Carrier modulation of BaSnO3 via field effect with various gate oxides and their interfaces

Abstract

The oxide materials were generally believed as one of the most promising materials for the electronic industry due to its exceptional novel characteristics such as high optical transparency and high electrical conductivity compared to conventional materials such as metals or ceramics. In fact, Sn:In2O3, Cu2O, ZnO, In2O3-ZnO, In2O3-Ga2O3, In2O3-Ga2O3-ZnO have been widely investigated and used as transparent conducting oxides and transparent oxide semiconductors. Additionally, during the past few decades, perovskite oxides such as YBa2Cu3O7, Pb(ZrxTi1-x)O3, and SrTiO3 with ABO3 formula have received a large amount of attention due to their diverse properties such as superconductivity, high-k dielectrics, ferromagnetic, photoconductivity and ferroelectricity. Many researchers have already demonstrated field effect transistors based on perovskite oxide materials in a metal-insulator-semiconductor heterostructure, which is a most fundamental component in oxide electronics. But even though the field effect transistor devices based on SrTiO3 and KaTaO3 were already demonstrated, it was impossible to achieve the high performance shown in devices based on binary oxides, due to their poor electron transport properties. Moreover, their lack of oxygen stability limited their applications. Recently, BaSnO3 has attracted large attention for many researchers due to its excellent properties compared with the materials mentioned above: the oxygen stability even at high temperatures, and the high electrical mobility at room temperature. BaSnO3 has a wide band gap with an optical band gap of 3.1 eV. La-doped BaSnO3 single crystals and thin films have a mobility value of about 300 cm2V-1s-1 and 70 cm2V-1s-1, respectively. It is known that the origin of the difference in mobility between single crystals and thin films is due to the existence of dislocation scatterings or grain boundaries in thin films. The oxygen diffusion constant of BaSnO3, evaluated by high temperature conductivity measurements, is determined to be 10-15 cm2s-1. Its value is lower than manganites, cupprates, and titanates by 3 ~ 10 orders-of-magnitude. By using these novel properties, the author performed carrier modulations of BaSnO3 via field effect with amorphous and epitaxial gate dielectrics. To obtain the appropriate channel layer properties, the author firstly investigated the influence of buffer layers, which is known to be able to reduce the dislocation density on La-doped BaSnO3 and improve the electrical transport. This is because the dislocations act as a defect (the source of charge traps or scattering centers) on the interface between BaSnO3 and the dielectric gate oxide. The author chose Al2O3 and HfO2 as gate dielectrics among the many amorphous gate dielectric candidates such as ZrO2, Y2O3, Ta2O5, TiO2. The author also investigated the dielectric properties of Al2O3 and HfO2, namely the breakdown field and dielectric constant. Furthermore, the author demonstrated a field effect transistor made with an undoped buffer layer of BaSnO3 on a SrTiO3 substrate using a lightly La-doped BaSnO3 channel, with the gate dielectrics as Al2O3 and HfO2. The performances of these devices, such as field effect mobility and Ion/Ioff ratio are consistent with the known material parameters of La-doped BaSnO3, Al2O3 and HfO2. And they also provided further evidence for the perfect surface quality of La-doped BaSnO3 and its stability. Through comparison with SrTiO3 or KTaO3 based field effect transistor devices, the author concluded again that these performances are evidences of the superior material properties of BaSnO3, and show the potential of BaSnO3 as a core material in a transparent, high-mobility field effect transistor device. To further enhance the performance of the device, the author used epitaxial gate dielectrics. Among the epitaxial gate dielectric candidates, such as SrZrO3, BaZrO3, SrHfO3, SrSnO3, the author chose LaInO3 and BaHfO3. Also, the author investigated the crystallinity and dielectric properties of these materials by X-ray diffraction and electrical measurements. Then, the author demonstrated field effect transistors using the chosen materials. Overall, the performance of the devices was improved by using epitaxial gate dielectrics. Especially, a remarkable performance was achieved in LaInO3/La-doped BaSnO3 devices. It is comparable with the performance of devices using binary oxides such as ZnO, In-Ga-Zn-O and SnO2. Additionally, during the fabrication of the devices, the author observed the sheet conductance enhancement of the La-doped BSO active layer in the LaInO3/La-doped BaSnO3 interface. The author believes a 2-dimensional electron gas is formed at its interface. To explain the origin of this phenomenon, the author investigated the La concentration dependence of BSO, the thickness dependence of LaInO3, the possibility of La diffusion, and the oxygen vacancy on the LaInO3/La-doped BaSnO3 interface. The author believes that these investigations on the LaInO3/BaSnO3 interface provide a better understanding of the origin of the 2DEG phenomena at the interface between two-band insulators.