Quantitative mineralogy by XRD and TOPAS Rietveld refinement helps improving manganese smelter operation in South Africa. Accurate knowledge of the phase composition of the ore allows establishing empirical relationships between the quantitative mineralogical data and the amounts of energy and reductant needed during the smelting process.
Quantitative Mineralogy of Manganese Ore
South Africa is a world leading Manganese producer. Most of the material is obtained by smelting ores whose mineralogy is complex (Figure 1). Traditionally, quality control of smelter feeds is done by bulk chemical methods. However, the amount of energy and reductant (e.g. coal) needed to process the material in the smelter is determined by the phase composition of the ore, since the minerals have different melting points and valence states of Manganese. Therefore, the quantitative mineralogy of Manganese ore from the Kalahari deposit (Northern Cape province) was recently studied applying XRD and DIFFRACplusTOPAS .
Features of XRD and TOPAS Rietveld Analysis
XRD and TOPAS Rietveld analysis are favored over scanning electron microscope (SEM) techniques for the mineral system dealt with  because:
- the preparation of the XRD samples is cost effective and feasible for a smelter,
- mineral identification by XRD is considered more robust than by EDS (XRD patterns for minerals like hausmannite and bixbyite are clearly different while the EDS signatures are virtually identical),
- fine-grained texture of particles down to the sub-micron level may hinder the determination of hematite and other minerals by EDS, while this is irrelevant for XRD,
- the acceptable accuracy limit for the major phases (1 - 2 wt-%) is sufficient for evaluating the smelting behaviour of the ore.
Equipment Setup for Quantitative Mineralogy
Drill core samples and two Mn-ore reference materials SARM 16 and 17  were measured applying a D500 diffractometer, using Cu radiation, a secondary graphite monochromator and a scintillation counter. Step scanned powder patterns were analyzed using TOPAS and applying the fundamental parameters approach .
TOPAS quantitative analysis does not require calibration, does not suffer from tube ageing, considers the whole powder pattern and hence, is not troubled by peak overlap. Furthermore, physically sound crystallite size and strain parameters are obtained in addition to the phase concentrations.
Figure 1. TOPAS fit of Mn-ore data. Commonly present minerals are: braunite Mn2+Mn3+6SiO12, hematite Fe2O3, hausmannite Mn2+Mn3+ 2O4, jacobsite (Mn2+,Fe2+,Mg)(Fe3+,Mn3+)2O4, calcite CaCO3, kutnohorite Ca(Mn2+,Mg,Fe2+)(CO3)2, dolomite CaMg(CO3)2, rhodochrosite MnCO3, bixbyite (Mn3+,Fe3+)2O3, andradite Ca3Fe3+2(SiO4)3, clinochlore (Mg,Fe2+)5Al(Si3Al)O10(OH)8, manganite Mn3+O(OH), tephroite (Mn2+,Mg)2SiO4, serpentine, and trace minerals.
Accuracy of the XRD and TOPAS Rietveld Method
Up to nine minerals were quantified in low-grade ore and up to twelve in high-grade ore, respectively . The 3sigma absolute standard deviation of the phase content is found to be below 1wt-%. The accuracy of the XRD and TOPAS Rietveld method, which is very fast, was validated by independent but slower and more demanding methodes (quantitative SEM and ICPOES). Element and oxide concentrations were calculated from the XRD phase composition according to the stoichiometry of the phases. They agree very well with the micro-analysis results. Figure 2 shows the very good agreement for the reference ores SARM 16 and 17.
Figure 2. Agreement plot of concentrations determined using TOPAS Rietveld  versus reference data for the standard manganese ores SARM 16 (high grade, filled symbols) and 17 (low grade, open symbols). The line displays 1:1 agreement.
XRD and TOPAS analysis provide quantitative mineralogy results that are suitable for quality control in Mn smelters. This is important since different ore compositions may satisfy the grade requirements of the ore but exhibit different process parameters. The TOPAS results allow the reliable prediction of smelter operation parameters, such as energy consumption of the furnace and the amount of reducing agent. Furthermore, phases such as carbonates are quantified that may act as natural flux agents.
A large potential for XRD and TOPAS based quality control can generally be predicted for ore quality control and optimising smelter operation. Additionally, the application of fast one-dimensional detectors with capability to discriminate fluorescence radiation (LynxEye™) improves data collection speed and makes routine quality control possible.
This information has been sourced, reviewed and adapted from materials provided by Bruker X-Ray Analysis.
For more information on this source, please visit Bruker X-Ray Analysis.