Simulation models of defects in MgAl₂O₄:Fe²⁺, Fe³⁺spinels

Abstract

A static computer simulation using the GULP code has been used to study dipole defects in iron doped spinels in normal and inverse structures. The occupation of tetrahedral (Mg2+) and octahedral (At') sites in the normal structure was simulated and the results indicate a preferential replacement of MgZ-by an AI3+ justified by the observed stoichiometric deviation in synthetic spinels. Besides this, A1zO3 excess produces Mg2+ vacancies and favours the hole centre production. Antisite disorder of the form [Mg2+]~:+ and [Al3+IM:+ is one of the possible mechanisms responsible for the existence of dipole defects in these materials. The obtained energy for the interchange between A13+ and Mg2+ was 0.52eV The calculated energies indicate that Fez+ replaces Mgz+ and Felt replaces A13' . The presence of Fe3+ in octahedral sites takes into account the oxidation process of Fez+ in these sites. Computer simulation has also been carried out on the inverse structure where the stability of the lattice was obtained for the Imma space group.

Grain‐Boundary Characterization in a Nonstoichiometric Fine‐Grained Magnesium Aluminate Spinel: Effects of Defect Segregation at the Space‐Charge Layers

“…In an equimolar synthetic spinel, due to their smallest energy increase in comparison with other calculated point defect energies, 16–20 the predominant intrinsic defects are those resulting from site exchange of cations (so‐called antisite defect pair). Furthermore, the Schottky defect has a smaller energy than either cation or anion Frenkel defects.…”

“…There are actually no reliable data or direct calculations for these formation‐free energies of intrinsic defects relative to the grain boundary. However, as the sign of the space‐charge potential is only necessary in our analysis, we have used the results of the recently calculated formation energies of these defects, relative to the lattice 17–20 . These calculations indicate the preferential replacement of Mg 2+ by Al 3+ in the normal structure (the calculated energy for Al 3+ replacing Mg 2+ is 2.3 eV and that for Mg 2+ replacing Al 3+ is 4.6 eV).…”

“…In the hyperstoichiometric spinel, MgO· n Al 2 O 3 (molar ratio n >1), excess point defects, as substitutional cations, are involved in the interchange of three Mg 2+ and two Al 3+ on tetrahedral sites. One would think that a tetrahedral vacancy should be left, as suggested by defect simulations, 17–20 but it was shown earlier from X‐ray measurements 22 that it is actually an octahedral vacancy that is more likely to form. The question of whether these charge‐compensating cation vacancies reside on octahedral sites, 22,23 tetrahedral sites, 24 or both 25 remains open.…”

“…As we have discussed previously, there are no reliable data or direct calculations for ; thus, to estimate this energy, we have used a method already described in the study of ionic double layers at the surface of Mg‐doped alumina 28,29 . From the theoretical calculations of the Schottky defect energies published by several authors 16–20 , we have determined taking the ratio and to be, respectively, equal to the lattice Al 3+ upon O 2− and Mg 2+ upon O 2− formation energy ratios. Finally, we have retained the lowest obtained value .…”

“…Thus, we confirm that the space‐charge potential in these spinels is positive, which corresponds to a negative grain‐boundary core adjacent to which defects are segregated, while and are depleted at equilibrium in the compensating positive space‐charge subgrain‐boundary region. All recent defect energy calculations in spinel show that vacancies have a greater liklihood of being created than ones or interstitials, 16–20 and the authors have quantified a significant depletion of the Mg 2+ ions at the grain‐boundary regions, rather than an increase of O 2− ions. These two factors together provide compelling proofs, for the accumulation of vacancies, rather than interstitials, at the grain‐boundary core, which contribute to its negative charge.…”