I have wide experience in XRD data analysis. In particular, my expertise goes from analytical methods to data processing using different software. I am interested in your proposal, so I would like to know what kind of analysis you are particularly interested in.
I have done extensive XRD Analysis on carbonates, if composition of clays you are after then it is easy. Check my master thesis, its on Research gate, titled: "Characterization of Clay Minerals and Their Impact on Reservoir Permeability"
A little skepticism. The critical diffraction line of graphite (002) coincides with the main quartz line. Well crystallized graphite has a strong tendency to textureation, poorly crystallized graphite, i.e. graphitic carbon has a very wide line that disappears in the background. These are pitfalls and sources of error. I measured many samples of black roofing slate and found no detectable graphite. But I use Bruker Topas.
Thank you all for your response especially Dr. Dalibor for his invaluable advice.
The followings are the key points (out of our email discussions) from Dr. Dalibor which is surely useful to the other researchers working on the XRD of shales:
I would just like to point out that at low levels it is not possible to determine some components completely reliably. These are mainly the basicity of plagioclase (albite or oligoclase) and polytype of mica. But it is definitely not slate and graphite is not there.
The fact that Rietveld analysis gives positive results for the contents of a phase does not necessarily mean that that phase is really present in the sample. The presence must first be demonstrated by qualitative analysis. Preferably by the presence of a peak that is unambiguous and does not overlap with anything. In the case of graphite and quartz, the problem is that graphite has very few lines. The main line (002) for Graphite-2H and -3R too overlaps with the quartz main line (101). Rietveld analysis here tends to interpret quartz peak residues as graphite. There is not much alternative here to reliably detect graphite. What is calculated as graphite is most likely a residue of the shape function of the quartz peak.
Another remark is that it is not realistic to specify graphite-2H and 3R at the same time. The deviations between them are only in very weak non-basal lines, which are almost never observable, even on conventional preparations of pure graphite. Preferred orientation is a major problem for graphite.
Particular graphite is separable by flotation with a non-polar organic substance. It is also possible to do this in the laboratory in a measuring cylinder. The finely ground sample is dispersed in water in a cylinder, a little ether or toluene is added and dispersed again. The graphite then accumulates in the organic matter on the surface of the water. Furthermore, dolomite and ankerite are isotypic. They are indistinguishable by powder diffraction analysis at low contents. In addition, according to the valid nomenclature, ankerite Fe0.2 is today iron dolomite. The only alternative is SEM / WDS microanalysis. In high contents, it can be oriented according to the cell parameters. The feldspar problem is similar. At high contents (> 30-50%) it is possible to orient according to the cell parameters, but at low contents it is simply plagioclase. But something must be added to the calculation.
Elevated background in the area below 20 deg. 2theta can have several causes. First, I hope that the knife slit, if used, was in the same position. Another reason may be the content of amorphous substances, both organic and, for example, some clay minerals (mixed layered, non-specific), organo-clay complexes, etc.
In Rietveld analysis, the most important thing is to test what the resulting model does with the data. At high magnification, the resulting image is scanned and the difference and refined region of the data is evaluated.