One of the analytical techniques commonly used to determine the mineralogy of an ore is X-ray diffraction (XRD). XRD instruments, equipped with fast detectors, in conjunction with TOPAS quantitative phase analysis (QPA) provide standardless analysis within minutes. For well crystalline samples, the typical accuracy of QPA can be better 1 wt%.
This is demonstrated through the comparison of the bulk chemical composition estimated from the concentration of the minerals and their stoichiometry to chemical analysis.
This article discusses the ability of TOPAS to handle variable Al/Fe stoichiometry. This enables the minerals industry to determine severe recovery losses in iron ore and bauxite processing.
XRD in Mining
QPA is widely used for exploring geologic materials in research and service laboratories, as well as in the quality control of mining operations. Gaining insights into the properties of ore (minerals from which metals are extracted) and gangue (deleterious minerals that need to be separated from the ore) is economically significant for the process mineralogy because physical properties that influence the processability of the material are directly related to the crystal structure of the minerals, and not to their chemical composition.
Hence, these properties have a direct impact on the beneficiation conditions such as the separation technique (magnetic washing, gravity or dissolution).
In mining operations, recovery estimates typically rely on chemical grade-estimates. Severe recovery losses may occur if the element of interest is present in one of the gangue minerals that is not accessible or removed during the processing of the ore.
Assessment of Accuracy
Using QPA involving the Rietveld technique and XRD data is a direct technique for determining the relative or absolute phase abundances of crystalline and non-crystalline (amorphous or nano-crystalline) components in a mixture.
The accuracy of the QPA result may be corroborated through comparison with conventional chemical analysis.
The XRD based chemical analysis of a multi-phase mixture arises from the phase abundances and the known stoichiometry of the crystalline phases. This technique requires well defined composition of the crystalline phases. Site occupancies and site multiplicities of the crystal structure quantify the mass of the unit cell of the mineral.
Hence, the mass and the volume of the unit cell are equal to a calibration coefficient in quantitative analysis and the crystal structure means a restraint or hard constraint. For that reason, accurately understanding all crystal structures is imperative for high accuracy measurement.
Different types of atoms may join at the same location of the crystal structure. Standard crystal structure refinement of a single phase by the Rietveld technique is capable of determining site occupancy parameters.
This is difficult in QPA due to the availability of limited measurement time in process control. This leads to short scan ranges, which may not allow independent site occupancy refinement but permit TOPAS QPA. High accuracy QPA of mixed crystals is achievable through the stabilization of the refinement using geometric or chemical restraints.
QPA Analysis of Bauxite BXMG- 3
In the below example, a D4 ENDEAVOR diffractometer was used to acquire diffraction data of reference bauxite BXMG 3. CETEM provided the diffraction data as well as chemical analysis. Figure 1 shows the characteristic TOPAS QPA. Hematite (Hem, Fe2O3), goethite (Gt, FeOOH), and gibbsite (Gb, Al(OH) 3) are the major phases. Gibbsite exhibited a bimodal size distribution, which was measured by refining two phases that have different crystallite size parameters.
In addition, trace amounts of kaolinite (Kln, Al2Si2O5 (OH) 4), quartz (Qtz, SiO2), and rutile and anatase (Rt, Ana: TiO2) were determined and refined as well.
Figure 1. TOPAS QPA of bauxite certified reference material BXMG-3.
BXMG-3 mineral concentrations from TOPAS refinement for different models of Fe/Al substitution in goethite are summarized in Table 1. The lower line provides the Al fraction at the Fe site in goethite. Table 2 shows the bulk elemental composition derived from the individual phase compositions as per the stoichiometry of the minerals.
The composition is compared to the respective reference values from the CETEM certificate. The lower part of Table 2 delineates the discrimination in the bulk elemental concentrations between the nominal mineral formulae and the various models of substitution.
Table 1. BXMG-3 mineral concentrations from TOPAS refinement for different models of Fe/Al substitution in goethite
||Phase concentration / Wt%
The results must typically agree within 1% or better. Table 2 depicts nearly 3% overestimation of iron and an underestimation of aluminum for the XRD data, revealing the existence of some systematic error in the model.
Table 2. Element concentrations calculated from the phase concentrations obtained for the different models of Fe/Al site occupancy in goethite
|Bias: Certificate - TOPAS calculated concentrations
The refined lattice parameter c of goethite is 2.9750(7) A, representing a highly substituted Al-goethite with the occupancy of Al at the Fe site of 0.28, as illustrated in Figure 2. This result is further corroborated by independent refinement of the Fe site-occupancy in goethite (column "free" of Table 1), yielding a value of 0.21. The underestimation of Fe is minimized from nearly 3% to 1.5% (Table 2).
Figure 2. Variation of the goethite, (Fe/Al)OOH, lattice parameter c with the Al substitution in %.
Lattice parameters are characterized by the peak positions, which are typically much better resolved in the diffraction pattern when compared to the slight intensity changes caused by variations in concentration (scale factor) and site occupancies. This leads to improved accuracy in QPA. Figure 2 shows the difference of the lattice parameter c with the concentration of Al in goethite, FeOOH.
The regression equation for the lattice parameter is utilized to describe the occupancy of Al and Fe (Figure 3). The resulting phase and chemical compositions are depicted in column "latt_constr c" in Tables 1 and 2. The Al fraction reaches 0.27 with an overall improvement of the discrimenation for element concentrations better 1 %.
Figure 3. Example of a lattice restraint in TOPAS. It couples the Fe/ Al site occupancy to the variation of the lattice parameter c in goethite, c_gt.
A new feature of TOPAS 5 is chemical restraints, which relate the crystal structures’ occupancy parameters to the sample’s known bulk chemistry. Figure 4 illustrates the implementation of this for Fe and Al in goethite. The QPA results for BXMG-3 bauxite are summarized in columns ‘chem_constr’ of Tables 1 and 2. The illustrative results for restrained elements Al, Fe and Al, Fe, Si, and Ti show the influence of the chemical restraints, causing a slightly higher Al occupancy of 0.29/0.30. The agreement with chemical analysis is in the order of 0.5%. The factors 0.6994 for Fe and 0.52923 for Al account for the transformation from oxide concentrations (reported in the certificate) to elemental concentrations utilized in TOPAS.
Figure 4. Examples of TOPAS code for chemical restraints. They force the Fe/Al site occupancy parameter xal towards minimal difference between Fe and Al refined for the crystal structure and the bulk chemical analysis for Fe2O3 and Al2O3.
The TOPAS refinement of XRD data was used to determine the quantitative mineralogy of CETEM certified reference bauxite BXMG-3, providing accurate results as the main chemistry values obtained from the XRD results are in good agreement with bulk chemical analysis from the certificates (better than 1 wt%). This yields knowledge on element partitioning into different minerals, which is useful in the process mineralogy.
Lattice and site occupancy parameters are interrelated in substitutional mixed crystals. The quantification of their ratio in TOPAS Rietveld quantification can be performed through the built-in macro language, and significantly improves the accuracy of the quantitative X-ray mineralogy results. The same level of accuracy is achieved utilizing chemical restraints.
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.