Dear T. Hosseini, one can do a lot using such software but again, all the factors are somehow interfering so that the result often depends on the sequence you are doing the corrections. It becomes more problematic with increasing number of prameters which are related to refineable parameters of the phases. The vendors of such software packages have to implement all these influencing effects as request of different customers, but often they would also apply only one or two but never all at the same time.

This is a quite complex question and one could perhaps write pages about this topic. The best way is of course to calibrate the system first before any measurement (zero shift). This is a constant shift of all peaks. The described sample displacement is more complicated since your sample is out of focus what you can theoretically correct by a mathematical equation (which is used to "adapt" a displacement) but still you sample is not in focus what will influence your peak shape negatively. The third effect re the phases itself since the lattice parameters define the peak positions in also a quite complex way. This depends of course from the symmetry, i.e. form the number of independend lattice parameters. The forth effect is related to the structure, i.e. the atomic comfiguration within each cell. If there are substitutions the intensities are affected so that the superimposed imformation of different reflections of different phases can cause mis-refinements. All these four factors affect the peak positions in a very complex way so that the best way is to calibrate the system and take care on the correct position of the sample in order to reduce the number of possible degrees of freedom.

You can of course apply the procedure you did, but I would do this in maximum for a single phase. Everything else I would see only as attempt to reduce the mismatch, but it does not automatically means that you found the correct minima but only one of perhaps many minima.

Dear Omar,

You can do that roughly by identification software like MDI Jade. In this software first you should find one of the major peaks in your sample (for example quartz) and use the Jade toolbar to shift your spectra to either left or right to be matched with predefined quartz peak. After shifting you can start fitting of your samples with other possible components. Then you can save your text file and export it to excel.

Dear T. Hosseini, one can do a lot using such software but again, all the factors are somehow interfering so that the result often depends on the sequence you are doing the corrections. It becomes more problematic with increasing number of prameters which are related to refineable parameters of the phases. The vendors of such software packages have to implement all these influencing effects as request of different customers, but often they would also apply only one or two but never all at the same time.

Dear Omar,

A couple more points:

regarding your question 'Sample Displacement Error and X-Offset. What would be the best way of correcting the horizontal shifting in this case?'

Read this from the EVA manual:

"X offset and displacement errors are difficult to distinguish from one another, especially for scans measured below 90 2-theta. By chance, there is no disadvatage for phase analysis using the X offset treatment to correct an error caused by a sample displacement. However, it is unsuitable for crystallography - in fact, none of these correction should be used for crystallography, because most crystallography software can refine similar corrections."

Some other things to think about. If you have a natural mineral assemblage then the main quartz 101 peak may be overlapped by muscovite 003. It might be worthwhile looking at the smaller quartz 100 peak which is usually free from any interference.

You should perhaps use a pure reference material such as the corundum standard to check the alignment of the diffractometer first, make sure that the peaks are where they should be before running your natural assemblages.

Tuning the cell is not an easy approach. It is quite common for the calcite to contain some Mg and this can strain the lattice that could be fitted by carefully tuning the cell. Other minerals that are solid solutions can also display this behaviour. I tend to steer away from tuning cells in EVA. It would be much better if you were to use a full pattern fitting routine based on crystal structures such as TOPAS from Buker. This, as hinted at in the EVA manual, will refine the x-offset, sample displacement and all cells as well as give you microstructural measures and full quants.

Only to add a graph showing the influence of each parameter...