Abstract: A flow injection spectrophotometric method for the determination of boron in ceramic materials is described. The method is based on spectrophotometric measurement of the decrease in the pH produced by the reaction between boric acid and mannitol in the presence of an acid-base indicator. A bichannel FI (flow injection) manifold in which the sample solutions were injected into deionized water (at pH 5.4) and the stream was later merged with the reagent stream (a mannitol solution containing 1 x 10^-4 mol L-1 bromocresol green at pH 5.4), was used. Transient signals were monitored at 616 nm. A theoretical model which describes the dependence between the absorbance values and boric acid concentration. is presented. The model predicts a nonlinear dependence between the absorbance or increment in absorbance and the boric acid concentration. In contrast, the model predicts a linear dependence between the inverse of the absorbance values and the boric acid concentration. The calibration graphs (1/A vs. µg mL-1 B2O3) were linear over the range 1-30 µg mL-1 of B2O3. The relative standard deviations were 0.7 and 0.4% for 4 and 8 µg mL-1 of B2O3, respectively. The limit of detection was 0.02 µg mL-1 of B2O3 (3s criterium). The method was used to determine boron in nine ceramic materials with very different nominal boron compounds. The results were compared with those obtained using a potentiometric titration method as reference method. No significant differences (at 95% probability level) were found between the proposed and reference methods. The method is rapid, reliable, precise and free of interferences.

Abstract: A versatile manifold is proposed which permits the online dilution of samples previously fused with metal carbonates. The manifold was used in the determination of Al, Fe, Ca, Mg, Na and K in ceramics, e.g., clay, kaolin, quartz, felspar, stoneware and porcelain. The effect of the flow injection parameters on the characteristics of the flame ASS determination was studied and results from analysis of real samples compared with results from batch analyzes.