Quantification of iridium and other platinum group metals in the presence of naturally occurring contaminants in geological ore
Chiweshe, Trevor Trymore
MetadataShow full item record
The present thesis presents the use of scandium as internal standard in the quantification of precious metals from the CRM, inorganic salts, organometallic compounds and geological mineral ores. This method was developed using both the liquid and the geological CRMs and applied to various inorganic salts, organometallic compounds and geological mineral ores. Quantitative determination of precious metals from both CRMs using scandium as internal standard yielded total recoveries (100 %) for all the precious metals present in the samples except for osmium (83 %). Osmium recovery in the liquid CRM was found to be affected by the instability of the osmium calibration standards (NH4)2[OsCl6]. This instability in the calibration standards were mainly due to the production of OsO2 or OsO4 which was found to result in a steady increase in the slope of the calibration curves. Experimental results showed improved stability and osmium percentage recoveries (+ 99 %) using (NH4)2[OsCl6] calibration standards kept in the dark and at ca. 10 ºC. Total percentage recoveries of osmium was obtained using the Sc as internal standard in the liquid CRM using the newly established calibration standards Cs2[OsO2(CN)4] which was found to be stable at room temperature and at 10 ºC for a period of up to 5 days. The application of scandium as internal standard in the quantitative determination of precious metals in inorganic, organometallic compounds and geological ores was shown to yield excellent recoveries (+ 99 %) compared to the direct calibration methods. Fluctuating percentage recoveries were obtained using the direct calibration method as shown by the larger standard deviation (± 8 %) compared to those obtained using scandium as internal standard (± 2 %). Accurate determination of precious metals from inorganic and organometallic compounds was found to be dependent on the purity and the chemical composition of these compounds. The use of scandium as internal standard was also found to be affected by unmatched HCl and the presence of Na+ ion concentrations (EIE’s) of more than 10 % and 10 ppm respectively. Complete digestion of the chromitite and the Merensky reef mineral ore was achieved using the Na2HPO4/NaH2PO4∙H2O mixture and (NH4)H2PO4 flux. However, the biggest drawback of using the Na2HPO4/NaH2PO4∙H2O flux mixture was the introduction of the Na+ ion (EIE’s) into the analyte solution which resulted in inaccurate quantification. The use of Sc as internal standard was found to be ineffective to compensate for all these matrix effects caused by this substantial increase in the Na+ ion concentrations in the analyte solutions. The use of (NH4)H2PO4 flux was found to be a better alternative since it had an advantage of not introducing EIE’s but precipitating chromium as Cr2O3 from both mineral ores. The quantitative determination of precious metals from the mineral ore using the ICP-OES revealed a total percentage content of 0.52 % for all the PGM (Ru, Os and Pt) in the chromitite mineral ore and 0.09 % for all the PGM (Ru, Os, Ir and Pt) in the Merensky reef mineral ore using Sc as internal standard. Validation results of the instrumental parameters and conditions used in this study were shown to be the optimum conditions for precious metal analysis. Excellent calibration curves with linear regression line (r2) values of above 0.997, low detection limits (of at least 0.0001 ppm) and a wide working range of between (0.5 - 10 ppm) were used. The obtained results of precious metals in CRMs, inorganic salts, organometallic compounds and the geological mineral ores were validated using the t-statistic and were found to be acceptable at 95 % confidence interval. All osmium results were validated using the ANOVA test. Osmium recoveries (+ 99 %) obtained using (NH4)2[OsCl6] and Cs2[OsO2(CN)4] calibration standards kept at ca. 10 ºC using scandium as internal standard were accepted at 95 % confidence interval.