Estimation of the proportions of clay minerals in a clay rock from XRD pattern treatment: the Arquant model

Estimating the amount of clay phases in a an argillaceous rock remains up to now a problem that is still not entirely solved in spite of the works already performed on that topic. Different models are able to calculate the XRD pattern of a clay phase, as a function of its composition like NEWMOD. The question of the amount of phases in a mixing (clay rock) is related to those models because XRD remain the main analytical tool for clay phases determination and quantification. But different approaches have been investigated there, depending on the authors. Rietveld analysis is proposed namely Hillier (2002), even if for some clay mineral phases (smectites), Rietveld refined structures are not available yet because of the poor crystallinity of theses phases. Lanson (1997) developed a deconvolution code that allow the author to follow the late stage of the smectite to illite transition, by considering the high angle domain. Improve modal calculations are also proposed by different authors, by including in the calculations, specific properties like CEC, or specific area surface. But the composition domains of many clay phases largely overlap and the values of the CEC and the specific area surface are highly variable, for a given clay mineral specie. Moore and Reynolds (1989) developed a different approach based on the intensity measurement of peaks in the low angle domain, an adapted version of the RIR method (Reference Intensity Ratio), called MIF (Mineral Intensity Factor). The method presented here is similar to the code developed by Aparicio and Ferell (2001) but the whole modeling process is simplified and more robust. It is based on a set of already simulated patterns that are combined together is order to display the closest fitting of the experimental data. However, fitting process have been reviewed, criticized and simplified. (authors)