We present a transport-reaction model (TRACTION) specifically designed to account for non-ideal transport effects in the presence of thermodynamic (e.g. salinity or temperature) gradients. The model relies on the most fundamental concept of solute diffusion, which states that the chemical potential gradient (Maxwell’s model) rather than the concentration gradient (Fick’s law) is the driving force for diffusion. In turn, this requires accounting for species interactions by applying Pitzer’s method to derive species chemical potentials and Onsager coefficients instead of using the classical diffusion coefficients. Electrical imbalances arising from varying diffusive fluxes in multicomponent systems, like seawater, are avoided by applying an electrostatic gradient as an additional transport contribution.
We apply the model to pore water data derived from the seawater mixing zone at the submarine Mercator mud volcano in the Gulf of Cadiz. Two features are particularly striking at this site: (i) Ascending halite-saturated fluids create strong salinity (NaCl) gradients in the seawater mixing zone that result in marked chemical activity, and thus chemical potential gradients. The model predicts strong transport-driven deviations from the mixing profile derived from the commonly used Fick’s diffusion model, and is capable of matching well with the profile shapes observed in the pore water concentration data. Even better agreement to the observed data is achieved when ion pairs are transported separately. (ii) The formation of authigenic gypsum (several wt%) occurs in the surface sediments, which is typically restricted to evaporitic surface processes. Very little is known about the gypsum paragenesis in the subseafloor and we first present possible controls on gypsum solubility, such as pressure, temperature, and salinity (pTS), as well as the common ion and ion pairing effects. Due to leaching of deep diapiric salt, rising fluids of the MMV are saturated with respect to gypsum (as well as celestite and barite). Several processes that could drive these fluids towards gypsum supersaturation and hence precipitation were postulated and numerically quantified. In line with the varied morphology of the observed gypsum crystals, gypsum paragenesis at the MMV is likely a combination of two temperature-related processes. Gypsum solubility increases with increasing temperature, especially in strong electrolyte solutions and the first mechanism involves the cooling of saturated fluids along the geothermal gradient during their ascent. Secondly, local temperature changes, i.e. cooling during the transition from MMV activity towards dormancy results in the cyclic build-up of gypsum.
The model showed that the interpretation of field data can be majorly misguided when ignoring non-ideal effects in extreme diagenetic settings. While at first glance the pore water profiles at the Mercator mud volcano would indicate strong reactive influences in the seawater mixing zone, our model shows that the observed species distributions are in fact primarily transport-controlled. The model results for SO4 are particularly intriguing, as SO4 is shown to diffuse into the sediment along its increasing (!) concentration gradient. Also, a pronounced gypsum saturation peak can be observed in the seawater mixing zone. This peak is not related to the dissolution of gypsum but is simply a result of the non-ideal transport forces acting on the activity profile of SO4 and Ca profiles.