Effect of operating parameters on a centrifugal partition chromatography separation

- Practical study of stationary phase retention versus mobile phase flow rate.
- Effect of the centrifugal field strength (rotor rotation).
- Effect of flow rate and field strength on peak efficiencies.
- Optimization of preparative productivity.

Centrifugal partition chromatography (CPC) is the branch of countercurrent chromatography (CCC) that works with single axis hydrostatic columns with rotary seals. The hydrodynamic of the liquid stationary phase-liquid mobile phase equilibrium in the CPC chambers has been studied theoretically and with specially designed CPC columns. In this work, we selected a simple analytical separation (no loading study) of three test solutes, coccine red, coumarin and carvone, with a commonly used heptane/ethyl acetate/methanol/water 1:1:1:1 v/v biphasic liquid system and two different rotors: a commercially available 30-mL CPC instrument and a 80-mL prototype rotor designed for productivity. We fully studied this separation in many possible practical operating conditions of the two rotors, aiming at a generic column characterization. The rotor rotation was varied between 1000 and 2800 rpm, the aqueous mobile phase flow rate was varied between 1 and 22 mL/min with the 30-mL rotor and 10 and 55 mL/min with the 80-mL rotor, the upper limits being mechanical constraints and some liquid stationary phase remaining in the rotor. The variations of Sf, the volume ratio of stationary phase in the rotor, were studied versus mobile phase flow rate and rotor rotation speed. A maximum mobile phase linear velocity was found to depend on the centrifugal field for the 30-mL rotor. This maximum velocity was not observed with the 80-mL rotor. Studying the changes in coumarin and carvone peak efficiencies, it is established that the number of cells required to make one theoretical plate, i.e. one chromatographic exchange, is minimized at maximal rotation speed and, to a lesser extent, at high mobile phase flow rate (or linear velocity). Considering the throughput, there is evidence of an optimal flow rate depending on the rotor rotation that is not necessarily the highest possible.