EXPERIMENTAL EVALUATION OF INDETERMINATE ERROR IN HEIGHT-WIDTH AND HEIGHT ONLY MEASUREMENTS OF CHROMATOGRAPHIC PEAKS

The four basic errors in the measurement of area of a chromatographic peak by height-width integration were isolated experimentally. The standard deviations (taken as measurements of the indeterminate errors) were: error in placement of the base line, about 0.010 cm; error in measurement of height from the base line, about 0.012 cm; error in positioning of a measuring instrument at an intermediate height, about 0.021 cm; and error in measurement of peak width, 0.0080 cm for sharp peaks and increasing as peaks become flatter. From these four basic errors in measurement the resultant errors in area were calculated according to the theory previously developed and found to agree well with the observed standard deviations in calculated areas for peaks of various shapes and sizes. The optimum peak shape for peaks of a given area has a ratio of height to widthat-half-height of about 3 to 5. For a peak of 15-cm2 area the standard deviation at this optimum is about 0.5%. For peaks of a given shape the relative error in area decreases according to the square root of the peak area. For peaks of a given height the error deinstrument creases with increasing width up to a limiting value of about 3- to 4-cm width. In general, peaks should be recorded so that peak height is near the maximum allowable by the chart paper and so that the width is about 3 to 4 cm. The use of peak height alone was also examined, theoretically and experimentally. Comparison of the peak-area and peak-height methods shows that in terms of indeterminate errors peak height is a much more precise measurement technique. Because of this inherently better precision, control of determinate errors is worthy of intensive effort. © 1968, American Chemical Society. All rights reserved.