your sample and its spectrum quality have a great impact on how reliable and trustworthy your calculation gonna be how ever if the sample is prepared homogeneously and the spectrum has taken with high step time,and permissible step 2theta angle ,then applying Rietveld analysis might be of use in your problem
This is an impractical result to obtain from XRD data alone. You need precise mass abundances for each phase with negligible error, before you can even think about the composition of the individual phases. You need to know the detail of phase structure with composition, assuming you know which end members are present in the many solid solutions that exist in most samples. I suggest you do it anothere way.
Your better option is to use XRF on the bulk sample if you wish to obtain wt% oxides.
Rietveld refinement is also a method but will not give you the percentages of oxides. It will give you the relative percentages of the mineral phases, and the oxides will be distributed among the different phases, so further calculation is needed. Since your sample is a combination of crystalline materials and disordered material (clay) it will not be an easy to obtain accurate phase composition, not impossible, just has to be done very carefully.
For some references please find some attachments, hope these will help you out: