After the perovskite family, the tetragonal tungsten bronzes (TTB) group is the largest group of ferroelectric materials. Although a complete understanding of the polarization mechanism(s) in TTBs does not exist, it is generally accepted that lead-containing TTBs often have an in-plane polarization leading to an orthorhombic distortion of the structure, while most lead-free TTBs have a uniaxial out-of-plane polarization and a tetragonal structure. These observations are rationalized based on the lead lone pair and a second-order Jahn-Teller effect, respectively. This mix of polarization mechanisms and structures (orthorhombic and tetragonal) opens the possibility to find and engineer morphotropic phase boundaries among the TTBs, and this becomes especially attractive if one can find a lead-free alternative with an in-plane polarization. Bismuth containing TTBs, namely K₂BiNb₅O₁₅ and Rb₂BiNb₅O₁₅ have recently been investigated for this purpose but were found to be non-ferroelectric. Here, we present an in-depth structural characterization study, combining X-ray powder diffraction and selected area electron diffraction data to study the incommensurately modulated structures of KBN and RBN. Niobium-octahedra tilting patterns along the c-direction are observed, in addition to significant in-plane displacement of bismuth away from its non-modulated position. This in-plane displacement does not order in a way that leads to switchable polarization, as it does for the lead-containing TTBs. As incommensurately modulated structures are a common feature of TTBs, we believe that it is vital to solve these structures, for a full picture of the structure-property relationship of these materials. This will for example enable a better understanding of the polarization mechanism(s) in this class of materials. We aim to gain a fundamental understanding of the structure-property relationship of TTBs in general, but also demonstrate the important proof-of-concept that incommensurately modulated structures can be solved with X-ray powder diffraction data.

The tetragonal tungsten bronze, Sr2NaNb5O15 (SNN), shows promise for high-temperature high-efficiency capacitors vital for the sustainable energy revolution. The structural complexity of tetragonal tungsten bronzes has obscured the mechanisms underpinning two large anomalies in the relative permittivity of SNN, which give rise to its exceptionally broad dielectric response. We use electron microscopy and diffraction to establish the reciprocal lattice and microstructure of SNN, and this result allows X-ray and neutron diffraction, together with symmetry-adapted distortion mode analysis and first principles electronic structure calculations, to unambiguously identify the structural origins of both anomalies. The high temperature phase (>300°C) has a non-polar orthorhombic √2 x 2√2 x 2 Amam structure, becoming polar along the c-axis in the room-temperature Amam structure. A second phase transition to a monoclinic Aa structure, with polarisation perpendicular to the c-axis occurs at around 0°C. All these phases appear slightly incommensurate. Despite the lowering of symmetry with respect to the tetragonal prototype, the unit cell remains unaltered through these phase changes and the small difference in a and b lattice parameters allows twinned regions to coexist with very little strain. A striking feature of the microstructure is the high density of anti-phase boundaries, which lie on parallel planes of the correct orientation to coincide with neutral domain walls.

Transition metal oxides exhibit a variety of functionalities that make them highly promising for the development of novel electronic devices including non-volatile memories, sensors, and flexible electronics. However, their integration into CMOS technology remains a significant challenge. A potential solution to this issue can be the transfer of epitaxial layers, released from their substrate, onto a target substrate. Among the strategies explored for releasing epitaxial oxides from their substrate, the chemical lift-off of a sacrificial layer grown between the substrate and the oxide has attracted significant attention. Here, we fabricate SrTiO_3 (STO) membranes, by growing a Sr_3 Al_2 O_6 (SAO) sacrificial layer and a 30nm-thick STO layer on a STO(001) substrate using pulsed laser deposition. The SAO layer was dissolved in deionized water, and the resulting STO membrane was transferred onto a Nb:STO(001) target substrate. The topography and the crystal structure of the membrane before and after release were probed by atomic force microscopy (AFM) and X-ray diffraction, respectively. AFM analysis revealed a RMS roughness of the membrane of ~ 0.5 nm, and X-ray diffraction confirmed that the crystallinity was preserved after release. To explore the nature of the membrane/target substrate interface bonding, electron energy-loss spectroscopy at atomic level precision (STEM-EELS) is employed, in particular the effect of the annealing temperature is investigated. Using the same approach, we are investigating ferroelectric oxide membranes (i.e. PbTiO_3) to study the behaviour of the ferroelectric polarisation and the domains structure before and after release from the substrate. These findings could pave the way for the development of novel electronic devices and for the integration of ferroelectric perovskites into CMOS-compatible platforms.