Methodology of Analysis of Unseparated Mixtures: Error Limits in Estimating the Total Analyte Concentration Recalculated to a Standard Substance

摘要

In estimating the total concentration of similar analytes recalculated to a standard substance X_(st), different sensitivities of their determination result in a systematic error δc, depending on the choice of X_(st) and the composition of the analyzed mixture. Two methods of predicting the limiting values of this error (~(δc)max), with and without the account of the properties of X_(st) are proposed and tested on model mixtures. The methods can be used if the total analytical signal linearly depends on the concentrations of analytes and is additive (spectrophotometry, conductometry, etc.). The value of ~(δc)max is independent of the ratio of analytes concentrations in the sample. To reduce δcmax, one should equalize the sensitivities of the determination of all compounds of a given type and use analytes with medium sensitivities as X_(st).