EXPERIMENTAL FLUCTUATIONS OF FLOW-INJECTION SIGNALS

An experimental factorial design is proposed for investigating the effects of flow-rate, injection volume and coil length on the fluctuations of flow-injection signals. The precision of the spectrophotometric data at different points along the profile produced by dispersing a dye solution or by processing the modified Griess reaction for NO2-N and its dependence on the single-line flow-injection parameters were evaluated. Large variations of the relative standard deviation were found along the profiles produced by medium and limited dispersed systems. The peak maximum and five subsequent points in the trailing portion of the dispersed signals were characterized by the best homogeneity of variance. Analysis of the fluctuations revealed that the interaction between the studied parameters was always significant. The errors inherent in peak-height and peak-width measurements are discussed.