Structure determination and refinement of three novel ternary phases in the Bi-Te-Ti-oxide system

Meden¹, A. Kocevar¹, M. Udovic², M. Valant², D. Suvorov²

¹Univ. of Ljubljana, Fac. of Chemistry & Chem. Tech., Askerceva 5, SI-1000 Ljubljana, Slovenia.

²”Jozef Stefan” Institute, Jamova 39, Si-1000, Ljubljana, Slovenia.

The ternary phase diagram of Bi₂O₃ – TeO₂ – TiO₂ was explored in order to locate possible candidates for use in the microwave wireless communication technology. Along with the desired microwave characteristics (appropriate dielectric constant, low dielectric loss, negligible temperature dependence of the resonant frequency), the possibility of sintering these materials at low temperatures was expected. This feature makes possible the use of the LTCC (Low Temperature Cofired Ceramics) technology where the whole module consisting of different ceramic electric elements (resistors, capacitors, conducting paths…) together with silver electrodes is assembled in green (pressed powder) state and sintered only once at the end.

All the phases of the corresponding binary systems, stable at room temperature in an oxygen atmosphere, were already known (four in Bi₂O₃ – TeO₂, four in Bi₂O₃ – TiO₂ and one in TeO₂ – TiO₂). This was, however, not the case for three ternary single phase materials with the expected compositions of 1Bi₂O₃-1TeO₂-1TiO₂ (phase 1), 1Bi₂O₃-1TeO₂-3TiO₂ (phase 2) and 2.5Bi₂O₃-1TeO₂-4TiO₂ (phase 3) as determined by electron microscopy and X-ray powder diffraction. For none of them it was possible to find any structure with similar powder diffraction pattern in the PDF-2, so they were considered new structure types and subjected to an ab-intio SDPD.

Indexing with CRYSFIRE suite was successful in all three cases and eventually helped to subsequently purify the phase 1 (eliminate unindexed peaks). The unit cells were: monoclinic a=13.1 Ǻ, b=5.00 Ǻ, c=5.25 Ǻ and b=108 ^o for phase 1, trigonal or hexagonal a=5.17 Ǻ and c=4.95 Ǻ for phase 2 and cubic a=9.40 Ǻ for phase 3.

At this point it was found by other methods, that during the synthesis in air, Te⁴⁺ was oxidized to Te⁶⁺, so that TeO₃ should be expected in the above compositions. This means more oxygen and smaller Te⁶⁺ ion which can share the crystallographic sites with Ti⁴⁺. This idea was used to put (Te,Ti)O₆ octahedra (Te : Ti ratio defined from the composition) and free Bi atoms into FOX. The number of each was varied to maintain the expected compositions, enable the occupation of special positions and satisfy the expected density (unit cell contents) for a selected space group. Several space groups were tested in each case and FOX gave reasonable solutions for all three phases. After the comparison of different solutions, the models in space groups P21/a, P-3 and Pn3 for phases 1, 2 and 3 respectively, were selected for the refinement.

Rietveld refinement using TOPAS was rather straightforward giving R_wp values between 10 and 12 % on the data from Bruker D4 diffractometer, Bragg-Brentano, CuKa up to 2q of 140 ^o. Interatomic distances and coordinations were reasonable without using restraints. In all three cases it was confirmed that the Te and Ti atoms occupy the same sites and the refined population parameters were very close to the ones expected from the composition. The number of oxygen atoms in the structural models confirms the oxidation of Te⁴⁺ to Te⁶⁺ without doubt in the phases 1 and 2, which are octahedral frameworks consisting of edge-sharing BiO₆ (larger) and (Te,Ti)O₆ (smaller) octahedra. In the phase 3 there remain some doubts about the coordination around Bi, which is larger than 6, rather irregular and includes oxygen atoms which are not part of the octahedral framework. The number and positions of these extra-framework oxygen atoms and the cation composition (in the current model it differs from the proposed one) will have to be clarified in future (X-ray single crystal and neutron powder diffraction are possible alternatives).

All three phases are “small” structures, but despite that, show a range of difficulty from a routine (phase 2) through moderate (phase 1) to challenging (phase 3) and show the role of crystallographic thinking and use of complimentary techniques in the process of SDPD.