Fine-Scale Nanostructure in γ-Al₂O₃

Abstract

We have applied a local structural technique, atomic pair distribution function (PDF) analysis of powder diffraction data to γ-Al 2 O 3 , to obtain a quantitative structure. Refinements of the PDF support previous findings that nonspinel model representations are more suitable to describe the average structure of γ-Al 2 O 3 , as opposed to the spinel model. Surprisingly, also we find a previously unknown fine-scale nanostructure with a domain size ∼ 1 nm. Modeling suggests that within these nanodomains the oxygen sublattice is modified from the average structure and retains aspects of the boehmite precursor that reflects a stacking fault in the γ-Al 2 O 3 matrix. This results in a novel and unexpected view of the γ-Al 2 O 3 structure because earlier controversies about it centered on the arrangement of Al ions among different cation sites, whereas the oxygen sublattice arrangement was rarely questioned. This average oxygen sublattice structure is recovered in our models on longer length scales by introducing aperiodic arrays of stacking faults. The use of γ-Al 2 O 3 in catalysis depends sensitively on its nanoporosity and defect structure. Here we present a new view which may allow for additional understanding and optimization of its functional properties.

Uranyl and arsenate cosorption on aluminum oxide surface

“…The pH of the aged suspension was 4.5-5.5. X-ray diffraction (XRD) and Fourier transform infrared spectroscopy (FT-IR) showed that the c-Al 2 O 3 surface slowly transforms to a layer of a-Al(OH) 3 (gibbsite) and b-Al(OH) 3 (bayerite) mixture upon hydration, in agreement with previous studies (e.g., Wijnja and Schulthess, 1999;Arai et al, 2001;Paglia et al, 2006;Roelofs and Vogelsberger, 2006;Yang et al, 2007).…”

Direct comparison between two structural models by DFT calculations

“…Using interatomic potentials and first-principles calculations, Paglia et al [19] verified that, in average, over 40% of the cations must occupy nonspinel positions to achieve an accurate structural model while the optimization of spinel-like structural models led to less favorable energies. More recently, a debate in literature [20][21][22][23] about the inability of a nonspinel model to describe the g-alumina bulk and other publications stating just the opposite [12,19,24], leaves some doubts about which model to choose when the goal is to simulate surfaces.…”

A simple semi sol-gel method for preparation of alumina monoliths with hierarchical pore structures

“…3, showing a platy and rhombus form with dimensions 24 nm (length) Â 16.5 nm (width) Â 6.4 nm (thickness). Similar modelling for the g-Al 2 O 3 samples (D500 and D800) was not possible due to higher degree of disorder in this phase [33,34]. For these samples, the broad diffraction peaks do not only result from coherent scattering of small size domains, but also from lower crystallinity of the crystallites [35].…”