X-ray study of two complex oxides, each having octahedral oxygen co-ordination of transition metals

Date of Completion

January 2008


Physics, Condensed Matter|Engineering, Materials Science




We have studied the system [Ba0.6 Sr0.4] [(Y Ta) 0.03 Ti0.94]O3, as produced using two different sintering temperatures. It was shown by others that for a sample sintered at 1550 °C the material is a relaxor, whereas for the same composition produced with sintering temperature of 1600 °C the sample is a normal ferroelectric. We have employed analysis of x-ray diffraction peak broadening, Ti K edge x-ray near edge spectroscopy, and extended x-ray absorption edge fine structure spectroscopy of Ta and Y sites in our study. We find that the 1550 °C sinter sample has over double the lattice strain as does the 1600 °C sample. For the lower temperature sinter material, both Ta and Y go to sites substitutional for Ti in the lattice, with a significant expansion (contraction) of the local perovskite structure about Y/Ta dopants. Thus, with only three percent B site addition of Y and Ta dopants, there is a strain associated relaxor behavior produced in a bulk sample. For the higher temperature sinter specimen, there is a marked change in the average Y environment relative to the lower temperature sinter sample. ^ We also have studied the structural transformations around both V and Cr atoms in (V1-xCrx) 2O3 across its metal-insulator transition (MIT) have been studied by extended x-ray absorption fine-structure technique. Our new results for Cr made possible by the use of a novel x-ray analyzer that we developed reveal the substitutional mechanism of Cr doping. We find that this system has a buckled structure with short Cr-V and long V-V bonds. This system of bonds is disordered around the average trigonal lattice ascertained by x-ray diffraction. Such local distortions can result in a long range strain field that sets in around dilute Cr atoms in microscopic regions. We suggest that such locally strained regions should be insulating even at small x. The possibility of local insulating regions within a metallic phase, first suggested by Rice and Brinkman in 1972, remains unaccounted for in modern MIT theories. ^ Also for the bulk single crystals, the glancing emergent angle method was used to minimize fluorescence distortion in V K-edge fluorescence XAFS. These data were compared to transmission results on finely ground powders. The powder XAFS is expected to be closely reproduced by weighting the c perpendicular to E data twice, the c parallel to E once, and dividing by three. For V-O and V-V distances, excellent agreement was obtained between the averaged single crystal and transmission powder data. Surprisingly, the amplitude of the weighed average glancing emergent angle data was in all cases greater, not less, than the amplitude of the corresponding powder data. ^