Anal. Chem. 2000, 72, 3363-3365
Technical Notes
Low-Speed Rotary Countercurrent Chromatography Using a Convoluted Multilayer Helical Tube for Industrial Separation Qizhen Du,† Pingdong Wu,‡ and Yoichiro Ito*,§
Tea Research Institute, Chinese Academy of Agricultural Sciences, Hangzhou, Zhejiang 310008, China, Department of Chemical Engineering, Zhejiang University, Hangzhou, Zhejiang 310027, China, and Laboratory of Biophysical Chemistry, National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, Maryland 20892
An industrial-scale countercurrent chromatographic system was developed using a slowly rotating helical device which can be automated and left unattended during the operation. It uses a multilayer coiled column consisting of long convoluted Teflon tubing of 0.85 cm i.d., which provides sufficient retention of the stationary phase under a relatively high flow rate. The column is mounted on a seal-less continuous-flow rotary device which eliminates various complications such as leakage and clogging as often caused by the use of a conventional rotary seal. Using a multilayer coiled column with a 10-L capacity, 150 g of crude tea extract was successfully purified, yielding 40 g of epigallocatechin gallate of over 92.7% purity at a recovery rate of 82.6%. The present system can be scaled up for a kilogram-scale separation. Since the 1970s, the coil planet centrifuge has been utilized for performing countercurrent chromatography (CCC) to separate various natural products mainly in research laboratories.1-4 In this centrifugal CCC system, scaling up of the column capacity for industrial preparation is limited by its relatively high speed column rotation, which is required to retain the stationary phase in the column. In the past, a slowly rotating coil has been used for preparative separation of dinitrophenyl amino acids with a stable chloroform solvent system;5,6 however, the retention of the stationary phase would become insufficient in other solvent systems especially for low interfacial tension solvent systems due to a lack of centrifugal force. Since the peak resolution in CCC highly depends on the amount of stationary phase retained in the †
Chinese Academy of Agricultural Sciences. Zhejiang University. § National Institutes of Health. (1) Mandava, N. B., Ito, Y., Eds. Countercurrent Chromatography: Theory and Practice; Marcel Dekker: New York, 1988. (2) Conway, W. D. Countercurrent Chromatography: Apparatus, Theory and Application; VCH: New York, 1990. (3) Ito, Y., Conway, W. D., Eds. High-Speed Countercurrent Chromatography; Wiley-Interscience: New York, 1996. (4) Menet, J.-M., Thiebaut, D., Eds. Countercurrent Chromatography; Marcel Dekker: New York, 1999. (5) Ito, Y.; Bhatnagar, R. J. Chromatogr. 1981, 207, 171-180. (6) Ito, Y.; Bhatnagar, R. J. Liq. Chromatogr. 1984, 7 (2), 257-273. ‡
10.1021/ac991423q CCC: $19.00 Published on Web 05/18/2000
© 2000 American Chemical Society
Figure 1. Seal-free rotary device equipped with a large convoluted multilayer coil.
Figure 2. Three types of Teflon tubing used in the present study. Convoluted tubing has a spiral ridge (resembling a miniature vacuum cleaner duct) which improves retention of the stationary phase and facilitates the preparation of helical column due to its flexibility.
column, retention of the stationary phase becomes a key parameter to improve the partition efficiency of the CCC system for industrial application. The present paper introduces a new column geometry by utilizing convoluted tubing which enables higher retention of stationary phase as well as enhanced mixing of two phases to improve peak resolution in a slowly rotating helical column. Performance of the present system is compared with a similar column made of standard wall tubing, and the feasibility of industrial application is demonstrated with a 10-L capacity column on the successful separation of 150 g of crude tea extract. EXPERIMENTAL SECTION A seal-free flow-through rotary device used in the present system is schematically illustrated in Figure 1. It horizontally holds a cylindrical column holder that can be rotated around its axis at a desired low speed from 0 to 100 rpm. A set of gears and pulleys provides a rotation ratio of 2:1 between the column holder and Analytical Chemistry, Vol. 72, No. 14, July 15, 2000 3363
Figure 3. Peak resolution of progesterone (1) and 11-hydroxylprogesterone (2) with three types of single-layer coiled columns by the sealfree rotary device. Experimental conditions: sample, a mixture of 5 mg each of the above in 10 mL of the lower phase; solvent system, n-hexane/ methanol/water (6:5:3); mobile phase, lower aqueous phase; elution mode, head to tail; flow rate, 5 mL/min; column rotation, 50 rpm.
Figure 4. Preparative chromatogram of pueraria root extract obtained by the present method. Experimental conditions: sample, 30 g of pueraria root extract in 300 mL of solvent consisting of equal volume of each phase; solvent system, n-hexane/ethyl acetate/1-butanol/water/ acetic acid (1:1:2:6:0.2); mobile phase, lower phase; elution mode, head to tail; flow rate, 5 mL/min; column rotation, 21 rpm.
Table 1. Retention of Stationary Phase of Three Types of Coiled Columnsa % retention of stationary phase solvent system
stationary phase
n-hexane 6 MeOH 5 H2O 3
revolution flow rate small large (rpm) (mL/min) std conv conv
upper phase
50
5
35.0 45.0
72.4
n-hexane EtOAc -BuOH AcOH H2O
1 1 2 upper phase 0.2 6
22
5
20.4 25.0
43.5
CHCl3 MeOH H2O AcOH
5 3 4 1
90
5
16.4 30.0
48.3
lower phase
a Key: MeOH, methanol; EtOAc, ethyl acetate; AcOH, acetic acid; std, standard wall Teflon tubing, 5.5 mm i.d., 8.5 m long and 202 mL capacity; small conv, convoluted Teflon tubing, 5.7 mm average i.d., 8.0 m long and 204 mL capacity; large conv, convoluted Teflon tubing, 8.1 mm average i.d., 4.0 m long and 207 mL capacity.
the rotary frame which allows the flow tubes to rotate without twisting.5,6 Consequently, the system permits continuous elution through the rotating column without a conventional rotary device 3364 Analytical Chemistry, Vol. 72, No. 14, July 15, 2000
that may cause leakage and contamination. PTFE tubing (Zeus Industrial Products, Raritan, NJ) is directly wound around the holder hub forming either a single-layer or multilayer coil separation column. For preliminary studies, three different single-layer coils were mounted onto the holder hub of 17-cm-diameter, i.e., standard wall tubing of 5.5 mm i.d. × 8.5 m (202 mL capacity), small convoluted tubing of 5.7 mm i.d. × 8.0 m (204 mL capacity), and large convoluted tubing of 8.5 mm i.d. × 4 m (207 mL capacity) (Figure 2). A series of experiments was performed with the above three coiled columns to measure retention of the stationary phase and peak resolution of steroid mixture. The retention of the stationary phase was measured with three different two-phase solvent systems each under an optimum experimental condition as indicated in Table 1. The partition efficiency of these three columns were tested on separation of a mixture of progesterone and 11-hydroxyprogesterone, each 5 mg, dissolved in 10 mL of lower phase. The two-phase solvent system composed of hexane, methanol, and water at a volume ratio of 6:5:3 was used. The separation was performed with the lower aqueous phase as a mobile phase in a head-to-tail elution mode at a flow rate of 5 mL/min at 50 rpm of coil rotation. The effluent was collected in test tubes (10 mL fraction), and the absorbance of each fraction was manually measured at 254 nm with a spectrophotometer to draw chromatogram.
Figure 5. Preparative chromatogram of crude extract of tea leaves obtained by the present method. Experimental conditions: sample, 150 g of tea leaves extract dissolved in 1.2 L of solvent consisting of equal volumes of each phase; solvent system, n-hexane/ethyl acetate/1butanol/water/acetic acid (0.5:1:2:6:0.2); mobile phase, lower phase; elution mode, head to tail; flow rate, 5 mL/min; column rotation, 21 rpm.
To demonstrate the preparative capability of the present system, a 10-L-capacity column was prepared by winding 200-mlong, 8.5-mm-i.d. convoluted tubing onto a 9-cm-o.d. column holder hub making seven layers each consisting of 60 loops (Figure 1). In each separation, the column was first completely filled with the organic stationary phase followed by sample charge. Then the aqueous mobile phase was eluted through the column at 5 mL/min in a head-to-tail elution mode while the column was rotated at an optimum rate. The effluent was collected into test tubes, and each fraction was analyzed by either TLC or HPLC to construct the chromatogram. RESULTS AND DISCUSSION Using three sets of two-phase solvent systems, the retention of the stationary phase was measured in three types of singlelayer coils. The results obtained at an optimum flow rate of 5 mL/ min are shown in Table 1. The small convoluted helical column (5.7 mm average i.d.) produced higher retention than the standardwall column of similar size (5.5 mm i.d.). The results also revealed that the large convoluted tubing (8.5 mm average i.d) gave the highest retention among all columns although this may be partly due to a relatively lower linear flow rate compared with the other two columns. Chromatograms of a steroid mixture obtained from these columns are shown in Figure 3. The result clearly indicates that the small convoluted column produces better peak resolution than the standard wall column, while the large convoluted column yielded the best peak resolution despite its relatively short length (4 m). The overall results suggest that the present system may be further scaled up using larger inner diameter convoluted tubing without loss of partition efficiency. The preparative capability of the present method was demonstrated with a 8.5-mm-i.d convoluted multilayer coil with a 10-L (7) Du, Q.-Z.; Ke, C.-Q.; Ito, Y. J. Liq. Chromatogr. 1998, 21 (1-2), 203-208.
capacity. In the first experiment, 30 g of crude root extract of Pueraria lobata containing puerarin at 60% was separated using a two-phase solvent system composed of n-hexane/ethyl acetate/ 1-butanol/water/acetic acid (1:1:2:6:0.2, v/v) at a flow rate of 5 mL/min at 21 rpm. The total elution time was 43 h, and 38% of the stationary phase was retained when measured at the end of the run. Figure 4 shows the chromatogram constructed from TLC analysis of each fraction. Two major components, puerarin and 3′-methoxypuerarin, were resolved with a minimum overlapping zone. This separation yielded 16.7 g of puerarin at 91.2% purity, which corresponds to a recovery rate of 84.6%. In the second experiment, 150 g of dried extract of tea leaves containing epigallocatechin gallate (EGCG)7 at 30% was separated with a two-phase solvent system composed of n-hexane/ethyl acetate/1-butanol/water/acetic acid (0.5:1:2:6:0.2, v/v) at a flow rate of 5 mL/min at 21 rpm. The separation required 72 h, and retention of the stationary phase was 33%. Each fraction was analyzed by HPLC to construct the chromatogram (Figure 5) where the EGCG peak was well resolved from the rest. In this separation, 40.05 g of EGCG was obtained at 92.7% purity, which corresponds to a 82.6% recovery rate. CONCLUSION The results of our studies demonstrated a preparative capability of the present system on separation of 150 g of a crude tea extract. The system can be readily scaled up for the industrial separation of kilogram sample size using a longer and/or larger inner diameter convoluted helical column. Although the separation requires days of continuous elution, the simplicity of the system provides an ease of automation while slow column rotation permits the system to be left unattended during the separation. Received for review December 13, 1999. Accepted April 13, 2000. AC991423Q
Analytical Chemistry, Vol. 72, No. 14, July 15, 2000
3365
