Difusión de metales tóxicos en disolución a través de "composites" mineral-hidrogeluna simulación en el laboratorio de interacciones fluido-mineral

Laburpena

The interaction of dissolved toxic metals with mineral surfaces influences their transport behaviour in soils and groundwater systems. Sorption, defined generally as a surface localisation irrespective of mechanism, represents a potentially significant way of immobilisation of toxic metals in natural porous media. Sorption kinetics in mineral-fluid systems is typically studied by reacting single-phase powders with aqueous solutions in various types of reactors. However, because the amount of material sorbed is small, unambiguous characterisation of the sorbed phase may be extremely tedious, if not impossible. Here we present an experimental study of the transport of various metals through artificial porous media consisting of mineral grains embedded in a matrix of silica hydrogel. The gel suppresses convection and advection, only allowing diffusion of the aqueous ions, which occasionally can be sorbed on the surface of the embedded minerals. The low mobility of aqueous species in the gel involves a drastic decrease in number but a significant increase in size of the sorbed entities. This increase allows characterising the sorbed phase and its crystallographic relation to the substrate. The experimental device consists of a typical U-tube arrangement in which two reservoirs are separated by a diffusion column. The horizontal branch of the U-tube was completely filled with mineral grains immersed in a sodium silicate aqueous-solution acidified with HCl 1N to a pH of 5.5. After some time, this solution polymerises to form a gel that agglutinates the grains. The silica hydrogel is a porous medium with a sheet-like structure that forms interconnecting cells. The water content in this kind of gel is about 95.6 wt% and the gel pores have diameters from less than 0.1 µm up to 0.5 µm. This structure allows the gel to be used as an aqueous diffusion medium where convection and advection are suppressed. Four different kinds of composite were used: calcite-gel, barite-gel, aragonite-gel, and quartz-gel. Calcite grains were prepared from optical-quality Iceland spar and are essentially cleavage fragments with only {101-4} faces exposed. Barite (BaSO4) grains are subhedral fragments dominated by cleavage surfaces. Finally, aragonite and quartz grains are, respectively, subhedral and anhedral fragments. In all cases grains with diameters ranging between 1 and 1.5 mm were selected. Diffusion experiments were carried out at 25 ¿C by filling the 'source reservoir' with an aqueous solution of the toxic metal (Cd, Cr(VI), or Pb), and the 'discharge reservoir' with deionized water. When the experiment starts, the diffusion column is free of contaminant, but gradients of concentration develop as diffusion time passes by. As a consequence, the pollutant concentration decreases in the source reservoir and increases in the discharge reservoir with diffusion time. This evolution was quantified by terminating diffusion experiments at fixed time intervals, removing the solutions from the reservoirs, and analysing them for the diffusing metal. Chemical analyses were carried out by inductively coupled plasma mass spectrometry. The concentration-time data series obtained in this way can be used to determine 'apparent' diffusion coefficients Da of the pollutants. After diffusion, the grains can be cleanly released from the matrix to characterise their surface. With this aim, back-scattered electron images and dispersive-energy microanalyses of the grain surfaces were obtained in a scanning electron microscope. In some cases the sorbate was characterized by glancing incidence X-ray techniques. The surface of the barite grains, observed after diffusion of CrO42- ions through a barite-gel composite, evidences clear dissolution signs and is covered by innumerable small crystals. These crystallites show the typical morphologies of chromium-rich Ba(CrO4,SO4) crystals grown in gels. Energy dispersive microanalyses yield compositions around BaCr0.89S0.11O4. Nucleation occurs in a non-random orientation with a strict parallelism between the crystallographic directions of both substrate and overgrowing crystals. This is to be expected if one considers that barite and barium chromate form a nearly ideal solid-solution series, where both end-members are isomorphous and crystallize in the same space group. Diffusion of Pb through barite-gel composites produces analogous outcomes: sorption occurs by precipitation of (Pb,Ba)SO4 crystallites on the barite surface. The nuclei are oriented according to the crystallographic directions of the isoestructural barite and show a moderately high Pb-content (»Pb0.75Ba0.25SO4). Similarly, the uptake of Cd2+ by calcite occurs via epitaxial overgrowth of lamellar crystallites of a (Cd,Ca)CO3 solid solution. These platelets are very Cd-rich, with compositions around Cd0.965Ca0.035CO3. According to the previous results, the overgrowth of a (Me,B)A solid solution seems to be the most effective mechanism of sorption, where the sorbed metal (Me) substitute for B in the substrate structure. In the present scenario, such a mechanism involves (1) the arrival of diffusing ions near the mineral surface, (2) the release of solute from the mineral surface to the pore fluid, and (3) the reaction between the released solute and the diffusing metal to form solid-solution nuclei. In summary, the present experiments appear to be an excellent tool to investigate whether metal sorption on mineral surfaces takes place by a true adsorption process or by surface precipitation of a metal-bearing solid. We find that long-term Cd2+ uptake by calcite occurs via formation of adherent precipitates of (Cd,Ca)CO3 solid solutions. Similarly, barite dissolution coexists with surface precipitation of Ba(CrO4,SO4) or (Pb,Ba)SO4, resulting in a sort of ion exchange between substrate and fluid. The influence of these solid-solution processes on the fate and transport of metals in natural environments could be determinant. Anyway, neglecting to consider such solid solution formation will lead to overestimates the availability and mobility of metals in the environment.