Comparative study of iron ore characterisation using a scanning electron microscope and optical image analysis

Abstract

In order to develop downstream processing routines for iron ore and to understand the behaviour of the ore during processing, extensive mineralogical characterisation is required. Microscopic analysis of polished sections is effective to determine mineral associations, mineral liberation and grain size distribution. There are two main imaging techniques used for the characterisation of iron ore, i.e. optical image analysis (OIA) and scanning electron microscopy (SEM). In this article, a QEMSCAN system is used as an example of SEM methodology and results obtained from it are compared against results obtained by the CSIRO Recognition3/Mineral3 OIA system. Both OIA and SEM systems have advantages and drawbacks. Even though the latest SEM systems can distinguish between major iron oxides and oxyhydroxides, it is still problematic for SEM systems to distinguish between iron ore minerals very close in oxygen content, e.g. hematite and hydrohematite, or between different types of goethite. Scanning electron microscopy systems also can misidentify minerals with close chemical composition, i.e. hematite as magnetite and vitreous goethite as hematite. In OIA, iron minerals with slight differences in their oxidation or hydration state are more easily and directly recognisable by correlation with their reflectivity. In both methods, the presence of microporosity can result in some misidentification, but in SEM methods misidentifications due to microporosity can be critical. Low resolution during QEMSCAN analysis can significantly affect the textural classification of particle sections. The main conclusion of this study is that, for low iron content ores or tailings, SEM systems can provide much more detailed information on the gangue minerals than OIA. However, for routine characterisation of iron ores with high iron content and containing a variety of iron oxides and oxyhydroxides, OIA is a faster, more cost effective and more reliable method of iron ore characterisation. A combined approach using both techniques will provide the most detailed understanding of iron ore samples being characterised.

Automated optical image analysis of natural and sintered iron ore

“…In this work, the authors reported segmentation of a mineral whose reflectivity area partially overlapped with the reflectivity area of epoxy. In these cases, the capability of some software packages to identify different textures can be utilized (see Donskoi et al, 2011Donskoi et al, , 2013a. In this example, the whole mineral reflectivity area is overlapped by some minor, extra differentiated areas.…”

Analysis of mineral grades for geometallurgy: Combined element-to-mineral conversion and quantitative X-ray diffraction

“…On the other hand, OIA-based automated mineralogy has the lower cost and is less time demanding. CSIRO (Donskoi et al, 2013) has developed an OIA-based automated mineralogy which has been used for iron ore characterization. In high grade samples, the OIA system results were close to the SEM-based automated mineralogy (QEMSCAN), but in lower grade samples and low resolutions the OIA results were drifting away from true values (i.e.…”

Mineral 4/Recognition 4: A Universal Optical Image Analysis Package for Iron Ore, Sinter and Coke Characterization

“…[9]), has better resolution for massive screening and better characterises porosity (for detailed comparison of the two techniques see Refs. [3,12]). …”

“…4 (reproduced from Ref. [12]) demonstrates the automated segmentation of two different types of hematite-martite and micro platy hematite-that have the same reflectivity but different morphology/texture. This feature can also be used for the segmentation of different morphologies of hematite and SFCA in iron ore sinter and IMDC (inert maceral derived components) and RMDC (reactive maceral derived components) in metallurgical coke.…”