The role of diffusion has generally been underestimated or ignored in sedimentary geochemistry, particularly diagenesis. For instance, reactions in shale have been suggested as a source of cements (carbonate, quartz and clay) and secondary porosity agents for adjacent sandstone intervals. However, most proposed models have used advective transport formulations that rarely provide sufficient fluid volumes to explain the observed mass transfer. Models based on diffusive transport can overcome these limitations. Given that shale and sandstone have different initial amounts of reactive detrital components such as organic matter, smectitic clays, biogenic silica and carbonate, mass transfer can result as the unstable components are heated past a threshold temperature during burial and the system becomes transport-limited. Units with more reactive component produce higher solute concentration per unit time, resulting in chemical gradients that transfer mass between the adjacent units. The process forms chemical gradients that can rapidly transfer mass equivalent to several volume % of the rock over distances of at least 5 meters. The mass transfer continues until the reactants are depleted and pore water chemistry is re-stabilized. The final product is authigenic mineral distributions related to the contact between units. The threshold temperature, where the local rate of surface reaction is greater than the local rate of transport, is different for each unstable component. Burial of units with multiple reactive components can produce sequential diffusion gradients throughout the burial history and may produce multiple generations of cement. Numerous examples of contact-related patterns of diagenetic carbonate, quartz and clay cements, secondary porosity and potassium depletion all support the diffusive model, rather than advective formulations.