This thesis includes a description of modifications of the Hilger atomic absorption apparatus, which was used for most of the work. These modifications were restricted to the atomiser and burner, which were replaced by a modified "Eel" flame. photometer atomiser-burner and resulted in improved sensitivity and instrumental stability for the atomic absorption of magnesium. A comparison of the performance of this unit with that of the unmodified Hilger apparatus is given. A "Handigas" butane-propane mixture)-air flame was used for most of this work, but a coal-gas-air flame was also studied and found to give slightly greater sensitivity. The method was found to be subject to interference from many elements. Strontium salts, employed as releasing agents to overcome the effect of other elements, were not completely effective as milligram amounts of several elements interfered even when strontium was present. Among the more serious interfering elements are: aluminium, iron, manganese and zirconium (less than 20 p.p.m. interfere); the alkali and alkaline earth metal salts (more than 200-500 p.p.m. interfere); phosphate (more than 100 p.p.m. Pā0ā uranium (more than 4,000 p.p.m.); arsenate and vanadate. An attempt is made to explain the mechanism of some of these interfering effects. A combination of strontium salt and acetyl acetone was found to over-come the effects of small amounts of several elements that form complexes with acetyl acetone (e.g. iron and aluminium) far more effectively than strontium alone. Larger amounts of many interfering elements are removed by a solvent extraction procedure employing acetyl acetone and chloroform. Elements which cannot be removed by this means may be separated by anion-exchange, volatilisation, electrolysis or precipitation. A spiking technique, which compensates JT/GB for the effects of small amounts of interfering elements, is described and enables many samples to be analysed without prior separations. The method described has good sensitivity (the limit of determination is approximately 1 microgram of magnesium in 50 ml. of solution). It has been applied to the analysis of clay samples, iron ore, limestone and uranium metal, oxides and processing solutions. The coefficient of variation of the method was determined using two clay samples and results of 2.0 and 4.6 percent, at magnesium oxide concentrations of 0.65 and 0.22 percent respectively, wore obtained. The speed of the method compares favourably with others described for the determination of microgram amounts of magnesium, but increases if large amounts of interfering elements are present.
| Identifer | oai:union.ndltd.org:netd.ac.za/oai:union.ndltd.org:rhodes/vital:4507 |
| Date | January 1963 |
| Creators | Taylor, John Douglas |
| Publisher | Rhodes University, Faculty of Science, Chemistry |
| Source Sets | South African National ETD Portal |
| Language | English |
| Detected Language | English |
| Type | Thesis, Masters, MSc |
| Format | 137 p., pdf |
| Rights | Taylor, John Douglas |
Page generated in 0.0026 seconds