We analyze the hypothetical link between octahedral straightening and increased conductivity inside the domain walls of BiFeO3. Our calculations for 109 degree walls predict a lattice parameter expansion of c.a. 1 percent in the direction perpendicul
ar to the wall, and an associated straightening of the octahedral rotation angle of 4 degrees, which is comparable to that observed in the high temperature metallic phase of BiFeO3. On the other hand, in the closely related family of rare-earth orthoferrites, straighter octahedra do not correlate with increased bandgap, which suggests that the correlation between octahedral straightening and bandgap reduction in BiFeO3 is perhaps fortuitous and not necessarily the cause of increased conductivity at the walls.
Multiferroic BiFeO3 ceramics have been doped with Ca. The smaller ionic size of Ca compared with Bi means that doping acts as a proxy for hydrostatic pressure, at a rate of 1%Ca=0.3GPa. It is also found that the magnetic Neel temperature (TNeel) incr
eases as Ca concentration increases, at a rate of 0.66K per 1%Ca (molar). Based on the effect of chemical pressure on TNeel, we argue that applying hydrostatic pressure to pure BiFeO3 can be expected to increase its magnetic transition temperature at a rate around ~2.2K/GPa. The results also suggest that pressure (chemical or hydrostatic) could be used to bring the ferroelectric critical temperature, Tc, and the magnetic TNeel closer together, thereby enhancing magnetoelectric coupling, provided that electrical conductivity can be kept sufficiently low.
We report X-ray structural studies of the metal-insulator phase transition in bismuth ferrite, BiFeO3, both as a function of temperature and of pressure (931 oC at atmospheric pressure and ca. 45 GPa at ambient temperature). Based on the experimental
results, we argue that the metallic gamma-phase is not rhombohedral but is instead the same cubic Pm3m structure whether obtained via high temperature or high pressure, that the MI transition is second order or very nearly so, that this is a band-type transition due to semi-metal band overlap in the cubic phase and not a Mott transition, and that it is primarily structural and not an S=5/2 to S=1/2 high-spin/low-spin electronic transition. Our data are compatible with the orthorhombic Pbnm structure for the beta-phase determined definitively by the neutron scattering study of Arnold et al .[Phys. Rev. Lett. 2009]; the details of this beta-phase had also been controversial, with a remarkable collection of five crystal classes (cubic, tetragonal, orthorhombic, monoclinic, and rhombohedral!) all claimed in recent publications.
We have analyzed the morphology of ferroelectric domains in very thin films of multiferroic BiFeO3. Unlike the more common stripe domains observed in thicker films BiFeO3 or in other ferroics, the domains tend not to be straight, but irregular in sha
pe, with significant domain wall roughening leading to a fractal dimensionality. Also contrary to what is usually observed in other ferroics, the domain size appears not to scale as the square root of the film thickness. A model is proposed in which the observed domain size as a function of film thickness can be directly linked to the fractal dimension of the domains.
The International Roadmap for Ferroelectric Memories requires three-dimensional integration of high-dielectric materials onto metal interconnects or bottom electrodes by 2010. We report the first integration of high-dielectric oxide films onto carbon
nanotube electrodes with an aim of ultra-high integration density of FeRAMs (Tb/in2).
