Considerations with QEMSCAN Grain Size Estimation

Most SEM-EDS systems have the ability to compute grain size. Not all systems use the same computational methods and we advise that you familiarise yourself with the details of your system of choice. In the case of QEMSCAN the presence of fine-grained inclusions in the mineral of interest (e.g.: a major ore-forming mineral like chalcopyrite, or key minerals in sedimentary systems such as quartz) has an affect on the computed mineral grain size, therefore, care must be taken when using automated grain size calculations.  

Figure 1: A single grain of chalcopyrite which contains inclusions of other minerals with a mean grain size of 12 µm. iDiscover calculates a grain size of 136 µm for this grain of chalcopyrite.

The example in figure 1 shows a single grain of chalcopyrite that hosts fine-grained inclusions averaging 12 µm in size, which account for 4.3% of the total pixel area, the bulk of which is chalcopyrite. Using the QEMSCAN size calculation this grain of chalcopyrite has a size of 136 µm. QEMSCAN computes a stereologically corrected size as the equivalent sphere diameter (ESD) of the surface to volume ratio, also known as the phase specific surface area (PSSA), which relies on randomly orientated and distributed particles.

Practically speaking, the calculation is based on the length of scan lines through the grain which start at the grain boundary on the left and ends at the point where a mineral composition other than the designated mineral  (chalcopyrite in this example) is encountered. Any one scan line has a surface area equal to 2 pixels, and a volume equal to the number of pixels in the line. As the grain of chalcopyrite in the example contains many small inclusions of other minerals, the size calculation is an average of line lengths measured from the grain boundaries to the inclusions, and the scan lines are therefore much shorter than the pixel width the actual grain.

If the presence of fine inclusions is ignored the grain size is typically underestimated. It is however possible to remove the inclusions for the purpose of size calculations. This is done by using two pre-processors. i.e.: the “granulator” to separate the mineral of interest from all other minerals, and then the “injector” to infill all the inclusion holes within that mineral grain. After these pre-processors are applied the size of the example grain is 287 µm. As always the size of single grain is an underestimation because of the effect of particle sectioning; however, the mean size for a population of particles processed in this way will be much more accurate. As a general comparison though, the example chalcopyrite grain measures 394 x 265 µm at the midpoint (Figure 2).

Figure 2: The same chalcopyrite grain as in figure 1 with small inclusions removed by pre-processing. iDiscover calculates a grain size of 287 µm for the processed grain of chalcopyrite.

The example above demonstrates a significant difference in grain size as a result of fine inclusions, which could have a serious impact where the data are used for mineral processing plant design or the interpretation of geological depositional environments. Using the above example, a theoretical grind size of 287 µm is required to liberate this particular grain of chalcopyrite from its surrounding mineral phases. Grinding the material to the originally proposed grain size of 136 µm is most likely a waste of valuable resources and an unnecessary load on the operational finances.