of1.0 mL/min. UV detection wasset at 267 nm.

Figure 5. Application of thetwo-phase solvent system inseparation of benzoylmesaco-nine from the crude extractof processed aconite roots. (A)Chemical profile of the crudeaconite extract by HPLC; (B)HSCCC chromatogram of thecrude aconite extract; (C) HPLCchromatogram of the obtainedHSCCC fraction (330–420 min)and its HRMS spectra; (D) TheHPLC examination was con-ducted on an Agilent 1100 seriesequipped with a Sunfire C18 col-umn (5 �m, 4.6 × 250 mm). Themobile phase was composed of0.5% v/v ammonium solutionand ACN in gradient program:0–20 min: 35% ACN; 20–50 min:35% ACN → 60% ACN; 50–60 min: 60% ACN → 95% ACN,at a flow rate of 1.0 mL/min. UVdetection was set at 230 nm.

(7.74 mg, purity 93.17%), mesaconitine (1.95 mg, 94.52%),and aconitine (8.07 mg, 93.69%) (Fig. 3) [13–15]. Because theK value of aconitine is much larger than 1, the resulted sepa-ration time is a little long. Several operations, such as gradientelution with the system of 3:0.5:1.5:2, and a higher flow rate,or a larger sample load could eliminate this weakness.

Similarly, chloroform/methanol/0.3%HCl 3:1:1.5:2yielded yunaconitine (253.59 mg, 98.65%) from 326.69 mg

of HCW crude extract (Fig. 4), and chloroform/methanol/0.3%HCl 4:1.5:2 yielded benzoylmesaconine(34.79 mg, 98.67%) from 400.01 mg of ZCW crude extract(Fig. 5) [13–15]. The identification of the CCC fractions wasconfirmed by comparison with reference compounds interms of HPLC retention time and the HRMS data.

Compared to the classical two-phase solvent system ofhexane/ethyl acetate/methanol/water, the system includes



chloroform whose toxicity is often a concern. This might beminimized through a tail-to-head separation in which thelower phase is used as the stationary phase.

4 Concluding remarks

In summary, an interesting two-phase solvent system,namely chloroform/ethyl acetate/methanol/water, is estab-lished here in which both phases contain sufficient organicsolvents to balance their dissolving capacities. This solventsystem succeeded in sample preparation of five aconitineanalogues directly from different aconite crude extracts. Theinfluences of solvent composition on the target compound’sK value are also investigated. It is easy to adjust the solventratio and get satisfactory K values.

The authors have declared no conflict of interest.

5 References

[1] Ito, Y., J. Chromatogr. A 2005, 1065, 145–168.

[2] Marston, A., Hostettmann, K., J. Chromatogr. A 2006,1112, 181–194.

[3] Cao, X. L., High-speed Counter-current Chromatogra-phy and Its Application, Chemical Industry Press, Beijing2005, p. 390.

[4] Shen, C. W., Yu, T., J. Chromatogr. A 2007, 1151,164–168.

[5] Shibusawa, Y., Takeuchi, N., Sugawara, K., Yanagida,A., Shindo, H., Ito, Y., J. Chromatogr. B 2006, 844,217–222.

[6] Chao, Z. M., Shibusawa, Y., Shindo, H., Ito, Y., J. Liq.Chromatogr. Relat. Technol. 2003, 26, 1895–1903.

[7] Sun, X. J., Tang, B., Xu, Y. M., Chin. Patent 2008 CN101323648.

[8] Song, G. L., Du, Q. Z., J. Chromatogr. A 2010, 1217,5930–5934.

[9] Jiang, Z. G., Du, Q. Z., Sheng, L. Y., Chin. J. Anal. Chem.2009, 37, 412–416.

[10] Zeng, Y., Liu, G., Ma, Y., Chen, X., Ito, Y., J. Chromatogr.A 2011, 1218, 8715–8717.

[11] China Pharmacopoeia Commission, China Pharma-copoeia 2010, Vol. 1, China Medical Science and Tech-nology Press, Beijing 2010, p. 36.

[12] China Pharmacopoeia Commission, China Pharma-copoeia 2010, Vol. 1, China Medical Science and Tech-nology Press, Beijing 2010, p. 37.

[13] Xiao, P. G., Wang, F. P., Gao, F., Yan, L. P., Chen, D. L., Liu,Y., Acta Phytotax. Sin. 2006, 44, 1–46.

[14] Xie, Y., Zhou, H., Wong, Y. F., Liu, Z., Xu, H., Jiang, Z., Liu,L., Chin. Med. 2008, 3, 6.

[15] Yue, H., Pi, Z., Song, F., Liu, Z., Cai, Z., Liu, S., Talanta2009, 77, 1800–1807.


an interesting two-phase solvent system and its use in preparative isolation of aconitines from...

Documents