Seleção de padrões internos para análise elementar em métodos instrumentais com plasma induzido

Abstract

This thesis approaches the challenges and strategies to use the internal standardization calibration method in two instrumental methods, i.e. inductively coupled plasma optical emission spectrometry (ICP-OES) and microwave induced plasma optical emission spectrometry (MIP-OES), for complex matrix analysis. In the first study, MIP-OES performance was evaluated for determination of 14 analytes with energies between 1.85 and 12.07 eV in urine. Three dilution levels (2, 20- and 200-fold) were evaluated to correct matrix effects, together with the internal standards (ISs) Ga, Ge, Pd, Rh, Sc and Y and with molecular species formed in nitrogen plasma (CN, N2, N2+ and OH). The limits of detection varied between 0.3 and 329 μg L-1. Using proper dilution, it was possible to use a single IS to correct matrix effects for all analytes, i.e. Ge, Rh or Sc for 20-fold dilution and N2+ for 200-fold dilution. In the next step was investigated interferences from two carbon sources, i.e. ethanol and oxalic acid, on the analytical signals of 10 elements by MIP-OES. Oxalic acid caused a signal increase for Al, Ca, Cd, Cu, Fe, Hg and Mg, and a signal suppression for As and P. High levels of ethanol (≥ 1 % m v-1) led to signal increments for all analytes and Bi, Ge and OH behaved as universal ISs. The external gas control module (EGCM) was also tested to correct matrix effects, however, it was not appropriate for Al, As, Hg and P. In the last study, the performance of both instrumental methods was evaluated for determination of 18 analytes in isotonic and energetic drinks. It was observed that argon plasma presented a higher robustness in comparison with nitrogen plasma, since only Si, Pb and Zn presented recoveries below 90 %. The criterion adopted to select internal standards in Ar-ICP and N2-MIP was verified and the best results were obtained through addition and recovery experiments rather than similar physical-chemical properties between analyte/IS. Multi-internal standard calibration (MISC), a recently proposed method in the literature, was also evaluated. It was demonstrated as a simple strategy for correcting matrix effects. The selection of the best IS is not needed because it uses several internal standards. Recoveries between 90-110 % were obtained for most analytes in both matrices and instrumental methods, and LODs were comparable to those obtained in traditional calibration strategies. MISC allowed the use of species from the plasma for obtaining analytical curves, e.g. Ar lines (ICP) or molecular species (MIP) and presented itself as a feasible alternative to the use of internal standardization.