The nature and timing of diagenetic reactions in marine sandstone/shale formations were modelled for a 50-120 degrees C temperature range, assuming at first chemical reactions in closed systems, in order to have a better understanding of the factors that control the illitization reaction when K-feldspar coexists with aluminous clay. The original sediment is constituted of quartz, muscovite, K-feldspar, Al-clays (kaolinite, beidellite, montmorillonite), with and without organic matter maturation. We compared the stable mineral parageneses predicted with those observed in natural sandstones and shales. We also tested the effects of an energy barrier on illite growth by allowing or not allowing the muscovite/illite precipitation reaction to occur and by using several illitization reaction rates. When compared with data for natural mineral assemblages, the results suggested that the illitization reaction depends on the nature of the reacting clays. Kaolinite conversion to end member illite involves high energy conditions which are not met when the pore water equilibrates with the mineral matrix from undersaturated conditions. To overcome this barrier, the fluid should be oversaturated with respect to the K-feldspar. An external source of potassium or a pH increase appears to be the most possible driving force in such reactions. Kaolinite conversion to end member illite appears then as a marker of fluid circulations and does not affect the K-feldspars when present. We also discuss the composition of the end member illite as an alternative explanation for the metastability oi the quartz-potassic feldspar-kaolinite assemblage. On the other hand, a smectite-illite conversion involves more limited energy requirements and the reaction progresses spontaneously in closed systems (as observed in most shales) by dissolving K-feldspars (the source oi potassium) and producing quartz overgrowth. The aqueous species of organic origin do not affect these reactions significantly.