APW+lo and TDPAC study of the Electric-Field Gradient at the cation sites of the (44Ti(EC)-->)44Sc-doped Sc2O3 Semiconductor

D. RICHARD, E. L. MUÑOZ, T. BUTZ, L. A. ERRICO, AND M. RENTERÍA

Ginebra

Conferencia; HFI/NQI2010, 15th International Conference on Hyperfine Interactions & 19th International Symposium on Nuclear Quadrupole Interaction; 2010

Executive Committee del 3th HFI/NQI International Conference

We report on an ab-initio study of the Electric-Field-Gradient tensor (EFG) at bothinequivalent Sc sites in the semiconductor Sc2O3. This semiconductor crystallizes inthe bixbyite structure with two six-fold coordinated cation sites called C and D. Thefirst is highly asymmetric whereas the second is axially symmetric. The calculationswere performed applying the Full-Potential Augmented-Plane Waves plus local orbitals(FP-APW+lo) method, in the framework of the Density Functional Theory (DFT), thatallows us to treat the electronic structure and the atomic structural positionrefinements in a fully self-consistent way. Our results are compared withexperimental data determined by Time-Differential gamma-gamma Perturbed-AngularCorrelation (TDPAC) spectroscopy using the Leipzig 6-detector TDPAC spectrometer withLaBr3(Ce) scintillators with the first excited I=1- state of the 44Ti(EC)->44Scisotope as radioactive tracer. There is excellent agreement between experiment andthe present ab-initio calculations. It is clear from the comparison of theexperimental electric-field gradients and the Point-Charge Model (PCM) that the PCMcan not describe even approximately the measured electric-field gradients at cationsites in pure scandium sesquioxide. In this simple situation, where the 44Sc probeatom is not an impurity in the material under study, the tracer does not introducestructural distortions that are usually not taken into account in the PCM when animpurity is concerned and that does not introduce impurity levels in the band gap ofthe semiconductor, which are usually critical for the origin of the electric-fieldgradient. Nevertheless, in this simple case, the PCM seems to fail. This can be onlydue to a poor description of the electronic distribution around the probe atom, whichis not taken into account with the Sternheimer antishielding factor that is proposedin the PCM to describe the polarization of the core electrons of the probe atom. Wealso found an experimental site preference for the 44Ti impurity which can beunderstood by performing ab-initio calculations for the impurity system using supercells.