Nanoparticles in natural systems I: The effective reactive surface area of the natural oxide fraction in field samples

Abstract

Information on the particle size and reactive surface area of natural samples is essential for the application of surface complexation models (SCM) to predict bioavailability, toxicity, and transport of elements in the natural environment. In addition, this information will be of great help to enlighten views on the formation, stability, and structure of nanoparticle associations of natural organic matter (NOM) and natural oxide particles. Phosphate is proposed as a natively present probe ion to derive the effective reactive surface area of natural samples. In the suggested method, natural samples are equilibrated (⩾10 days) with 0.5 M NaHCO 3 (pH = 8.5) at various solid–solution ratios. This matrix fixes the pH and ionic strength, suppresses the influence of Ca 2+ and Mg 2+ ions by precipitation these in solid carbonates, and removes NOM due to the addition of activated carbon in excess, collectively leading to the dominance of the PO 4–CO 3 interaction in the system. The data have been interpreted with the charge distribution (CD) model, calibrated for goethite, and the analysis results in an effective reactive surface area (SA) and a reversibly bound phosphate loading Γ for a series of top soils. The oxidic SA varies between about 3–30 m 2/g sample for a large series of representative agricultural top soils. Scaling of our data to the total iron and aluminum oxide content (dithionite–citrate–bicarbonate extractable), results in the specific surface area between about 200–1200 m 2/g oxide for most soils, i.e. the oxide particles are nano-sized with an equivalent diameter in the order of ∼1–10 nm if considered as non-porous spheres. For the top soils, the effective surface area and the soil organic carbon fraction are strongly correlated. The oxide particles are embedded in a matrix of organic carbon (OC), equivalent to ∼1.4 ± 0.2 mg OC/m 2 oxide for many soils of the collection, forming a NOM–mineral nanoparticle association with an average NOM volume fraction of ∼80%. The average mass density of such a NOM–mineral association is ∼1700 ± 100 kg/m 3 (i.e. high-density NOM). The amount of reversibly bound phosphate is rather close to the amount of phosphate that is extractable with oxalate. The phosphate loading varies remarkably ( Γ ≈ 1–3 μmol/m 2 oxide) in the samples. As discussed in part II of this paper series ( Hiemstra et al., 2010), the phosphate loading ( Γ) of field samples is suppressed by surface complexation of NOM, where hydrophilic, fulvic, and humic acids act as a competitor for (an)ions via site competition and electrostatic interaction.