Considerations about the Equipment Mainly the Centrifuge

Various prototypes of counter-current chromatographs have been ingeniously designed and developed by Ito. However, only four types are used to measure partition coefficients due, at least in part, to their commercial availability. These are multichannel cartridges CPC (Sanki Engineering, Kyoto, Japan), toroidal coil planet CPC, flow-through multilayer coil planet CPC (also called Ito multilayer separator-extractor, P. C.

Inc., Kim Place, Potomac, Maryland, USA) and horizontal flow-through multilayer CPC (model CCC-1000, Pharma-Tech Research, Baltimore, Maryland, USA). In contrast to the coil type column where the tubing diameter and length and the total volume can be varied, the cartridges in multichannel CPC apparatus are not readily modifiable.

For hydrodynamic reasons such as pressure build-up in the column [17], our laboratory has given the preference to the CPC systems employing hydrodynamic equilibrium mechanisms. This however does not imply that other types of CPC systems cannot be used for lipophilicity measurements as evidenced in the next section. The Ito multilayer coil separator-extractor is a simple centrifuge with a single coiled column balanced by counter-weights. This suggests the need for a readjustment of counter weights whenever the volume ratio of organic and aqueous phases is changed, since a fine balance is of particular importance in stabilizing the system. A defect in balancing may lead to a continuous "bleeding" of the stationary phase and perturbs the measurement. In addition, when filling a vertical coiled column with volatile organic solvents such as chloroform, air bubbles are likely to accumulate in the tubing and will equally perturb the measurement. In contrast, the horizontal flow-through multilayer CPC uses three coiled columns to avoid the problems of imbalance when changing the volume ratio. The problem of air bubble accumulation in the column is also easier to overcome when using horizontal coiled columns. However, this type of CPC has its own limitations: the increasing number of intercolumn connections would increase the frequency of tubing disconnection and the leakage of phase solutions. Fortunately, the design of intercolumn connections was much improved during the past few years, and tubing disconnection is no longer a major practical problem.

It is immediately apparent that the diameter and length of tubing are important parameters in influencing the partitioning equilibrium of solutes. For the purpose of measuring the partition coefficients of very lipophilic or very hydrophilic compounds, it is important to have a maximal retention of the stationary phase. In our experience, a length of 50-60 m of tubing is sufficient to ensure a partitioning equilibrium of solutes between the two phases. Increasing the internal diameter of tubing generally results in a greater retention volume of the stationary phase, but the use of tubing with an i. d. > 2.6 mm does not improve further the retention. The polytetrafluoroethylene (PTFE) tubing with an i. d. of 2.6 mm and an o. d. 3.4 mm (#10, Zeus Industrial Products, Orangeburg, Southern California, USA) has thus become our optimal choice. This internal diameter together with a length of 50-60 m yields a total volume of ca. 300-320 ml.

Another factor critical to the accuracy of measurements is the quality of tubing. The chemical and mechanical properties of tubing must be carefully considered. To allow most organic solvents and solutes at all pH values, it must be chemically resistant. As the tubing is under centrifugal forces, mechanical resistance should also be taken into account. Also, the hardness of tubing would surely influence the adsorption of solutes on the tubing and hence the peak symmetry of chromatograms. The PTFE tubing fulfills satisfactorily these criteria and is economically reasonable.

6.4.3 Experimental Design

Experimental layout for the measurement of partition coefficients using CPC is similar to that of capacity factors using HPLC, except that CPC uses a centrifuge mounted with coiled columns rather than a solid phase column (Fig. 6). However, experimental design in various types of CPC can be quite different. The differences are particularly marked for HSES and HDES. Experimental designs from our laboratory with different CPC systems have been described elsewhere [10], only those with flow-through multilayer CPC being detailed here.

A Kontron model 420 HPLC pump (Kontron Instrument, Ziirich, Switzerland) was used to propel the mobile phase at a flow rate of 0.5-10 ml/min and a Kontron model 432 UV-visible detector (variable wavelength) coupled with a Hewlett-Packard 3392A integrator (Hewlett-Packard, Avondale, Pennsylvania, USA) to detect the eluate. A flowmeter (Phase Separations, Queensferry, UK) is used to measure precisely the flow-rate. It must be noted that a stable flow-rate is of critical importance for the measurements.

It is important to estimate the partition coefficient of the investigated compounds in order to set up optimal experimental conditions. We have used the CLOGP algorithm of Hansch and Leo [24] to estimate partition coefficients in «-octanol/water systems. As for partition coefficients in the other systems, a good guess based on the log Poct value and hydrogen-bonding capacity is often needed. For the operation procedures, we use a "normal mode" process. Namely, for compounds of log P values > 0, the organic phase is used as the mobile phase in a column with a total volume of c. 300 ml. For compounds of log P values < 0, the aqueous phase is used as mobile phase. Measurement begins by filling the columns at a flow-rate of 5 ml/min, the stationary phase being presaturated with the mobile phase. When filling the columns with volatile solvents such as «-heptane, the flow-rate should be reduced to prevent the accumulation of air bubbles in the columns. When the columns are full, the centrifuge is rotated at a speed of c. 800-1000 r.p.m. and the mobile phase is propelled into the columns. For different log P ranges, the flow-rate of mobile phase should be accordingly adjusted in

Pump flowmeter

UV/vis detector

outlet of the mobile phase solution

Inlet of the mobile phase solution recorder/integrator centrifuge

Figure 6. Experimental layout of centrifugal partition chromatographs for partition coefficient measurements.

order to elute the solute at an appropriate retention time. Here we recommend an approximate scheme for the use of flow-rate at different log P ranges (see also Fig. 7):

Under flow rates of 0.5,1, 3 ad 6 ml/min ca. 88 %, 86 %, 83 % and 77 % of the stationary phase of an octanol/water system can be retained.

After the system has reached its equilibrium, i. e., no more stationary phase exudes from the columns, a Merck injector is used to inject 20 [xl of samples containing 0.1-5 mM of solutes dissolved in the mobile phase. The amount of sample injected should be appropriately increased by increasing either the concentration or the injected volume when using a higher flow-rate.

A precise determination of the dead volume or retention time of non retained solutes (/0) is of critical importance for the accurate measurement of partition coefficients, particularly for compounds with log P >2.5 or < -2.5. In the past, we used either anthracene or biphenyl as the nonretained solute when the organic solvent was the eluent. However, anthracene is easily oxidized in solution, while biphenyl is more stable and has hence become our preferred standard of t0 determination. As for nonretained hydrophilic compounds when water is the eluent, potassium dichromate is used satisfactorily in our laboratory. However, potassium dichromate appears to undergo

For log P > 2.5 or log P < -2.5 For 1.5 < log P < 2.5 or -2.5 < log P < -1.5 For 0.5 < log P < 1.5 or -1.5 < log P < -0.5 For 0 < log P < 0.5 or -0.5 < log P < 0

Organic phase as eluent

Aqueous phase as eluent

U (ml/min, the flow rate of mobile phase)

Figure 7. A proposed scheme for the operational flow-rates of the mobile phase in different log P ranges. The flow-rates of the mobile phase are selected and adjusted according to the different log P ranges of the investigated compounds.

chemical reactions and/or protonation at pH <ca. 3 and partition into the organic phase. In this case, cobalt chloride can be used to replace potassium dichromate for dead time determination.

In principle, UV-vis spectroscopy is not the only mode to detect eluates. Refracto-metry should allow detection of compounds lacking a chromophore. However, we have not yet investigated UV-inactive compounds for this type of measurement.

It is important to note that it is not necessary to empty and refill the columns each day provided that the same solvent system is used. The organic solvents can be reused after distillation and saturation with aqueous solution. Before changing the solvent system, it is advisable to wash the coils with methanol and dry them with a flow of air.

The above-mentioned experimental design and procedures are not applicable if a small column of say 30 ml is to be used. In such a case, the operation mode - called "reversed mode" in contrast to the "normal mode" described above - is completely different. For example, the aqueous phase is recommended as the mobile phase when measuring compounds of log P > 0. Since this type of experimental design would limit the measured range of log P values (ca. -2 to 2), it was not employed in our laboratory. However, this mode of operation may be used in combination with a column having a small total volume.

0 0

Post a comment