Data-processing method to reduce error coefficients for membrane-based analytical systems. 1. Amperometric-based sensor evaluated for quantification of oxygen.

Abstract

This paper describes the use of a predictive, curve-fitting method to reduce the effects of experimental variables on results obtained with membrane-based devices. Multipoint data from the transient regions of responses are used with suitable models and curve-fitting methods to predict the signal that would be measured for the system at equilibrium. The resulting equilibrium response usually is much less dependent on experimental variables than the transient responses used to predict it. The approach is evaluated for the membrane-based amperometric electrode for oxygen. Current vs time data are used to predict the equilibrium current expected when oxygen concentrations are the same on both sides of the membrane. Predicted equilibrium currents vary linearly with oxygen concentration. Relative to the more common steady-state method, the sensitivity of the predictive method is about 5-fold higher, the measurement time is about 17-fold shorter and the dependencies on membrane thickness and stirring rate are 125- and 8-fold lower, respectively. Pooled standard deviations (n = 40) correspond to uncertainties in oxygen concentration of about 0.009 mmol L-1.