Comments on methods for evaluating SEC column performance

Abstract

The parameter that is used most frequently to summarize SEC column performance is plate number. The number of theoretical plates, N, is defined by N = ( ? ) 'where tR is the peak retention time and o' is the variance. The column plate number is determined experimentally by injecting a low molecular weight material into the apparatus and calculating N according to eq. Plate count has limited use in selecting SEC columns since the experimental solute usually has much smaller size than the polymeric species that are of prime interest in these analyses.Balke' has recently reviewed techniques for assessment of resolution of SEC columns and their use in comparing the fractionation quality of different columns. Of these methods, the most useful appear t o be the resolution, the valley-to-peak ratio, and the peak separation.Resolution, R,, is frequently used in chromatography as a measure of peak overlap. This parameter is defined by eq. (3).where the numerator is the difference between the retention times of neighboring peaks (subscripted 1 and 2) and the denominator is the mean width of the two peaks.This method of determining resolution is simple and quantitative if peak widths are measured by drawing tangents to the curves at half heights, rather than by assuming Gaussian forms?If the peaks are indeed Gaussian then R , can be related to the number of theoretical plates by assuming that the plate number is consistant for neighboring peaks. Then R , = 0.25m( -)(a kHere k is the mean capacity factor for the two peaks, where the capacity factor for peak c, k,, is defined by