When is a ferroelectric not a ferroelectric?

Ferroelectrics are insulating materials with an electrical polarisation that can be switched by an applied voltage. Ferroelectricity cannot occur in metals because it would be screened by the conduction electrons. In an article published in Nature Materials, we report the discovery of a new material called lithium osmate (LiOsO3) which remains a metal down to the lowest temperatures and yet undergoes a structural phase transition that is identical to the ferroelectric transition in the well-known ferroelectrics LiNbO3 and LiTaO3.

Using a variety of techniques, including neutron diffraction at the ISIS Facility, Oxfordshire, Yanfeng Guo, Andrew Princep and Andrew Boothroyd, together with co-workers from ISIS, Japan and China, found that the phase transition in LiOsO3 is characterised by a large shift in the position of the Li ions, a structural effect which has been known for many years to cause ferroelectricity in LiNbO3 and LiTaO3.

The discovery represents the first clear-cut example of a so-called “ferroelectric” metal, a concept first postulated over 50 years ago by Nobel prize-winner Philip Anderson and co-worker Blount. It is scientifically interesting because the mechanisms for structural phase transitions are usually quite distinct in metals and insulators, so it is surprising to find a metal (LiOsO3) that undergoes the same structural transition as occurs in the insulating analogues. The discovery of a “ferroelectric” metal establishes a new class of materials which could have interesting properties, possibility including non-centrosymmetric superconductivity stabilised by the “ferroelectric” structural instability.

Reference: Y. Shi et al. Nature Materials doi:10.1038/nmat3754
News & Views article: doi:10.1038/nmat3774