Dispersed Mobile-Phase Countercurrent Chromatography - MDPI

separations Article

Dispersed Mobile-Phase Countercurrent Chromatography Timothy Yiu-Cheong Ho and Hong Xue * Division of Life Science and Applied Genomics Center, Hong Kong University of Science and Technology, Clear Water Bay, Hong Kong, China; [email protected] * Correspondence: [email protected]; Tel.: +852-2358-8707 Academic Editor: Alain Berthod Received: 27 July 2016; Accepted: 11 October 2016; Published: 1 November 2016

Abstract: Countercurrent distribution based on liquid–liquid partition is a powerful separation method with minimal incurrence of loss of solutes, but its industrial application has been limited by cumbersome shifting of immiscible solvents. Although centrifugation has been employed to facilitate equilibration between phases, process scaling-up remains difficult. In this study, a dispersed mobile-phase countercurrent chromatography (DMCC) method has been developed to adapt the countercurrent distribution principle to a continuous column chromatography format. Continuous solute exchange between two immiscible phases within a series of separation columns is achieved by mechanical dispersion of an influx of mobile phase into an upward stream of small droplets travelling through the columns filled with stationary phase. The diameter, length, and number of columns, and the number of stationary phases employed in the different columns can be varied to match the requisite scale and resolution of operation. Illustrations of DMCC were provided by examples of solute separations where the fractionated solutes could be collected either from the eluate of the series of columns, or from drainage of the stationary phases in the individual columns at the end of a chromatographic run. Keywords: liquid–liquid partition; countercurrent chromatography; DMCC

1. Introduction Countercurrent distribution provides high-resolution separation for various classes of compounds [1–3], including the breakthrough purification of tRNA for sequencing [4,5]. While its avoidance of irreversible adsorptive losses represents an important advantage, its requirement of shifting and reequilibration of upper and lower phases from multiple tubes is mechanically complicated. In droplet countercurrent chromatography (DCCC), the passage of single droplets of mobile phase through a column of stationary phase of comparable diameter facilitates solute distribution [6], but the small diameter of the column renders scaling-up difficult. A range of modifications incorporating centrifugation, such as centrifugal partition chromatography (CPC) and high speed countercurrent chromatography (HSCCC) can speed up separation [3,7–17], but the use of centrifugation increases the cost of equipment, especially on an industrial scale. In controlled-cycle pulsed liquid-liquid chromatography (CPLC), mixing of upper and lower phases is conducted in columns segmented into a cascade of chambers by horizontal perforated plates, yet equilibration of solutes between the phases still has to be achieved on a discontinuous basis with intervening pauses for phase separation in each transfer cycle [18–20]. In order to combine the advantages of countercurrent distribution with those of column chromatography, in this study, we have developed a dispersed mobile-phase countercurrent chromatography (DMCC) method that enables continuous chromatography based on the dispersal of mobile phase into a stream of fine droplets travelling through a column of stationary phase (Figure 1). Dispersion of the mobile phase can be brought about by devices such as magnetic stirring or ultrasound.

Separations 2016, 3, 32; doi:10.3390/separations3040032

www.mdpi.com/journal/separations

Separations 2016, 3, 32 Separations 2016, 3, 4

2 of 8 2 of 8

1). Dispersion of the mobile phase can be brought about relative by devices such as magnetic stirring or Since DMCC employs columns of much larger diameters to the dispersed droplets, it can ultrasound. Sinceup DMCC employs columns of industrial much larger diameters to the dispersed be readily scaled to provide preparative and separations. Asrelative illustrated in the present droplets, it can be readily scaled up to provide preparative and industrial separations. As illustrated study, the fractionated solutes can be collected from different fractions of eluate emerging from a series in the presentcolumns, study, the fractionated solutes collectedphases from in different fractionscolumns of eluate of separation or from the solutes left incan thebe stationary the individual at emerging a series of separation columns,stationary or from the solutes left in thecolumns, stationaryand phases in the the end offrom a run. The placement of different phases in different the effects individual columns at theflow end rate of a of run. The placement different phases in different of column dead-volume, mobile phase, andofthe numberstationary of columns employed were columns, and the effects of column dead-volume, flow rate of mobile phase, and the number of also investigated. columns employed were also investigated.

Figure 1. Setup of Figure 1. Setup of of dispersed dispersed mobile-phase mobile-phase countercurrent countercurrent chromatography chromatography (DMCC): (DMCC): A A stream stream of dispersed lighter mobile-phase (yellow) is shown ascending through a series of separation columns, dispersed lighter mobile-phase (yellow) is shown ascending through a series of separation columns, each denser stationary stationary phase phase (green). (green). each containing containing aa denser

2. Materials and Methods 2. Materials and Methods 2.1. 2.1. Reagent Reagent and and Materials Materials Distilled employed throughout. throughout. Equilibrated Distilled water water was was employed Equilibrated phenol phenol (pH (pH 8.0, 8.0, Ultrapure Ultrapure MB MB grade) grade) was obtained from USB Corp. (Cleveland, OH, USA), o-cresol from Riedel-de Haën (Seelze, was obtained from USB Corp. (Cleveland, OH, USA), o-cresol from Riedel-de Haën (Seelze, Germany), Germany), tripotassium from Nacalai Tesque (Kyoto, Japan), benzylbenzoic alcohol, acid, benzoic acid, tripotassium phosphate phosphate from Nacalai Tesque (Kyoto, Japan), benzyl alcohol, sodium sodium hydroxide, and sodium hydrogencarbonate from Sigma-Aldrich (St. Louis, MO, USA), nhydroxide, and sodium hydrogencarbonate from Sigma-Aldrich (St. Louis, MO, USA), n-butanol from butanol from VWR Chemical (Radnor, PA, USA), baicalein from Indofine Chemical Co. Inc. VWR Chemical (Radnor, PA, USA), baicalein from Indofine Chemical Co., Inc. (Hillsborough, CA, (Hillsborough, CA, USA), and wogonin from Wako Pure Chemical Industries (Osaka, Japan). USA), and wogonin from Wako Pure Chemical Industries (Osaka, Japan). 2.2. Selection of Two-Phase System A quantity quantity of ofeach eachofofthe thekey key compounds to separated be separated by DMCC dissolved in a twocompounds to be by DMCC was was dissolved in a two-phase phase such that their concentrations inphases the two phases could be determined accurately by systemsystem such that their concentrations in the two could be determined accurately by absorbance absorbance measurement. mixing and phase-settling, the concentration of theincompound in the measurement. After mixingAfter and phase-settling, the concentration of the compound the upper phase upper phase (A1) and phase in the (A2) lowerwere phase (A2) wereby determined byatabsorbance 254the nmdistribution to yield the (A1) and in the lower determined absorbance 254 nm to at yield distribution D = A1/A2. selected upperphases and lower phases KD values within coefficient KDcoefficient = A1/A2.KThe selectedThe upper and lower provided KDprovided values within the range of the range of 0.5–12 (Table 1) for the differentto compounds to be separated. 0.5–12 (Table 1) for the different compounds be separated.

Separations 2016, 3, 32

3 of 8

Table 1. KD of each solute in different solvent systems. Solvent System

Sample

KD

n-butanol–1% NaOH (1:1, v/v)

cresol benzyl alcohol

1.06 4.78

n-butanol–0.1 M NaHCO3 (1:1, v/v)

benzyl alcohol phenol benzoic acid

4.77 4.68 0.51

n-butanol–0.1 M NaOH (1:1, v/v)

benzyl alcohol benzoic acid benzyl alcohol

4.77 0.49 0.50

n-butanol–0.1 M K3 PO4 (1:1, v/v)

wogonin baicalein

12.1 2.56

2.3. DMCC Setup The DMCC system consisted of ten Econo-Column® chromatography columns (except where specified otherwise), each with 25 mm i.d., length of 20 cm, and volume of 117 mL (BioRad, Hercules, CA, USA). They were connected in series with 35 cm lengths of MasterFlex silicone tubing with 1.6 mm i.d. (except in the high dead-volume runs in where 3.1 mm i.d. tubing and 46 cm tubing lengths were employed instead). All of the columns were filled with approximately 116 mL of the denser of a selected pair of immiscible solvents serving as stationary phase, and all connecting tubings were subsequently filled with the lighter solvent serving as mobile phase by pumping the mobile phase through all of the columns of stationary phase. Prior to the start of a DMCC run, the lighter solvent—serving as the mobile phase—was pumped into the columns at 4 mL/min (except where specified otherwise) through an influx port at the bottom of column-1 by an easy-load™ MasterFlex® model 7518-00 (Cole-Parmer, Vernon Hills, IL, USA). Upon entry into column-1, as well as each successive column, this inflow of mobile phase was broken up by a magnetic stir bar rotating at 350 rpm, driven by a Thermolyne Mirak stirrer from Sigma-Aldrich (St. Louis, MO, USA) into a stream of dispersed droplets ascending through the column of stationary phase (Figure 1). Insofar that a stirring device can be positioned more readily at the bottom of the column compared to the top of the column, transit of mobile phase through the stationary phase in DMCC is more easily implemented in the ascending mode than in the descending mode. An hour later, the inflow of mobile phase was halted, and a sample solution containing a mixture of solutes was injected into the bottom of column-1, followed by the resumption of inflow of mobile-phase. The absorbance of the eluate emerging from the last column was monitored at 254 nm using a Biologic™ LP (BioRad, Hercules, CA, USA). At the end of the DMCC run, the stationary phases remaining in all the different columns were individually drained and analyzed at 254 nm with a UV-1201 spectrophotometer (Shimadzu, Kyoto, Japan). 3. Rusults 3.1. Separation of o-Cresol and Benzyl Alcohol Using an n-Butanol–NaOH(aq) Two-Solvent System In this DMCC run, water-saturated n-butanol was employed as the mobile phase. An n-butanol–1% NaOH (1:1, v/v) mixture was shaken and allowed to settle, and its lower phase was employed as the stationary phase in the columns. Following the passage of mobile phase through the columns to pre-equilibrate the phases, a 4 mL sample mixture consisting of 2 g of o-cresol and 2 g of benzyl alcohol was injected into column-1 at the start of DMCC run, followed by resumption of influx of mobile phase. Monitoring of the eluate from the last column showed that passage of 320 mL of mobile phase through the column system was required to elute all the benzyl alcohol, and o-cresol began to appear after passage of 580 mL mobile phase (Figure 2a, solid line).

Separations 2016, 3, 32 Separations 2016, 3, 4

4 of 8 4 of 8

Figure 2. Separation Separation of ofbenzyl benzylalcohol alcoholand ando-cresol: o-cresol: Benzyl alcohol peak in mobile-phase eluate Figure 2. (a)(a) Benzyl alcohol peak in mobile-phase eluate was was followed by beginning elution of o-cresol; (b) o-Cresol distribution in the stationary phases followed by beginning elution of o-cresol; (b) o-Cresol distribution in the stationary phases drained drained from individual separation columns. Solid line: regular setup; dashed line: high dead-volume from individual separation columns. Solid line: regular setup; dashed line: high dead-volume setup. setup.

At the end of the run, drainage of the stationary phases in the individual separation columns At the end of the run, drainage of the stationary phases in the individual separation columns enabled the recovery of o-cresol in columns 1–10, peaking at column-5 (Figure 2b, solid line). enabled the recovery of o-cresol in columns 1–10, peaking at column-5 (Figure 2b, solid line). Importance of increased dead-volume in the columns was examined by repeating the run, filling Importance of increased dead-volume in the columns was examined by repeating the run, filling at at the start 80% (v/v) of each column (around 95 mL) with the denser stationary phase and 20% the start 80% (v/v) of each column (around 95 mL) with the denser stationary phase and 20% (v/v) (v/v) with the lighter mobile phase, and connecting the columns with longer and wider-bore tubings with the lighter mobile phase, and connecting the columns with longer and wider-bore tubings (see (see Methods). Total volume in the ten columns plus connecting tubings was around 1170 mL in Methods). Total volume in the ten columns plus connecting tubings was around 1170 mL in the the regular runs, and around 1190 mL in the high dead-volume runs. In this instance, the benzyl regular runs, and around 1190 mL in the high dead-volume runs. In this instance, the benzyl alcohol alcohol peak was widened substantially, and the spread of o-cresol in the stationary phases in the peak was widened substantially, and the spread of o-cresol in the stationary phases in the columns columns widened to a smaller extent (Figure 2a,b, dashed line), showing the need to limit excessive widened to a smaller extent (Figure 2a,b, dashed line), showing the need to limit excessive deaddead-volume of mobile phase over and above that which is present inside the connecting tubings volume of mobile phase over and above that which is present inside the connecting tubings between between columns, in order to restrict peak widening—especially for the eluted solutes. columns, in order to restrict peak widening—especially for the eluted solutes. 3.2. Separation of Benzoic Acid, Phenol, and Benzyl Alcohol Using n-Butanol–NaHCO3 (aq) and 3.2. Separation of Benzoic Acid, Phenol, and Benzyl Alcohol Using n-Butanol–NaHCO3(aq) and n-Butanol– n-Butanol–NaOH(aq) Two-Solvent Systems NaOH(aq) Two-Solvent Systems In this run, water-saturated n-butanol was employed as the mobile phase. An n-butanol–0.1 M In run, water-saturated n-butanol was employed as the mobile phase. An n-butanol–0.1 M NaHCOthis 3 (1:1, v/v) mixture was shaken and allowed to settle, and its lower phase employed as the NaHCO 3 (1:1, v/v) mixture was shaken and allowed to settle, and its lower phase employed as the stationary phase in columns 1–5. An n-butanol–0.1 M NaOH (1:1, v/v) mixture was shaken and allowed stationary phase in columns An n-butanol–0.1 M NaOHphase (1:1, v/v) mixture6–10. was shaken and to settle, and its lower phase 1–5. was employed as the stationary in columns The sample allowed to settle, and its lower phase was employed as the stationary phase in columns 6–10. The mixture consisting of 1 g of benzoic acid, 1 g of phenol, and 3 g of benzyl alcohol was injected into sample mixture consisting of 1 g of benzoic acid, 1 g of phenol, and 3 g of benzyl alcohol was injected column-1 to begin the run. Monitoring of the eluate from the last column showed that the passage of into column-1 to begin run. of thewas eluate from the last column showed that the passage 340 mL of mobile phasethe into theMonitoring column system required to elute all the benzyl alcohol (Figure 3a, of 340 mL of mobile phase into the column system was required to elute all the benzyl alcohol (Figure solid line). At the end of the run, drainage of the stationary phases in the individual separation 3a, solid line). At the of theofrun, drainage stationary phases in individual columns enabled the end recovery benzoic acid of in the columns 1–5, peaking atthe column-2, andseparation phenol in columns enabled the recovery of benzoic acid in columns 1–5, peaking at column-2, and phenol in columns 7–10, peaking at column-8 (Figure 3b, solid line). columns 7–10, peaking at column-8 (Figure 3b, solid line). Thus the use of two different stationary phases—one in columns 1–5 and the other in columns 6–10—made possible the separate retention of benzoic acid in columns 1–5, and phenol in columns 7–10. The importance of dead-volume in the columns was tested again by filling at the start 80% (v/v) of each column with the denser stationary phase and 20% (v/v) with the lighter mobile phase, and the use of longer and wider-bore connecting tubing between columns. In agreement with the results of Figure 2, the eluted peak of benzyl alcohol was widened substantially, whereas the spreads of benzoic acid and phenol in the stationary phases were only somewhat widened, possibly owing to a compression effect by the stationary phase (Figure 3a,b, dashed line).

Separations Separations 2016, 2016, 3, 3, 32 4 Separations 2016, 3, 4

55 of of 88 5 of 8

Figure 3. Separation of benzyl alcohol, phenol, and benzoic acid: (a) Benzyl alcohol peak was followed Figure 3.3.Separation of alcohol, phenol, acid: (a) alcohol peak was Figure Separation ofbenzyl benzyl alcohol, phenol,and andbenzoic benzoic (a)Benzyl Benzyl alcohol peakphenol wasfollowed followed by initial appearance of phenol in mobile-phase eluate; acid: (b) Benzoic acid (Z) and (P) in by initial appearance of phenol in mobile-phase eluate; (b) Benzoic acid (Z) and phenol (P) in by initial appearance of phenol in mobile-phase eluate; (b) Benzoic acid (Z) and phenol (P) in stationary stationary phases drained from individual separation columns. Solid line: regular setup; dashed line: stationary phasesfrom drained from individual separation Solidregular line: regular dashed phases drained individual separation columns.columns. Solid line: setup;setup; dashed line: line: high high dead-volume setup. high dead-volume setup. dead-volume setup.

Thus the use of two different stationary phases—one in columns 1–5 and the other in columns Thus the use of two different stationary in columns 1–5 and the other in columns 3.3. Separation of Baicalein Wogonin Using of aphases—one n-Butanol–K (aq) Two-Solvent System 3 PO 6–10—made possible the and separate retention benzoic acid in4columns 1–5, and phenol in columns 6–10—made possible the separate retention of benzoic acid in columns 1–5, and phenol in columns 7–10. The importance of dead-volume in the columns was tested again by filling An at the start 80% (v/v) this run, water-saturated n-butanol employed as theagain mobile n-butanol–0.1 M 7–10. In The importance of dead-volume in the was columns was tested by phase. filling at the start 80% (v/v) of3each column with the denser stationary phase and 20% (v/v) with the lighter mobile phase, and the K PO (1:1, v/v) mixture was shaken and allowed to settle, and its lower phase was employed as the of each4 column with the denser stationary phase and 20% (v/v) with the lighter mobile phase, and the use of longer and wider-bore connecting tubing between columns. In agreement with the results of stationary phase the columns. A sample mixture consisting of 20 mg baicalein and 20the mgresults wogonin use of longer andinwider-bore connecting tubing between columns. In agreement with of Figure 2, the eluted peak of benzyl alcohol was widened substantially, whereas the spreads of benzoic (Figure2,4) dispersed in 8ofmL n-butanol–0.1 K3 PO4 (1:1, v/v) mixture was injected intoofcolumn-1 Figure the eluted peak benzyl alcohol wasMwidened substantially, whereas the spreads benzoic acid and phenol in the stationary phases were only somewhat widened, possibly owing to a to begin run. in Monitoring of thephases eluate from lastsomewhat column showed flow-rate dependence acid and the phenol the stationary were the only widened, possibly owing to of a compression effect by the stationary phase (Figure 3a,b, dashed line).the passage of 240 mL, 360 mL, the DMCC chromatogram. At a flow rate of 4, 12, or 20 mL/min, compression effect by the stationary phase (Figure 3a,b, dashed line). or 540 mL of mobile phase through the column system was required to elute all the wogonin from the 3.3. Separation of Baicalein and Wogonin UsingmL, a n-Butanol–K 3PO4(aq) Two-Solvent System columns, and of the passageand of 1200 mL, Using 1400 or 1600 mL of4(aq) mobile phase was required to elute all 3.3. Separation Baicalein Wogonin a n-Butanol–K 3PO Two-Solvent System the baicalein, respectively (Figuren-butanol 5a). In this run, water-saturated was employed as the mobile phase. An n-butanol–0.1 M In this run, water-saturated n-butanol was employed as the mobile phase. An n-butanol–0.1 M the flow rate ofshaken the mobile was significant stationary phase retention, K3PONotably, 4 (1:1, v/v) mixture was and phase allowed to asettle, and itsfactor lowerinphase was employed as the K3PO4 (1:1, v/v) mixture was shaken and allowed to settle, and its lower phase was employed as the which wasphase foundintothe becolumns. 95.8%, 87.9%, 83.4%, and 79.0% for flow of 4, 12, 16, 20 mL/min, stationary A sample mixture consisting of 20rates mg baicalein andand 20 mg wogonin stationary phase in the columns. A sample mixture consisting of 20 mg baicalein and 20 mg wogonin respectively. The correlation between flow M rate stationary phase retention (Figure 4) dispersed in 8 mL n-butanol–0.1 K3(x) POand 4 (1:1, v/v) mixture was injected (y) intoconformed column-1 to (Figure 4) dispersed in 8 mL n-butanol–0.1 M K3PO 4 (1:1, v/v) mixture was injected into column-1 to 2 the linear y = −of 1.0409x + 100,from withthe R = 0.9987 (Figure 5b). Thus, a fastdependence flow rate brought begin the relationship run. Monitoring the eluate last column showed flow-rate of the begin the run. Monitoring of the eluate from the last column showed flow-rate dependence of the about loss of stationary by mL/min, increasedthe mobile phase the columns DMCCincreased chromatogram. At a flow phase, rate of replaced 4, 12, or 20 passage of within 240 mL, 360 mL, or and 540 DMCC chromatogram. At a flow rate of 4, 12, or 20 mL/min, the passage of 240 mL, 360 mL, or 540 tubings, thereby possibly delaying the migration of the solutes and adversely affecting separation. mL of mobile phase through the column system was required to elute all the wogonin from the mL of mobile phase through the column system was required to elute all the wogonin from the The importance ofpassage employing an adequate of columns was illustrated in Figure 5c,to where columns, and the of 1200 mL, 1400number mL, or 1600 mL of mobile phase was required eluteten all columns, and the passage of 1200 mL, 1400 mL, or 1600 mL of mobile phase was required to elute all columns wererespectively required to separate the wogonin (W) and baicalein (B) peaks. the baicalein, (Figure 5a). the baicalein, respectively (Figure 5a).

Figure Structures of Baicalein and Wogonin. Figure4.4. 4.Structures Structuresof ofBaicalein Baicaleinand andWogonin. Wogonin. Figure


6 of 8


6 of 8

Figure 5. 5. Separation Separation of wogonin and and baicalein: baicalein: (a) baicalein (B) (B) in in Figure of wogonin (a) Separation Separation of of wogonin wogonin (W) (W) and and baicalein eluate at a flow rate of 4, 12, or 20 mL/min using ten columns; (b) Effect of flow rate on stationary eluate at a flow rate of 4, 12, or 20 mL/min using ten columns; (b) Effect of flow rate on stationary phase retention; retention; (c) (c) Effect Effect of of number number of of columns columns on on separation. separation. phase

Notably, the flow rate of the mobile phase was a significant factor in stationary phase retention, 4. Discussion which was found to be 95.8%, 87.9%, 83.4%, and 79.0% for flow rates of 4, 12, 16, and 20 mL/min, The DMCC adapts the powerful countercurrent distribution [21] a continuous respectively. Themethod correlation between flow rate (x) and stationary phaseprinciple retention (y)toconformed to 2 column chromatographic format for solute separations, which makes continuous countercurrent the linear relationship y = −1.0409x + 100, with R = 0.9987 (Figure 5b). Thus, a fast flow rate brought equilibration between immiscible phases possible without the use of centrifugation. to about increased loss oftwo stationary phase, replaced by increased mobile phase within the Scaling columnsup and large scale operations can therefore be readily achieved through the use of large columns, avoiding tubings, thereby possibly delaying the migration of the solutes and adversely affecting separation. the costs ofof centrifugal Asnumber illustrated in Figures 2 and 3, fractionated can Thehigh importance employingequipment. an adequate of columns was illustrated in Figuresolutes 5c, where be from required either thetomobile-phase eluate from series of columns or from the stationary tencollected columns were separate the wogonin (W) aand baicalein (B) peaks. phases in the individual columns at the end of a chromatographic run. Collection from stationary phases eliminates the need to elute all the solutes from the separation columns. By shortening the 4. Discussion chromatographic run and reducing the solvent requirement, it is particularly well-suited for the The DMCC method adapts the powerful countercurrent distribution principle [21] to a fractionation of slow-eluting solutes. continuous column chromatographic format for solute separations, which makes continuous Furthermore, in column chromatography, gradient elution can be introduced to improve countercurrent equilibration between two immiscible phases possible without the use of separation. The use of multiple solid stationary phases in stationary phase-optimized selectivity centrifugation. Scaling up to large scale operations can therefore be readily achieved through the use liquid chromatography (SOSLC) furnishes another approach to enhance separation [22–25]. In DMCC, of large columns, avoiding the high costs of centrifugal equipment. As illustrated in Figures 2 and 3, separation can be enhanced by the use of multiple liquid stationary phases. In separating benzyl fractionated solutes can be collected from either the mobile-phase eluate from a series of columns or alcohol, phenol, and benzoic acid, for example, the distribution coefficients were notably very close for from the stationary phases in the individual columns at the end of a chromatographic run. Collection benzyl alcohol (KD = 4.77) and phenol (KD = 4.68) in an n-butanol–0.1 M NaHCO3 two-solvent system, from stationary phases eliminates the need to elute all the solutes from the separation columns. By and they were very close for phenol (KD = 0.49) and benzoic acid (KD = 0.50) in an n-butanol–0.1 M shortening the chromatographic run and reducing the solvent requirement, it is particularly wellNaOH solvent system. Therefore, the separation of benzyl alcohol, phenol, and benzoic acid is suited for the fractionation of slow-eluting solutes. difficult to achieve using DMCC with either one of these two-solvent systems. However, when DMCC Furthermore, in column chromatography, gradient elution can be introduced to improve was carried out in Figure 3, placing the lower phases from these two-solvent systems separately in separation. The use of multiple solid stationary phases in stationary phase-optimized selectivity columns 1–5 and columns 6–10 as stationary phases, separation of these three solutes was obtained liquid chromatography (SOSLC) furnishes another approach to enhance separation [22–25]. In using only ten columns with a combined column volume of ~1200 mL. Likewise, either a succession of DMCC, separation can be enhanced by the use of multiple liquid stationary phases. In separating multiple mobile phases—or a stream of mobile phase incorporating a concentration gradient in one or benzyl alcohol, phenol, and benzoic acid, for example, the distribution coefficients were notably very more of its chemical components—may also be employed during a DMCC rum. close for benzyl alcohol (KD = 4.77) and phenol (KD = 4.68) in an n-butanol–0.1 M NaHCO3 two-solvent system, and they were very close for phenol (KD = 0.49) and benzoic acid (KD = 0.50) in an n-butanol– 0.1 M NaOH solvent system. Therefore, the separation of benzyl alcohol, phenol, and benzoic acid is


7 of 8

Complex solute mixtures are commonly encountered in plant extracts, and numerous medicinal herbs are known to contain a wide range of pharmacologically-active ingredients, many of which being present in small amounts that have not been fully investigated. For such potentially important medicinal ingredients from plants, mixtures of cellular RNAs and complex products of organic synthesis, etc., DMCC provides a convenient and useful method for both laboratory investigation and industrial production, with only minimal obstacles posed by extraneous factors such as inconvenience of operation [18], adsorptive losses [20], and high equipment cost. In conclusion, the DMCC method combines the outstanding advantages of countercurrent distribution and continuous column chromatography. Both the number of columns employed and the dimensions of the columns can be varied to suit the scale of operation. As in any countercurrent distribution system, the separation attainable between any pair of solutes is predicted almost exclusively by the difference between their distribution coefficients in the immiscible mobile and stationary phases. Accordingly, provided that the solutes are not chemically affected by the two solvents, and that there is a sufficient difference between their distribution coefficients, DMCC may be expected to provide a basis for their separation. Acknowledgments: We thank J. Tze-Fei Wong for valuable discussion and Peggy Lee for expert support. We would like to indicate that a Provisional US Patent on the method has been filed by the HKUST R & D Corp of the Hong Kong University of Science and Technology of which Timothy Yiu-Cheong Ho and Hong Xue are inventors. Author Contributions: The article was written and approved by both authors. Timothy Yiu-Cheong Ho and Hong Xue participated in DMCC design and data analysis; Timothy Yiu-Cheong Ho performed the experiments. Conflicts of Interest: The authors declare no conflicts of interest.

Abbreviations The following abbreviations are used in this manuscript: CPC CPLC DCCC DMCC HSCCC KD RNA SOSLC

Centrifugal Partition Chromatography Controlled-cycle Pulsed Liquid-liquid Chromatography Droplet Countercurrent Chromatography Dispersed Mobile-phase Countercurrent Chromatography High Speed Countercurrent Chromatography Distribution coefficient Ribonucleic acid Stationary-phase Optimized Selectivity Liquid Chromatography

References 1. 2. 3.

4. 5. 6. 7.

Craig, L. Identification of smalll amounts of organic compounds by distribution studies: II. Separation by counter-current distribution. J. Biol. Chem. 1944, 155, 535–546. Kimura, Y.; Kitamura, H.; Hayashi, K. A method for separating commercial colistin complex into new components: Colistins pro-A, pro-B and pro-C. J. Antibiot. (Tokyo) 1982, 35, 1513–1520. [CrossRef] [PubMed] Li, H.B.; Chen, F. Isolation and purification of baicalein, wogonin and oroxylin a from the medicinal plant scutellaria baicalensis by high-speed counter-current chromatography. J. Chromatogr. A 2005, 1074, 107–110. [CrossRef] [PubMed] Apgar, J.; Holley, R.W.; Merrill, S.H. Purification of the alanine-, valine-, histidine-, and tyrosine-acceptor ribonucleic acids from yeast. J. Biol. Chem. 1962, 237, 796–802. [PubMed] Holley, R.W.; Apgar, J.; Everett, G.A.; Madison, J.T.; Marquisee, M.; Merrill, S.H.; Penswick, J.R.; Zamir, A. Structure of a ribonucleic acid. Science 1965, 147, 1462–1465. [CrossRef] [PubMed] Tanimura, T.; Pisano, J.J.; Ito, Y.; Bowman, R.L. Droplet countercurrent chromatography. Science 1970, 169, 54–56. [CrossRef] [PubMed] Berthod, A. Countercurrent Chromatography: The Support-Free Liquid Stationary Phase; Elsevier Science Ltd.: Boston, MA, USA, 2002; Volume 38, p. 397.


8. 9. 10. 11. 12. 13. 14.

15. 16. 17. 18.

19. 20. 21. 22.

23.

24. 25.

8 of 8

Berthod, A.; Maryutina, T.; Spivakov, B.; Shpigun, O.; Sutherland, I.A. Countercurrent chromatography in analytical chemistry (IUPAC Technical report). Pure Appl. Chem. 2009, 81, 355–387. [CrossRef] Conway, W.D. Countercurrent Chromatography: Apparatus, Theory and Applications; VCH Publishers: New York, NY, USA, 1990. Du, Q.; Wu, P.; Ito, Y. Low-speed rotary countercurrent chromatography using a convoluted multilayer helical tube for industrial separation. Anal. Chem. 2000, 72, 3363–3365. [CrossRef] [PubMed] Ito, Y.; Conway, W.D. High-Speed Countercurrent Chromatography; Wiley-Interscience: New York, NY, USA, 1996. Ito, Y.; Sandlin, J.; Bowers, W.G. High-speed preparative counter-current chromatography with a coil planet centrifuge. J. Chromatogr. A 1982, 244, 247–258. [CrossRef] Sutherland, I.A. Recent progress on the industrial scale-up of counter-current chromatography. J. Chromatogr. A 2007, 1151, 6–13. [CrossRef] [PubMed] Wu, S.; Sun, A.; Liu, R. Separation and purification of baicalin and wogonoside from the Chinese medicinal plant Scutellaria baicalensis georgi by high-speed counter-current chromatography. J. Chromatogr. A 2005, 1066, 243–247. [CrossRef] [PubMed] Foucault, A.P. Centrifugal partition chromatography. In Chromatographic Science Series; Foucault, A.P., Ed.; Marcel Dekker: New York, NY, USA, 1995; Volume 68, pp. 25–49. Marchal, L.; Legrand, J.; Foucault, A. Centrifugal partition chromatography: A survey of its history, and our recent advances in the field. Chem. Rec. 2003, 3, 133–143. [CrossRef] [PubMed] Murayama, W.; Kobayashi, T.; Kosuge, Y.; Yano, H.; Nunogaki, Y.; Nunogaki, K. A new centrifugal counter-current chromatograph and its application. J. Chromatogr. A 1982, 239, 643–649. [CrossRef] Kostanyan, A.E.; Voshkin, A.A.; Kodin, N.V. Controlled-cycle pulsed liquid–liquid chromatography. A modified version of Craig’s counter-current distribution. J. Chromatogr. A 2011, 1218, 6135–6143. [CrossRef] [PubMed] Belter, P.A.; Speaker, S.M. Controlled-cycle operations applied to extraction processes. Ind. Eng. Chem. Process Des. Dev. 1967, 6, 36–42. [CrossRef] Ito, Y.; Bowman, R.L. Countercurrent chromatography. Anal. Chem. 1971, 43, 69A–75A. [CrossRef] Craig, L.; Otto, P. Apparatus for countercurrent distribution. Anal. Chem. 1949, 21, 500–504. [CrossRef] Nyiredy, S.; Szucs, Z.; Szepesy, L. Stationary phase optimized selectivity liquid chromatography: Basic possibilities of serially connected columns using the “prisma” principle. J. Chromatogr. A 2007, 1157, 122–130. [CrossRef] [PubMed] De Beer, M.; Lynen, F.; Chen, K.; Ferguson, P.; Hanna-Brown, M.; Sandra, P. Stationary-phase optimized selectivity liquid chromatography: Development of a linear gradient prediction algorithm. Anal. Chem. 2010, 82, 1733–1743. [CrossRef] [PubMed] Chen, K.; Lynen, F.; Szucs, R.; Hanna-Brown, M.; Sandra, P. Gradient stationary phase optimized selectivity liquid chromatography with conventional columns. Analyst 2013, 138, 2914–2923. [CrossRef] [PubMed] Chen, K.; Lynen, F.; De Beer, M.; Hitzel, L.; Ferguson, P.; Hanna-Brown, M.; Sandra, P. Selectivity optimization in green chromatography by gradient stationary phase optimized selectivity liquid chromatography. J. Chromatogr. A 2010, 1217, 7222–7230. [CrossRef] [PubMed] © 2016 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC-BY) license (http://creativecommons.org/licenses/by/4.0/).