Error Sources in the Determination of Chromatographic Peak Size Ratios

Is it a problem to determine true peak sizes and peak size ratios? The reader may think the answer is, “no,” since today’s chromatographic separation systems are of excellent performance, and powerful integrators are on the market. However, let us take a closer look at the problem presented in Fig. 1 . A peak pair with true areas of 1:4 or a ratio R of 0.250 needs to be analyzed. This value is unknown to the analyst. In (a), a column with 10,000 theoretical plates (N) giving symmetrical peaks (tailing T = 1.0) is used. The separation is performed without programming (i.e., under isocratic or isothermal conditions), and the chromatographic resolution obtained is 1.52. The integrator gives a peak area ratio of 0.251 or 100.4% of the true value. In (b), the column deteriorated and gives asymmetrical peaks with T = 1.5. As a consequence, R increased and is now determined as 0.260 or 104.0%. In (c), the plate number dropped to 5000 giving R = 0.282 or 112.8%. If the analyst uses a column with identical performance, but opposite selectivity as in (d), (e.g., by taking a chiral stationary phase of S configuration instead of its R counterpart in enantioselective analysis) the result changes markedly: now R is 0.224 or 89.6%. Finally, in (e), the analysis is performed in a programmed mode giving peaks of equal width, but differing plate number; also tailing is suppressed by this approach. Again a different result is obtained and R is found to be 0.246 or 98.4%.