Particle morphological and roughness controls on mineral surface charge development

Abstract

Effects of mineral particle morphology and roughness on potential determining ion (p.d.i.; H+, OH−) loadings achieved at synthetic lepidocrocite (γ-FeOOH) surfaces were predominantly investigated by potentiometry and thermodynamic modeling. Nanosized rod- (RL) and lath-shaped (LL) particles exhibiting different proportions of the same predominant crystallographic faces acquired largely comparable pH, ionic strength and counterion (NaCl, NaClO4) dependencies on p.d.i. loadings. These results supported previous claims that faces ideally containing proton silent sites only, are likely populated by additional proton active sites. This concept was supported further by results of roughened LL-like particles (LLR) also showing highly congruent pH-, ionic strength- and composition-dependent p.d.i. loadings with those of LL and RL. These loadings thereby correspond to maximal levels allowed by net attractive and repulsive forces at each solution composition, irrespective of particle morphology. Contrasting equilibration times required to achieve these loadings revealed considerably slower exchange of p.d.i. and electrolyte ions near the point of zero charge in the rough LLR than in the more idealized LL and RL particles.Thermodynamic modeling was used to test various concepts accounting for these results. The model made use of a novel framework capable of isolating electrostatic contributions from different faces, and of accounting for ion-specific double-layer properties within a single crystallographic face. These efforts made use of capacitance values for each electrolyte ions within the framework of a recently developed Variable Capacitance Model. Attempts at modeling all three particle types were used to suggest that the (010) face contains ∼0.9sitenm−2 of proton active sites, a value notably constrained by recently published Na+, Cl−, and ClO4− loadings derived by cryogenic X-ray photoelectron spectroscopy. The model presented in this work thus provides a means to predict p.d.i. loadings on multifaceted mineral particle surfaces, and can therefore be used to constrain further our understanding of mineral/water interface reactivity.