Experimental design methodology applied to the study of channel dimensions on the elution of red blood cells in gravitational field flow fractionation.

Field flow fractionation (FFF) separation techniques have gained considerable success with micron-sized species. Living red blood cells (RBCs) of any origin have emerged as ideal models for cell separation development. Their elution mode is now described as "Lift-Hyperlayer". Certain separator dimension parameters are known to play a key role in the separation and band spreading process. Systematic studies of channel dimensions effects on RBC retention, band spreading, peak capacity and on a novel parameter described as "Particle Selectivity" were set up by means of a two-level factorial experimental design. From experimental results and statistical calculations it is confirmed that channel thickness plays a major role in retention ratio, peak variance, peak capacity and particle selectivity. Channel breadth strongly influences plate height, with lower impact on peak capacity and particle selectivity. Retention ratio, peak variance and peak capacity observed results are modulated by second-order interactions between channel dimensions. Preliminary rules for channel configurations are therefore set up and depend on separation goals. It is shown that a very polydisperse population is best disentangled in a thin and narrow channel whatever its length. If a mixture of many different micron-sized species is considered (each of limited polydispersities); a thick and broad channel should be preferred, with length modulating peak capacity to disentangle this polymodal mixture.