RESUMO
Understanding calcium silicate hydrates (CSHs) is of paramount importance for understanding the behavior of cement materials because they control most of the properties of these man-made materials. The atomic scale water content and structure have a major influence on their properties, as is analogous with clay minerals, and we should assess these. Here, we used a multiple analytical approach to quantify water distribution in CSH samples and to determine the relative proportions of water sorbed on external and internal (interlayer) surfaces. Water vapor isotherms were used to explain the water distribution in the CSH microstructure. As with many layered compounds, CSHs have external and internal (interlayer) surfaces displaying multilayer adsorption of water molecules on external surfaces owing to the hydrophilic surfaces. Interlayer water was also quantified from water vapor isotherm, X-ray diffraction (XRD), and thermal gravimetric analyses (TGA) data, displaying nonreversible swelling/shrinkage behavior in response to drying/rewetting cycles. From this quantification and balance of water distribution, we were able to explain most of the widely dispersed data already published according to the various relative humidity (RH) conditions and measurement techniques. Stoichiometric formulas were proposed for the different CSH samples analyzed (0.6 < Ca/Si < 1.6), considering the interlayer water contribution.
RESUMO
Na/Ca ion-exchange isotherms were performed on a Na-saturated montmorillonite for different constant normalities of the aqueous chloride solution and at two solid/solution ratios. The experimental data suggest that the affinity of Na(+) for the solid increases with total normality and m/V ratio and that a significant proportion of CaCl(+) is also sorbed. A thermodynamic modeling procedure with one sorption site and three sorbed cations (Na(+), Ca(2+) and CaCl(+)) was applied to interpret the data. We show that by accounting for the activities of aqueous species, the experimental data obtained at different total normalities for a given solid/solution ratio can be fitted using a unique set of selectivity coefficient values. However, when the m/V ratio is decreased from 25 to 2.5 g/L, an increase in the log Kc(Na(+)/Ca(2+)) of up to 0.4 ± 0.05 was required to interpret the data with a constant log Kc(Na(+)/CaCl(+)) value. The same behavior concerning the increase in the log Kc(Na(+)/Ca(2+)) with a m/V ratio decrease was observed when using a multi-site model taking into account the specific sorption of H(+). The results clearly indicate that the Na(+)/Ca(2+) selectivity coefficients strongly depend on the solid/solution ratio but are independent of the exchanger composition and the total normality of the solution. Such findings provide possible pitfalls when using selectivity coefficients obtained in diluted suspension to assess the sorption in compacted clayrocks and allow a better prediction of the sorption of trace elements in competition with major cations.
RESUMO
The aim of the present paper is to clarify previous results showing that selectivity coefficients determined for the exchange of Na(+) for Ca(2+) in montmorillonite were dependent on the solid/solution ratio. The organization of montmorillonite suspensions upon Na(+)/Ca(II) exchange was analyzed by combining optical microscopy, small-angle X-ray scattering and X-ray diffraction. All samples displayed flocculated characteristics, eliminating the possibility of contrasting accessibility of sorption sites with the solid/solution ratio. Modeling of experimental X-ray diffraction patterns was used to quantify the relative proportions of interlayer Ca(2+) and Na(+) cations along the exchange isotherm. The results further confirmed the influence of the solid/solution ratio on the degree of interlayer Ca(II)-for-Na(+) exchange, and specific selectivity coefficients for interlayer sites were determined. The effect of the solid/solution ratio was finally interpreted by the resulting local changes in the solution chemistry. We demonstrated that by accounting for the Donnan effect, the different data can be interpreted using a single selectivity coefficient. The obtained Kc constant was successfully applied to interpret existing hydrogeochemical data on a natural aquitard. This most likely represents a more constrained and valid approach for the modeling of reactive element transport in natural media than does the poorly defined Kd parameter.