SPECTROPHOTOMETRIC MULTICOMPONENT DETERMINATION OF YTTRIUM, BARIUM AND COPPER IN MICROSAMPLES OF YBA2CU3OX CERAMICS WITH NONADDITIVE BEHAVIOR OF ANALYTES

The stoichiometry of thin films and microcrystals of high-T(c) superconductive YBa2Cu3O(x) ceramics was characterized by means of a spectrophotometric multi-component determination with multivariate calibration. Partial least-squares (PLS) regression was applied as a mathematical algorithm. The applied coloured complex system Y-Ba-Cu-Arsenazo III at pH 4.7 in water-acetone solution shows non-additive behaviour owing to the formation of binuclear complexes, e.g., of Cu and Ba, which was proved by Job's method. The optimum wavelength selection was done heuristically on the basis of the sensor selectivity definition formulated by Kaiser. The concentration matrix in the form of the orthogonal factorial design for three variables with three levels including fifteen mixtures as calibration points in a relatively small concentration range was shown to be appropriate for approximation of the relationships in this system to the linear regression model. In microsamples of YBa2Cu3O(x) ceramics with a mass > 100-mu-g the elaborated method allows the determination of the stoichiometric coefficients referred to the Cu content within the error ranges Y1.00 +/- 0.03Ba2.00 +/- 0.07Cu3.00O(x).