ABSTRACT
The sorption of Co(II) on colloidal hematite was studied as a function of pH, ionic strength, and Co(II) concentration. Two different techniques were used, yielding two different sets of information: (i) potentiometric titrations that provide information on the number of protons released as a function of pH owing to the sorption of Co(II) and (ii) measurement of the amount of cobalt sorbed on the surface as a function of pH using a radioactive tracer, (60)Co. At low Co(II) concentrations (10(-8) M), the sorption was found to be independent of ionic strength but there seems to be a weak ionic strength dependence at higher Co(II) concentrations (10(-4) M). The adsorption edge moved to higher pH with increasing Co(II) concentration. For the high Co(II) concentration, the number of protons released per cobalt sorbed increased from zero to approximately 1.5. The basic charging properties of hematite were modeled with four different surface complexation models. The 1-pK Basic Stern Model (BSM), with binding of electrolyte ions to the Stern plane, seems to be the most reasonable model if the ambition is to describe experimental data at different ionic strengths. The sorption of cobalt was modeled with the 1-pK BSM. By introducing a low concentration of high affinity surface sites for cobalt sorption it was possible to model the sorption in very wide cobalt concentrations, ranging from 10(-8) M to 10(-4) M. Copyright 2000 Academic Press.
ABSTRACT
The sorption and desorption of Th onto TiO(2) (10 g/L) was studied as a function of pH (1-11), ionic strength (0.005-0.1 M NaCl, NaClO(4)), thorium concentration (from 1 x 10(-9) to 3 x 10(-2) M), and carbonate concentration (up to 10(-2) M) using (234)Th tracer. The results indicate the reversible formation of an inner sphere complex with a strong pH dependence. No effect of the carbonate on sorption could be measured in the pH range investigated (5-10.5). The data were fitted with five different conceptual models (constant capacitance, diffuse layer, Stern, triple layer, and a nonelectrostatic model) using FITEQL. Different hydroxy complex surface reactions were tried, one at a time, varying the site density (1-12 sites/nm(2)). The reaction that provided the best fit depended strongly on the surface site density and the model used. To be able to fit a reaction to the pH-dependent data with the compact layer models, the effect of the electrostatic term had to be decreased by decreasing the site density below 5 sites/nm(2). None of the electrostatic models could be used to fit all of the isotherm data. The nonelectrostatic model could be used to reasonably fit both the pH-dependent and isotherm data. Copyright 1999 Academic Press.
