Calibration

In the context of chemical metrology, calibration is the process of relating aknown quantity of an analyte to the corresponding measured instrumental responsethrough a mathematical relationship. Calibration permits the assignment of analytelevels in unknown samples based on the known levels of the calibrants. Details ofthe calibration model are important to achieve accurate results. Several commonapproaches are used in calibrating methods. Most frequently, calibration models arebased on linear instrumental response, with mathematical models that include zerointercept, fixed intercept, unconstrained (fitted), and bracketed models. Wheninstrumental response is nonlinear, a linear model may still provide accurateresults if the calibration range is sufficiently limited. This presentation willprovide an overview and application of various calibration models, withrecommendations of ways to improve measurement accuracy. Examples are presented thatillustrate advantages and disadvantages for each of these models as applied to lowlevel samples and to unknowns with levels that span several orders ofmagnitude.