Metodologias eletroquímicas e espectroeletroquímicas de ordem superior para análise de matrizes complexas

Abstract

The effect of interferences is an important problem in Analytical Chemistry. Some approaches to deal with this problem are usually separation methods, deve-lopment of selective sensors, among others. There are also ways to mathematically separate the analyte signal, with it being acquired in multiple instrumental ways, extracting relevant information and gaining robustness to the interferences through mathematical properties. These are called multiway or higher-order methods. In this work, three methods based in electrochemical and spectroelectrochemical te-chniques were used to obtain multiway signals, applying them in the study of different systems. The first is a spectroelectrochemical method, combining chro-noamperometry and fluorescence. A proprietary cell and an interfacing system were developed for the experiment, with dopamine/epinephrine mixtures being used to test the methodology. Data was acquired in three ways (emission-time-potential) and the third-order models provided predictions with relative errors of prediction (REPs) of 1 to 2%, compared to the second-order models, presenting REPs of 9 to 40%, highlighting the advantages of expanding data dimensionality beyond second-order. The second method was the expansion of electrochemical impedance spectroscopy to a second order technique by varying the direct current potential. This method was applied to the quantification of paracetamol in urine, in the presence of added uric and ascorbic acid as interferents. Using the combi-nation of the real part of the impedance with the phase shift of the signal, it was possible to determine paracetamol in the range of 0.5 to 3.0 pmol.L¹ with a root mean squared error of prediction (RMSEP) of 0.38 pmol.L. Finally, the third method consisted of the application of multiway voltammetric techniques, creating models for the identification of synthetic cannabinoids in street samples (n = 48) The best performance was attained by models created with differential pulse vol-tammetry with gradual changes in pulse times, allowing the identification of drugs based on 8-quinoline indole-3-oate, indazole-3-amide and negative samples, with a 98.3% accuracy.