We have studied the polarization-electric-field hysteresis, the dielectric permittivity dispersion, the piezoelectric properties, the electric-field-induced strain, and the interrelations between these properties for bismuth ferrite (BiFeO3) ceramics. The results indicate that the domain-wall movement in BiFeO3 is strongly inhibited by charged defects, most probably acceptor-oxygen-vacancy defect pairs. The domain-wall mobility can be considerably increased by preventing the defects from migrating into their stable configuration; this can be achieved by thermal quenching from above the Curie temperature, which freezes the disordered defect state. Similarly, Bi2O3 loss during annealing at high temperatures contributes to depinning of the domain walls and an increase in the remanent polarization. The possible defects causing the pinning effect are analyzed and discussed. A weakening of the contacts between the grains in the ceramics and crack propagation were observed during poling with constant field at 100 kV/cm. This is probably caused by an electrically induced strain associated with ferroelastic domain reversal. A relatively large piezoelectric d33 constant of 44 pC/N was obtained by “cyclic poling,” in which the electric field was released after each applied cycle with the purpose to relax the mechanical stresses and minimize the problem of cracking.

We report on the crystal structure of a new polymorph of Li(2)FeSiO(4) (prepared by annealing under argon at 900 degrees C and quenching to 25 degrees C) characterized by electron microscopy and powder X-ray and neutron diffraction. The crystal structure of Li(2)FeSiO(4) quenched from 900 degrees C is isostructural with Li(2)CdSiO(4), described in the space group Pmnb with lattice parameters a = 6.2836(1) A, b = 10.6572(1) A, and c = 5.0386(1) A. A close comparison is made with the structure of Li(2)FeSiO(4) quenched from 700 degrees C, published recently by Nishimura et al. (J. Am. Chem. Soc. 2008, 130, 13212). The two polymorphs differ mainly on the respective orientations and alternate sequences of corner-sharing FeO(4) and SiO(4) tetrahedra.

Highly defective LiFePO4 powders were synthesized via a modified version of the co-precipitation in aqueous medium method using oxidizing experimental conditions. A pure olivine phase containing 44 at.% of Fe 3+ is obtained after only 10 minutes at 108 °C, and the evolution of the structure and purity is followed during reaction. The nature of the native defects and their influence on the crystallographic structure and on the electrochemical reaction mechanisms are thoroughly studied by a combination of ex situ and in situ methods, using high-resolution transmission electron microscopy, inductive coupled plasma, X-ray diffraction and Mössbauer spectroscopy. The high concentration of defects induces a unit-cell volume 4 Å 3 smaller than that of stoichiometric LiFePO4, with a complete cationic redistribution over the M1 and M2 crystallographic sites, and a completely new electrochemical signature. A precise structural model during electrochemical operation of the pristine defective "LiFePO4" is built.

As part of a broad project to explore Li4MM′O6 materials (with M and M′ being selected from a wide variety of metals) as positive electrode materials for Li-ion batteries, the structures of Li4FeSbO6 materials with both stoichiometric and slightly deficient lithium contents are studied here. For lithium content varying from 3.8 to 4.0, the color changes from yellow to black and extra superstructure peaks are seen in the XRD patterns. These extra peaks appear as satellites around the four superstructure peaks affected by the stacking of the transition metal atoms. Refinements of both XRD and neutron scattering patterns show a nearly perfect ordering of Li, Fe, and Sb in the transition metal layers of all samples, although these refinements must take the stacking faults into account in order to extract information about the structure of the TM layers. The structure of the most lithium rich sample, where the satellite superstructure peaks are seen, was determined with the help of HRTEM, XRD, and neutron scattering. The satellites arise due to a new stacking sequence where not all transition metal layers are identical but instead two slightly different compositions stack in an AABB sequence giving a unit cell that is four times larger than normal for such monoclinic layered materials. The more lithium deficient samples are found to contain metal site vacancies based on elemental analysis and Mössbauer spectroscopy results. The significant changes in physical properties are attributed to the presence of these vacancies. This study illustrates the great importance of carefully determining the final compositions in these materials, as very small differences in compositions may have large impacts on structures and properties.

No abstract