Characterization of temperature correction method for kinetic methods of analysis with the extended Kalman filter

An extended Kalman filter for compensating for within-run temperature variations in kinetic methods of analysis is critically evaluated with first-order synthetic data. The kinetic data represent values ranging from 0.7 to 14 half-lives and include a random noise contribution to give approximate signal-to-noise ratios of 100, 1000, and 10,000. The theoretical and experimental performances of the method are compared by evaluating the data with three data analysis programs based on the extended Kalman filter algorithm. The effects of variations in temperature data morphologies are also considered. Criteria for selection of an appropriate data analysis method are presented, based on the results obtained in the comparative study.