Search

authoren We have employed a range of atomistic simulation methods to explore aspects of defect chemistry in ABO3 (where A2+= Ba2+ or Sr2+, and B4+= Zr4+ or Hf4+) perovskites, placing emphasis on processes relevant for application of these materials as high performance scintillators. Specifically, we examined intrinsic defect reactions, A and B excess nonstoichiometry and the solution of Me3+ rare earth cations. As has been predicted in previous studies, we find that Schottky disorder is the lowest energy intrinsic process. For nonstoichiometry, we predict that AO-excess is compensated by oxygen vacancies and BO2-excess is charge compensated by A vacancies (see abstract figure). Finally, for Me3+ solution, we considered several reactions for Me3+ cations ranging in size from Lu3+ to La3+, and the preferred reaction depends on the specific Me3+ cation and whether or not phase separation occurs. Nonstoichiometry mechanisms in ABO3 perovskites. Red spheres correspond to O2−, blue to A2+, and green to B4+.

Defects can decrease the efficiency of scintillators by trapping electrons. Here, point defects in REAlO3 and RE3Al5O12 are predicted with pair potential simulations, where RE is yttrium or a trivalent rare earth cation. It was found that REAlO3 shows a preference for Al2O3-excess whereas RE3Al5O12 most readily exhibits RE2O3-excess. Also, lattice volume changes for the energetically favorable intrinsic mechanisms are relatively invariant as a function of RE cation size in RE3Al5O12, but not in REAlO3. However, in non-stoichiometric RE3Al5O12, the energetically preferred disorder mechanism results in an increasing lattice expansion with increasing RE radius whereas, in non-stoichiometric perovskites, a relatively small, radius independent, lattice contraction is predicted. These results illustrate that defect behavior in REAlO3 perovskites and RE3Al5O12 garnets is quite disimilar.

It has been recently shown that the addition of Ga to Lu3Al5O12 garnet (LuAG) removes the electron trapping associated with cation antisite defects. In this paper, we show via atomic scale simulations that the replacement of Al with Ga in LuAG actually lowers the energy of the antisite disorder process. Thus, we predict that Ga additions will lead to an increase of the antisite defect concentration and, in the absence of the electronic structure changes, would actually degrade the performance of the material. Since there are two crystallographically distinct Al sites in the garnet structure, we not only present results for complete replacement of Al with Ga, but also partial replacement for 40% (on 16a) and 60% (on 24d) of the Al with Ga. Our results support the explanation for Ga-doping leading to improvement in garnet scintillator performance that relies on variations of the electronic structure rather than reduction of the antisite defect concentration. Cation antisite defects in RE3Al5O12 garnet, where small light gray atoms and large dark gray atoms are lattice Al and RE atoms, respectively. The constituents of the antisite defect pairs are labeled, and it should be noted that the Ga atom (blue) is similar in size to RE (green). Oxygen atoms are not shown for clarity.

RE 3 Al 5 O 12 garnets (where RE represents rare earth atoms Lu-Gd and Y) are technologically important, particularly for optical applications. However, the performance of these materials suffers from the existence of point defects, which are responsible for both delayed and reduced light output in Ce:Y 3 Al 5 O 12 (Ce:YAG) scintillators. The complex garnet crystal structure prevents a straightforward description of the defects responsible for decreased performance. In this paper, we employ atomistic simulation techniques to reveal non-intuitive features of the extrinsic defect structure. Specifically: Me 2+ dopants (ranging in ionic size from Mg 2+ -Ba 2+ ) are predicted to reside on RE sites with oxygen vacancies as charge compensating defects and Me 4+ dopants (from Ti 4+ -Pb 4+ ) reside on both RE and Al sites and are predicted to be charge compensated by RE vacancies. These results predict the defects resulting from common Me 2+ impurities, as well as explain why Me 4+ doping is not effective in improving RE 3 Al 5 O 12 scintillator performance. The predicted extrinsic defects in RE 3 Al 5 O 12 associated with Me 2+ (bottom left) and Me 4+ (upper right) doping. Small and large grey atoms represent Al and RE lattice atoms respectively. Oxygen atoms are not shown.

The cover picture of Rapid Research Letters refers to the article [1] by Chris Stanek et al. The image visually describes the results of calculations pertaining to the defect structure in rare earth aluminum garnet corresponding to Me2+ and Me4+ dopants. In particular, this result aids the development of garnet scintillators.

The efficiency of scintillator materials is decreased by processes associated with electrons or holes trapped at point defects. It therefore follows that minimizing the concentration of the defects responsible for trapped electrons/holes will increase the sciatillator efficiency. It has been proposed that annealing or doping oxide scintillators with aliovalent ions can change the concentration of point defects (M. Niki, phys. stat. sol. (a) 178, 595 (2000) [1]). Here, we predict the defect structures corresponding to ZrO 2 doping in a series of rare earth aluminate (REAlO 3 ) perovskites. From these calculations, we predict the mechanism leading to the decrease in oxygen vacancy concentration (the predominant electron trap in REAlO 3 ). We propose this mechanism to be an example of defect engineering, where a particularly egregious defect is traded for a less deleterious defect. These results can be used to further optimize such scintillator materials. A perfect REAlO 3 orthorhombic perovskite unit cell (a), compared to the same unit cell with the defects associated with ZrO 2 doping, namely Zr RE , Zr Al and O i (b).