Prediction of Nanoparticle Photoreactivity in Mixtures of Surface Foulants Requires Kinetic (Non-equilibrium) Adsorption Considerations.

Abstract

The adsorption of foulants on photocatalytic nanoparticles can suppress their reactivity in water treatment applications by scavenging reactive species at the photocatalyst surface, screening light, or competing for surface sites. These inhibitory effects are commonly modeled using the Langmuir-Hinshelwood model, assuming that adsorbed layer compositions follow Langmuirian (equilibrium) competitive adsorption. However, this assumption has not been evaluated in complex mixtures of foulants. This study evaluates the photoreactivity of titanium dioxide (TiO2) nanoparticles toward a target compound, phenol, in the presence of two classes of foulants ─ natural organic matter (NOM) and a protein, bovine serum albumin (BSA) ─ and mixtures of the two. Langmuir adsorption models predict that BSA should strongly influence the nanoparticle photoreactivity because of its higher adsorption affinity relative to phenol and NOM. However, model evaluation of the experimental phenol decay rates suggested that neither the phenol nor foulant surface coverages are governed by Langmuirian competitive adsorption. Rather, a reactivity model incorporating kinetic predictions of adsorbed layer compositions (favoring NOM adsorption) outperformed Langmuirian models in providing accurate, unbiased predictions of phenol degradation rates. This research emphasizes the importance of using first-principles models that account for adsorption kinetics when assumptions of equilibrium adsorption do not apply.