Correction for long‐term instrumental drift

A method of correcting for instrumental drift must be associated with any calibration procedure in order to validate the stored calibration data (slopes and intercepts) over a long period of time. Indeed, it can usually be observed that the drift is negligible for a 24 h period, but over longer periods corrections to the measured intensities have to be made. These corrections are based on the measurement of special specimens known as drift monitors or simply monitors. Physically, the monitor can comprise one to several specimens, each containing one or several analytes of which the intensity of each analyte is slightly higher than the highest intensity in the analyte concentration range. The other two essential properties of a monitor are its stability over time and reproducibility of its intensity measurements. A drift correction must be applied to the measured intensities of every analyte using one, two or several monitors depending on the spread of the intensity ranges. Some drift correction methods are proposed and it is explained how to combine them with the calibration procedure in order to obtain precise analytical results over long periods of time from the same set of calibration data. Copyright © 2002 John Wiley & Sons, Ltd.