Experimental and first-principles study of point defects, domain walls, and point-defect/domain-wall interactions in ferroelectric oxides
Ferroelectric oxides, such as lead zirconate titanate, have proved invaluable due to their excellent dielectric and piezoelectric properties. These classes of materials possess a large electric polarization below the Curie temperature. Regions of the crystal lattice having different directions of polarization are separated by nano-scale interfaces known as domain walls. These interfaces, which can be moved by electric and mechanical fields, are not only important for electromechanical properties but they also exhibit unique structural and electronic properties that may be exploited through novel application in nanoelectronics and photovoltaics. The mobility of domain walls is strongly dependent on the defect chemistry of the material due to defect-domain wall interactions. Indeed, aliovalent doping of these materials has been used to tune the properties of ferroelectric oxides for specific applications. In this work, we investigate the origin of the so-called â hardâ and â softâ behavior in lead zirconate titanate. Experimental evidence suggests that small concentrations of donor dopants (such as niobium) results in enhanced domain wall motion while doping with acceptor dopants (such as iron) leads to an inhibited domain wall response. The microscopic origin of this phenomena is best investigated using first-principles simulations of the atomistic properties of defects and domain wall in these materials to bring to light novel structural and electronic properties. First, we show that polar defect complexes are likely to exist in both acceptor-doped and undoped PbTiO3 . These defects and defect associates are attracted to 180o domain walls and cause pinning of such interfaces. Donor- doped materials, on the other hand, do not show the presence of polar defect complexes. A closer investigation of the 180o domain wall in PbTiO3 reveals the presence of a â Blochâ component of polarization at the domain wall. In other words, due to polarization rotation, there exists a ferroelectric phase within the domain wall itself. We characterize the strain dependence of this phenomena and calculate the piezoelectric properties of such domain walls. A complete study of domain walls in PbTiO3 also entails looking closer at the properties of the ferroelastic â head-to-tailâ 90o domain boundary. We show the presence of an asymmetry in the variation of the lattice parameter across the domain wall. This asymmetry is verified using high resolution aberration corrected electron microscopy. We look at the energy landscape of oxygen vacancies in the vicinity of these walls to explain the pinning effect in terms of random-bond and random-field defects. Next we look at the electronic properties of the recently discovered charged domain walls. We characterize the band bending phenomena in 90o head-to-head and tail-to-tail domain wall in PbTiO3 . We then look at the long range effects of clusters of oxygen vacancies in BaTiO3; especially in terms of the formation of 180o tail-to-tail domain wall. In parallel to the first priciples work, we also prepared high quality samples of PZT 50/50 with different concentrations of donor dopant. We characterized the microstructure and hysteresis loops as a function of donor dopant concentration. Based on experimental observations and ab initio calculations we propose ideas on the origins of hardening and softening in PZT.
EPFL_TH6931.pdf
openaccess
12.2 MB
Adobe PDF
0fe389ce59a0cc09791d27562ad17492