APPLICATION OF RECURSIVE ESTIMATION TO THE REAL-TIME ANALYSIS OF TRACE-METAL ANALYTES BY LINEAR SWEEP, PULSE, AND DIFFERENTIAL PULSE ANODIC-STRIPPING VOLTAMMETRY

The application of a Kalman filter to the analysis of data derived from the determination of trace quantities of lead in municipal and sea water analytes using differential pulse, pulse, and linear sweep anodic stripping voltammetry at the hanging mercury drop electrode and linear sweep anodic stripping voltammetry at the thin film mercury electrode is demonstrated. A critical comparison of this recursive estimator to other analysis techniques is presented. For these real analytes where substantial faradaic and nonfaradaic modeling errors occur, the concentration estimates computed by the real time Kalman filter and nonreal time multiple regression were generally more precise than those estimates computed by other digital methods for analyses where S/N <~ 5. At higher S/N, concentrations computed using more standard analysis algorithms were comparable to those determined by the real time Kalman filter or nonreal time multiple regression. © 1979, American Chemical Society. All rights reserved.