Two profiles in Miocene fluvio-lacustrine deposits consist of sandy, marly, lutitic and carbonatic levels constituted by variable percentages of quartz, calcite and clay minerals as major components, and feldspars, dolomite and occasionally gypsum and anhydrite as minor ones. The clay minerals are inherited and consist mostly of micas, with minor quantities of chlorites, pyrophyllites and kaolinites. The crystallochemical parameters of the micas indicate muscovitic compositions and their uniformity through both the different rocks and their silt and clay fractions suggest the same provenance source area, possibly located northward.
Clay minerals concentrate preferentially Li, Sc, V, Cr, Co, Ni, Cu, Zn, Rb, Cs, Ba, Zr, Hf, Th, U and REE whereas the authigenic carbonates concentrate Mn and Sr. The Sc, Cr, Th, Y, Zr and REE values in clay minerals indicate that the provenance source area of these deposits was similar in composition to the average continental upper crust, probably as a result of sedimentary recycling processes.
Zeolitic levels constituted by different proportions of analcime and smectite as major components outcrop at the top of the profiles. The analcimes show anhedral to euhedral morphologies, with grain-size ranging between 1 and 20 μm, and Si/Al ratios ranging from 2.2 to 2.5. The smectites are dioctahedral and beidellitic in composition. The zeolitic levels present significant chemical differences relative to the other ones, such as higher overall REE contents, more pronounced negative Eu anomalies and higher (La/Yb)n, Th/Sc and La/Sc ratios, suggesting a different provenance source area. Their chondrite-normalized REE patterns reflect the possibility that the starting materials were pyroclastic eruptive rocks originating from intracrustal partial melting. The variable analcime and smectite percentages are attributed to variations in H+/(Na+ + K+) and K+/(Na + Ca2+ + Mg2+) activity ratios and silica and water activities in the pore-waters during diagenetic processes.