A Perspective on the Particle-Based Crystal Growth of Ferric Oxides, Oxyhydroxides, and Hydrous Oxides

Abstract

The diversity of iron oxide, oxyhydroxide, and hydroxide materials in natural settings is remarkable. Hereafter referred to simply as the iron oxides, these materials exhibit myriad textures and morphologies, and such features provide evidence that classical growth cannot adequately explain the formation and growth of iron oxides. The iron oxides often form as a result of iron leaching from iron-containing minerals (e.g., biotite) through both abiotic and biotic weathering processes (Barker et al. 1998). Furthermore, iron oxides can form when natural water containing Fe(II) encounters oxidizing conditions (e.g., Waychunas et al. 2005). With its abundance in near-surface materials and its redox reactivity, iron plays important roles in biogeochemical cycling of a wide range of species, including metals and molecular species. The capacity of the iron oxides for sorption of metals and polyatomic anions makes these materials important players in the fate and transport of a wide range of contaminants (Waychunas et al. 2005). Elucidating how these minerals form, transform, aggregate, and grow is critical to understanding their geochemical reactivity. Particle-based crystallization has been featured prominently in the recent crystal growth literature (De Yoreo et al. 2015 and references therein). Of the diverse crystal growth mechanisms known, both classical crystal growth and particle-based crystal growth are particularly important in the iron oxides. Classical crystal growth 257 258 R.L. Penn et al. can be simply described as the monomer-by-monomer addition of molecular-scale species to a growing crystal. Oriented attachment is a special case of particle-based crystal growth and has been recognized since at least the late nineteenth century (Ivanov et al. 2014). In oriented attachment, primary particles associate to reversibly form complexes that are analogous to the outer sphere complexes described in inorganic chemistry. The primary particles composing these complexes lack direct contact, with solvent molecules and other molecular-scale species residing in the spaces separating them. The primary particles can rearrange and reorient through Brownian motion within this intermediate structure. If the primary particles achieve a common crystallographic orientation, the intermediate structures, which are sometimes referred to as mesocrystals (Cölfen and Mann 2003; Yuwono et al. 2010; Rao and Cölfen 2017, Chap. 8), can either dissociate or irreversibly bond together to form new secondary crystals. These new crystals can have symmetry-defying morphologies and contain defects like dislocations, stacking faults, and twin boundaries (Penn 2004). Numerous reviews describing crystal growth by oriented attachment have appeared in the relatively recent literature (De Yoreo et al. 2015; Ivanov et al. 2014; Xiong and Tang 2012; Dalmaschio et al. 2010; Zhang et al. 2010; Zhang et al. 2009; Niederberger and Cölfen 2006; Penn 2004), among others. These reviews provide concise descriptions of the fundamental mechanism as well as observations of oriented attachment, reporting numerous examples of oriented attachment in synthetic materials, such as titanium dioxide, iron oxides, metal selenides and sulfides, and more. In addition, evidence for oriented attachment (OA) has been observed in natural environment (Banfield et al. 2000; Hochella et al. 2008; Penn et al. 2001b). Currently, there is no universal description of how the iron oxides grow. In fact, the iron oxides literature is rife with contradictions, even when observed morphologies, textures, and microstructures are similar. An excellent example is the case of pseudocubic hematite crystals, which have been prepared by several research groups. Kandori et al. (1991) prepared synthetic hematite by aging an acidic solution of ferric chloride at 100 ı C for 20 days, and they described their product pseudocubic crystals as polycrystals composed of smaller and oriented subcrystals. They concluded that the pseudocubic hematite formed by aggregation of hydrous ferric oxide crystallites is followed by recrystallization and dehydration. Whether the hydrous ferric oxides precursor particles were oriented prior to recrystallization and dehydration was not addressed (Kandori et al. 1991). Similarly, Sugimoto et al. (1993) described their synthetic pseudocubic hematite particles, which were prepared by aging a partially neutralized solution of ferric chloride at 100 ı C for