The long-standing problem of growing a commensurate crystalline oxide interface with silicon has been solved. Alkaline earth and perovskite oxides can be grown in perfect registry on the (001) face of silicon, totally avoiding the amorphous silica phase that ordinarily forms when silicon is exposed to an oxygen containing environment. The physics of the heteroepitaxy lies in establishing a sequenced transition that uniquely addresses the thermodynamics of a layer-by-layer energy minimization at the interface. A metal-oxide-semiconductor capacitor using SrTiO 3 as an alternative to SiO 2 yields the extraordinary result of t eq , 10 Å. [S0031-9007(98)07238-X] PACS numbers: 81.15. Hi, 73.40.Qv, 77.55. + f Since the advent of the integrated circuit in 1959 and the introduction of metal-oxide-semiconductor (MOS) capacitors in the early 1960s, electronic technology has relied on silica ͑SiO 2 ͒ as the gate dielectric in a field effect transistor. However, silica-based transistor technology is approaching fundamental limits. Feature-size reduction and the ever-demanding technology roadmaps have imposed scaling constraints on gate oxide thickness to the point where excessive tunneling currents make transistor design untenable; an alternative gate dielectric is needed [1].While now it is especially clear, with SiO 2 thicknesses in the sub-50-Å regime, the argument for alternative gate oxides is not new; it has been made from different perspectives for over 40 years [2][3][4][5]. Quite aside from the "physical" thickness limits that tunneling currents make obvious, the amorphous SiO 2 interface with silicon leaves dangling bonds as electronic defects disrupting translational symmetry at the interface. An alternative crystalline gate oxide would, in principle at least, uniquely maintain a one-to-one correspondence between physical and electrical structure preserving translational symmetry to atomic dimensions.Crystalline oxides on silicon (COS), simply by virtue of their high dielectric constants, could fundamentally change the scaling laws for silicon-based transistor technology. More importantly COS introduces the possibility for an entirely new device physics based on utilization of the anisotropic response of crystalline oxides grown commensurately on a semiconductor. In this Letter, we report that high dielectric constant alkaline earth and perovskite oxides can be grown in perfect registry with silicon. Commensurate heteroepitaxy between the semiconductor and the oxide is established via a sequenced transition that uniquely addresses the thermodynamics of a layer-by-layer energy minimization at the interface. The perfection of the physical structure couples directly to the electrical structure, and we thus obtain the unparalleled result of an equivalent oxide thickness of less than 10 Å in a MOS capacitor.An equivalent oxide thickness t eq can be defined for a MOS capacitor asin which´S iO 2 and´0 are the dielectric constants of silica and the permittivity of free space. ͑C͞A͒ ox is the specific capacitance of the...

No abstract

Perovskite‐spinel epitaxial nanocomposite thin films are commonly grown on single crystal perovskite substrates, but integration onto a Si substrate can greatly increase their usefulness in devices. Epitaxial BiFeO3–CoFe2O4 nanocomposites consisting of CoFe2O4 pillars in a BiFeO3 matrix are grown on (001) Si with two types of buffer layers: molecular beam epitaxy (MBE)‐grown SrTiO3‐coated Si and pulsed‐laser‐deposited (PLD) Sr(Ti0.65Fe0.35)O3/CeO2/yttria‐stabilized ZrO2/Si. The nanocomposite grows with the same crystallographic orientation and morphology as that observed on single crystal SrTiO3 when the buffered Si substrates are smooth, but roughness of the Sr(Ti0.65Fe0.35)O3 promoted additional CoFe2O4 pillar orientations with 45° rotation. The nanocomposites on MBE‐buffered Si show very high magnetic anisotropy resulting from magnetoelastic effects, whereas the hysteresis of nanocomposites on PLD‐buffered Si can be understood as a combination of the hysteresis of the Sr(Ti0.65Fe0.35)O3 film and the CoFe2O4 pillars.

Interfaces between transition metal oxide compounds provide a setting where correlated behavior can emerge. Essential to interface studies are substrates that have a single atomic plane termination. By tuning the vapor pressure of La above the surface of La0.18Sr0.82Al0.59Ta0.41O3 (LSAT) during annealing, single‐unit‐cell steps and predominant A‐site (SrO) termination can be achieved.