There is emerging evidence that growth of synthetic and natural phases occurs by the aggregation of prenucleation clusters, rather than classical atom-by-atom growth. Ferrihydrite, an iron oxyhydroxide mineral, is the common form of Fe3+ in soils and is also in the ferritin protein. We isolated a 10 angstrom discrete iron-oxo cluster (known as the Keggin ion, Fe13) that has the same structural features as ferrihydrite. The stabilization and manipulation of this highly reactive polyanion in water is controlled exclusively by its counterions. Upon dissolution of Fe13 in water with precipitation of its protecting Bi3+-counterions, it rapidly aggregates to ~22 angstrom spherical ferrihydrite nanoparticles. Fe13 may therefore also be a prenucleation cluster for ferrihydrite formation in natural systems, including by microbial and cellular processes.

Iron oxides and oxyhydroxides are ubiquitous in the environment and serve vital roles in interrelated phenomena of contaminant transport, pH control of surface and ground water, and microbial activity (1–4). Relevant phases include hematite, magnetite, goethite, and ferrihydrite. The structure of ferrihydrite, the most common iron oxyhydroxide in soil and in the core of the ferritin protein, is isostructural to the Al-mineral akdalaite (5). The ferrihydrite structure has been highly debated (6–9), but the general framework contains linked and fused iron-oxyhydroxide Keggin units. The Keggin ion or molecule is a metal-oxo structural motif in natural and synthetic materials (10). Even before the proposed structure of ferrihydrite, an iron Keggin cluster (henceforth referred to as Fe13) was presumed to be synthetically attainable—analogous to the Al13-Keggin cluster [AlO4Al12(OH)24(H2O)12]7+, which was first crystallized and structurally characterized over 50 years ago. (11)

The identified nonclassical growth behavior of iron oxides in both nature (3) and the laboratory (12)—defined by the aggregation of prenucleation clusters rather than atom-by-atom growth—supports the existence of a discrete Fe13 ion as a precursor to ferrihydrite and magnetite (Fig. 1). However, the higher reactivity (acidity) of Fe3+-bound H2O as compared with Al3+-bound H2O in the Al13 Keggin ion has thus far thwarted all synthetic efforts to capture a discrete Fe13 ion from water. Instead, iron oxyhydroxide nanoparticles and precipitates are obtained, bypassing the intermediate discrete cluster state. The closest specie to Fe13 reported thus far is [FeO4Fe12F24(OCH3)12]5– (13). Although possessing the Keggin structure, this cluster was synthesized in anhydrous conditions and is surface-passivated entirely with nonaqua ligands; it is neither a likely precursor for iron oxide nucleation, nor provides identification of the aqua ligands and ion-charge of Fe13 derived from water.

Science

1359

03/2015

http://m.sciencemag.org/content/347/6228/1359.full