The role of surface functionality on the adsorption of divalent metal ions

The toxicity of heavy metals has been extensively studied and is reasonably well understood. Heavy metals have been recognised for their toxicological effects on humans and other animal life. As a means of remediation, adsorption of heavy metals onto solid substrates has been extensively studied, but is perhaps not so well understood. Metal adsorption involves the addition of a solid material to an aqueous metal solution, creating a suspension. Metal species bind on to the solid surface, decreasing the aqueous metal concentration. The resultant solid can then be removed by sedimentation, filtration or flotation. In order to observe metal adsorption behaviour, investigate the mechanisms of adsorption and to better understand which adsorbent and metal properties influence the extent of metal removed from solution, a suite of seven different metals (Ca(II), Cd(II), Cu(II), Mg(II), Ni(II), Pb(II), Zn(II)) was adsorbed onto a series of substrates, including six metal oxides (Al2O3, two different samples of Fe2O3, SiO2 and two different samples of TiO2) and three different amphoteric polystyrene latices. In each case, the adsorption took place under constant conditions of sorbent:sorbate ratio, temperature, exposure time and electrolyte conditions, producing one of the largest systematic studies of metal adsorption to date. In most cases, adsorption was studied as a function of pH. The adsorbents were characterised according to their surface charge, zeta potential, morphology, particle size, surface area and dielectric constant and used in conjunction with both surface complexation and thermodynamic modelling in order to determine the experimental parameters that have a substantial effect on metal adsorption behaviour. The TiO2 adsorbents were found to be the most efficient at removing metals from solution, followed by the Fe2O3 adsorbents, then SiO2 and finally Al2O3. The degree to which the amphoteric polystyrene latices adsorbed metals could be altered by varying the proportion of carboxyl functional groups present on the adsorbing surface. It has been postulated that using an adsorbent with a large dielectric constant improves metal adsorption and results in less variation between different metals adsorbing onto a particular surface – a prediction based on the importance of solvation barriers to adsorption (James & Healy, 1972c). The TiO2 adsorbents – specifically chosen for their high dielectric constants – showed an enhanced adsorption ability possibly attributable to their low iso-electric point (IEP) at a pH of approximately 4.4, or another undetermined property. However, the enhanced adsorption ability was found to not directly correlate with their high dielectric constant. A high correlation was observed between the pH at which 50% of the metal was removed from solution (pH50) and the first hydrolysis constant for the metal (pK1) for metal adsorption onto the metal oxide surfaces. In addition to this, it was found that a high correlation was observed between the pH50 and pKc (equilibrium constant describing metal-carboxyl affinity) for metal adsorption onto the polystyrene latex substrates. It was discovered that the TiO2 adsorbents did not show this strong correlation. This observation will need to be the subject of future study. A study of the specific surface area, particle size and surface charge revealed that the Brunauer Emmett Teller (BET) gas adsorption method of surface area determination under-estimated the available surface area in an aqueous environment. The extent of underestimation observed in this study could not be linked to any universal attribute, but may, as suggested in the literature, be due to changes in the surface structure, such as the formation of 'hairy' particle surfaces (Allison, 2009). For all metals adsorbing onto each of the nine substrates, a good correlation was achieved between surface complexation modelled equilibrium constants and the experimental adsorption data. In all cases, only one reaction was required to model the experimental adsorption isotherms; [] where X represents the solid surface. A strong correlation was observed between KX-_MOH* (the adsorption equilibrium X −MOH constant) and the pK1 of each metal species, when adsorbing onto metal oxide surfaces. Similarly, a strong correlation was observed between K COO-_MOH* (the adsorption equilibrium constant) and the pKc of each metal species, when adsorbing onto the amphoteric polystyrene latices. These observations indicate that there is a strong relationship between a metal’s affinity for a particular ligand in solution and it’s affinity for that same ligand present as part of an adsorbing surface. The experimental data was modelled using three different thermodynamic models – the James Healy model (James & Healy, 1972c), the James-Healy-Levine model (Lyklema et al., 1971) and the Agashe Regalbuto model (Agashe & Regalbuto, 1997). For most adsorption isotherms, a good correlation was achieved between the experimental and modelled adsorption data. The thermodynamic modelling of the adsorption process showed that a strong correlation typically existed between ΔGOchem and pKX, indicating that the ΔGOchem term accounts for the variation in a metal’s affinity for the adsorbing surface functional groups.