University of Connecticut University of UC Title Fallback Connecticut

Tin-based ferroelectrics to replace lead titanate

This project is dedicated to exploring the realm of novel multifunctional tin-containing oxide materials, where the 2+ oxidation state of tin (Sn) is associated with the presence of electron “lone pairs.” Large ionic displacements induced by these sterically active, but chemically inert electrons lead to large elastic deformations and strong polar properties. Our synergistic approach includes both materials theory and computation, and growth and characterization as essential interconnected parts. Utilizing the combination of all of these tools, we investigate complex coupled phenomena − including elastic, polar and electronic ones − in bulk, thin-film and nanostructured forms of tin-based oxides, identify rational pathways for achieving enhanced or yet unknown new properties in these compounds that will lead to advanced functionalities, and develop synthetic processes to grow tin-based oxides utilizing atomic layer / chemical vapor deposition (ALD/CVD), as well as processing technologies to incorporate them into a variety of nanostructures. Naturally, this is a collaborative effort that includes this group at the University of Connecticut (theory and simulation), the group of Prof. Christos Takoudis at the University of Illinois at Chicago (ALD and CVD growth, and sample characterization) and our DoE Lab partner Seungbum Hong at Argonne National Laboratory (piezo force microscopy). Images below are loosely borrowed from the paper on First-principles studies of misfit strain-stabilized ferroelectric SnTiO3, coauthored by William Parker, James Rondinelli and Serge Nakhmanson in 2011, as well as a number of conference presentations given around that time and a new paper that came out in Jan of 2015.

Related publications:

An investigation of stability of various SnTiO3 polymorphs done by calculating their optimized structures within the DFT/LDA. Plausible polymorphs include polar and nonpolar perovskites and trigonal R3c [lithium niobate (LN)] phases, all of which have corner-sharing TiO6 octahedra. Layered hexagonal phases are also feasible: R3 (ilmenite), P63/mmc, and P63mc structures, the former containing edge-sharing and the latter two face-sharing TiO6 octahedral networks.The connectivity of the octahedral network strongly affects the presence or absence of polar distortions in a given polymorph. The phases with corner-sharing octahedra are, in general, more stable than the hexagonal polymorphs. The polar perovskite (P4mm) phase has the lowest energy of all structures, and the two other low-energy phases are monoclinic perovskite Cm and LN-type R3c.The figure shows energy differences per formula unit for various SnTiO3 polymorphs with respect to the polar perovskite P4mm phase, which has the lowest energy. Energy variations with respect to slight phase volume change are also presented. Layered hexagonal phases have much higher energies, compared to the other phases, and their volume-energy curves are not shown here explicitly.
Polarization (left vertical axis) and tetragonality (right vertical axis, star data points) of epitaxially strained P4mm and Cm SnTiO3 phases as functions of biaxial misfit strain
ε. Polarization of the R3c LN-type structure is also marked on the left as an open square for comparison.The strained polar structures of SnTiO3 have polarization in excess of 0.9 C/m2 throughout the whole range of applied strains. In the P4mm phase, high polarization is accompanied by large tetragonality, which reaches 1.134 in the stress-free structure. Even at 2% tensile strain, the Pz component in the Cm phase remains rather large (∼40% of the total polarization). The polarization values obtained for all the polar SnTiO3 phases and especially the tetragonality of the P4mm structures are higher than those computed for PbTiO3 with the same methodology.
Evolution of the Ti dxy and O px orbital projected density of states (PDOS) in epitaxially strained P4mm and Cm SnTiO3 with biaxial strain. Data for the corresponding P4/mmm structures are also shown in dotted lines. The degree of epitaxial strain (in %) is shown at the top of each panel.The strength of the interaction between these orbitals determines the size of the optical band gap in this material. Under compressive strain, the P4mm structure retains fourfold rotational symmetry, which prevents the hybridization of the Ti dxy state and the O px,y states, suppressing further broadening of the band gap. Under tensile strain, the fourfold symmetry is lifted by the polar displacements, allowing the Ti dxy and the O px,y states to mix. This mixing leads to an additional repulsion between the VBM and CBM, which opens up the band gap.The strong dependence of the size of the gap on the direction of polarization (in Cm vs P4mm phase) opens up interesting opportunities for dynamical tuning of the gap opening by growth of SnTiO3 films under an epitaxial condition that is close to the boundary between the phases, e.g., on a SrTiO3 substrate.
Partial charge density maps within the (110) plane, depicting the characteristic asymmetric charge-density lobes associated with stereochemically active electron lone pairs in ATiO3 (A = Pb, Sn). A2+ cation centered cells are used here, with A2+, Ti, and O ions represented by gray, light blue and red spheres, respectively. SnTiO3 structures are shown in the top row: (a) P4mm, (b) Amm2, (c) Cm and (d) non-polar Pm3m. PbTiO3 structures are assembled in the bottom row: (e) P4mm, (f) Amm2, (g) Cm and (h) non-polar Pm3m. Contour levels shown are between 0 (blue) and 0.24 (red) for SnTiO3, and 0 (blue) and 0.16 e/Å3 (red) for PbTiO3.