This research is focused on discovering and understanding new physics in novel types of nanomaterials. Access to high-performance computing facilities is required for understanding the following novel nanostructure materials that possess unusual properties of technological importance: (1) nanostructures of ferroelectric (FE) and piezoelectric oxides (which exhibit many electrical, mechanical, and structural properties that are not shared by other materials, see below); (2) semiconductor nanomaterials (which are currently one of the major pursuits in nano-technology). The state-of-art first-principles densityfunctional theory (DFT) will be the main computational tool. This theory is known to be very accurate. However, applications of this theory to nanomaterials are exceptionally difficult since meaningful modeling of nano-sized materials requires the consideration of several hundred (or even thousands) of atoms, which far exceeds the size range (~100 atoms) that normal DFT can handle. For this reason, many existing theoretical studies of nano-dots rely on semi-empirical or model approaches. By comparison, the first-principles approach has much greater prediction power and is able to handle a realistic atomic-scale environment. While the direct DFT study of nanomaterials no doubt poses a great challenge, overcoming it, on the other hand, implies a new area of unknown science to be explored and discovered.
Ferroelectric and piezoelectric perovskite oxides distinguish themselves from other classes of materials by being able to efficiently convert electricity into mechanical energy or vice versa. Consequently, they have been widely utilized in energy-conversion devices such as transducers and actuators, robotics, and micro-electric mechanical systems. These materials have also been ubiquitously used in ultrasonic imaging for human health, arrays for telecommunications, and military sensors for national security, as well as in nonvolatile random accessed memories. Fundamentally, the spectacular modern theory of polarization using geometric phase as well as the significance of polarization in density functional theory [g8, g9] are examples indicating the importance of the field. Compared to bulk ferroelectrics, FE nanostructures (e.g., thin films, wires, nanotubes, dots) are poorly understood (in fact, most of their properties remain entirely unknown). Recently, the Fu and Bellaiche group predicted that a novel vortex phase may exist in FE particles, using first-principles derived effective Hamiltonian methods [g14, g15]. The properties in FE nanostructures were found to be very different from those of the bulk, indicating new physics and knowledge in nano-FE. The goal here is to be able to model the FE nanostructures using the direct DFT theory, since the direct approach is reliable and accurate down to very small size (e.g., a few atoms) and can also potentially take into account the realistic surface and boundary conditions. It is well known that the relevant energy scale of ferroelectric instability is only a few milli-eV, a property that requires very accurate modeling.
Surface conditions are expected to be increasingly important when the size become smaller, in which the direct approach thus becomes necessary. Though an accurate and efficient mix-basis DFT code has been developed (which has been successfully applied to study electrical, dielectric, electromechanical, optical and elastic properties, its applications to FE nanostructures have not been attempted because of the lack of sufficient computing power. This challenging problem in materials science can be tackled with access to high performance computing clusters.
Investigator: Huaxiang Fu (UAF)