New Analytical Approach to Determining Minerals in Coal and Ash Products Using X-Ray Diffraction

Background Coal mineral matter, particularly its composition, plays an important role in all phases of coal science from mine geology, resource development, preparation and utilisation. In the past, techniques available for determining individual mineral phases lacked the necessary accuracy to provide quantitative data. Of the techniques available, X-ray diffraction (XRD) is the most specific, in that it is based on the unique characteristic diffraction of x-rays from the crystal structure of each mineral. This was however, at best, semi quantitative because of the problems of imperfect crystalline phases, particularly prevalent in clay minerals.

The development of XRD analysis based on the Rietveld full profile opened the way for coal mineral quantitative analysis. This has been achieved with the development of SIROQUANT(tm), a computer software package capable of accurate multiphase analysis.

Objectives The aim of this project was to apply XRD to the determination of mineral phases in coal and ash products and quantify the minerals using SIROQUANT(tm). Specifically the objectives were as follows:

• Establish an accurate method for determining coal minerals in low temperature ashes and ash products.
• Apply the technique to the direct analysis of minerals in raw coals, that is, without the need for ashing.
• Develop a specific SIROQUANT(tm) Coal Package which would be user-friendly by minimising operator input and processing time.
  * Demonstrate the potential application to process studies such as coal preparation plants.
• Study mineral transformations using hot stage XRD and SIROQUANT(tm).

Conclusions Accurate quantitative data was obtained for minerals in a range of bituminous coals using X-ray diffraction and the SIROQUANT(tm) software package with the clay package. The quantification of major minerals was validated by analysis of synthetic mixtures of clays and quartz where the accuracy was better than 5 percent. For the analysis of low temperature ashes, the data was validated by comparison of the element oxide contents as determined by chemical analysis and those calculated from the minerals as determined by SIROQUANT(tm). Additional validation was demonstrated by comparison of the siderite content on some ashes compared to that inferred by chemical leaching. By the use of a spiking procedure, the SIROQUANT(tm) program allows the total amorphous material to be determined. The detection limits for most minerals was around 0.1 percent.

Raw coal was successfully analysed by the development of a structure model to simulate the organic fraction. This allowed the direct determination of the coal matter as well as the individual minerals. A good correlation (r2 = 0.89) was obtained between the radiofrequency ash value for the coals and the mineral matter as measured by the sumation of the individual mineral contents as found by SIROQUANT(tm). The data was validated by the good recoveries (>95 percent) obtained on samples of demineralised coal spiked with known amounts of kaolin and quartz. There were only minor differences in the shapes of the organic diffraction patterns of coals of different rank and it has been shown that there was little difference in the final results when various models for the coal humps were used. The detection limits for individual minerals was around 0.2 percent. The detection limits in raw coals were not substantially affected by the dilution of the mineral matter with the organic material. This was because of the very low mass absorption by the carbon, hydrogen and oxygen which are the constituents of the organic material.

A SIROQUANT(tm) Coal Package was developed based on a database of 50 minerals using a template task file. This package considerably reduces operator input and processing time. Good agreement was obtained between the mineral contents determined using the package and manual operation of the software. This package is being further refined and will be available commercially as an add-on to SIROQUANT(tm). The value of quantitative XRD was demonstrated in the determination of minerals in various streams of a coal preparation plant. This allowed the partitioning of minerals in the process to be determined and has important implications for coal preparation.

Analysis of laboratory and combusted coal ashes enabled the mineral transformations to be quantified. Specifically the conversion of clay minerals to mullite and the increased amorphous contents could be quantified.

Hot stage diffraction patterns of heated ash samples were analysed by SIROQUANT(tm) and it was possible to quantify the mineral transformations over the temperature range 800-1200?C. Quantitative analysis of coal ash over the temperature profile provided data on the production of mullite, cristobalite and magnetite and the reduction in quartz, hematite and illite.

In view of the importance of minerals transformations in studies of ash fusibility and the results obtained in this project, the Division has commissioned a state-of-the-art X-ray diffractometer with a hot stage capable of reaching temperatures of up to 1800?C. Further research, based on this instrument, will be undertaken to study mineral transformations in coal mineral matter over temperature ranges similar to those occurring in PF fired boilers.

This project has demonstrated that X-ray diffraction combined with SIROQUANT(tm) provides accurate quantitative data for minerals and crystalline phases in coal and ash products. This new capability will augment research associated with coal geology, mining, preparation and utilisation.