Ionically conducting polymers (polymer electrolytes) are under intensive investigation because they form the basis of all solid-state lithium batteries, fuel cells, and electrochromic display devices, as well as being highly novel electrolytes. Little is known about the structures of the many crystalline complexes that form between poly(ethylene oxide) and a wide range of salts. The crystal structure is reported of the archetypal polymer electrolyte poly(ethylene oxide) ₃ :LiCF ₃ SO ₃ , which has been determined from powder x-ray diffraction data. The poly(ethylene oxide) (PEO) chain adopts a helical conformation parallel to the crystallographic b axis. The Li ⁺ cation is coordinated by five oxygen atoms—three ether oxygens and one from each of two adjacent CF ₃ SO ₃ ^- groups. Each CF ₃ SO ₃ ^- in turn bridges two Li ⁺ ions to form chains running parallel to and intertwined with the PEO chain. There are no interchain links between PEO chains, and the electrolyte can be regarded as an infinite columnar coordination complex.

ABO₃ perovskites have fascinated solid-state chemists and physicists for decades because they display a seemingly inexhaustible variety of chemical and physical properties. However, despite the diversity of properties found among perovskites, very few of these materials are ferroelectric, or even polar, in bulk. In this Perspective, we highlight recent theoretical and experimental studies that have shown how a combination of non-polar structural distortions, commonly tilts or rotations of the BO₆ octahedra, can give rise to polar structures or ferroelectricity in several families of layered perovskites. We discuss the crystal chemical origin of the polarization in each of these families – which emerges through a so-called ‘trilinear coupling’ or ‘hybrid improper’ mechanism – and emphasize areas in which further theoretical and experimental investigation is needed. We also consider how this mechanism may provide a generic route for designing not only new ferroelectrics, but also materials with various other multifunctionalities, such as magnetoelectrics and electric field-controllable metal-insulator transitions.

A detailed investigation using variable temperature powder neutron diffraction demonstrates that ${\mathrm{BiFeO}}_3$ undergoes a phase transition from the ferroelectric $\alpha$ phase (rhombohedral, $R3c$) to a paraelectric $\beta$ phase (orthorhombic, $Pbnm$) between $820°\mathrm{C}$ and $830°\mathrm{C}$. Coexistence of both phases over a finite temperature interval, together with abrupt changes in key structural parameters, confirms that the transition is first order. The $\beta$ phase corresponds to a ${\mathrm{GdFeO}}_3$-type perovskite structure.

The structures of the ferroelectric two-layer Aurivillius phases Bi ₂ ANb ₂ O ₉ (A=Ba, Sr, Ca) have been refined using a combination of X-ray and neutron powder diffraction data. Bi ₂ BaNb ₂ O ₉ is not significantly distorted from idealised symmetry and has been refined in tetragonal space group I4/mmm, a=3.9362(1) and c=25.6582(7) Å. The Sr and Ca compounds have been refined in orthorhombic space group A2 ₁ am, with a=5.5193(3), b=5.5148(3), c=25.0857(6) Å and a=5.4833(1), b=5.4423(1), c=24.8984(6) Å, respectively. The orthorhombic distortion increases with decreasing A ²⁺ cation size and originates from bonding requirements at the perovskite A site, in agreement with previous work. However, in contrast to earlier work, we find a partial mixing of Bi and A cations on their respective sites, which increases in the order Ca<Sr<Ba. This effect is also discussed in terms of bonding requirements at the metal sites and bond valence sum analysis. The combined use of X-ray and neutron powder diffraction data is found to be essential in determining this subtle effect, which may have important consequences in the interpretation of the ferroelectric behaviour of this family of materials.