No abstract

No abstract

A new Schottky barrier (SB) nonvolatile nanowire memory is reported experimentally with efficient low-voltage programming and erasing. By applying an SB source/drain to enhance the electrical field in the silicon gate-all-around nanowire, the nonvolatile silicon-oxide-nitride-oxide-silicon (SONOS) memory can operate at gate voltages of 5 to 7 V for programming and −7 to −9 V for erasing through Fowler-Nordheim tunneling. The larger the gate voltage is, the faster the programming/erasing speed and the wider the threshold-voltage shift are attained. Importantly, the SB nanowire SONOS cells exhibit superior 100-K cycling endurance and high-temperature retention without any damages from metallic silicidation process or field-enhanced tunneling.Index Terms-Gate-all-around nanowire, Schottky barrier (SB), silicon-oxide-nitride-oxide-silicon (SONOS) memory.

Using graded silicon-germanium heterojunctions, the green tunnel field-effect transistor (TFET) can be scaled down into sub-10 nm regimes without short-channel effects. This work elucidates numerically the physical operation and device design of extremely short-channel TFETs with graded silicon-germanium heterojunctions for future low-power and high-performance applications. Critical device factors, such as the drain profile and bandgap engineering, were examined to generate favorable characteristics in the on-current, on-off switching, and off-leakage of very short TFETs. A mildly doped drain with a pure Ge source is preferred in designing the graded TFETs to optimize a desirable green transistor for low-power integrated circuits.

A coupled-channel analysis of the 18,20,22 O(p, p ) data has been performed to determine the neutron transition strengths of the 2 + 1 states in oxygen targets, using the microscopic optical potential and inelastic form factor calculated in the folding model. A complex density-and isospin-dependent version of the CDM3Y6 interaction was constructed, based on the Brueckner-Hartree-Fock calculation of nuclear matter, for the folding model input. Given an accurate isovector density dependence of the CDM3Y6 interaction, the isoscalar (δ 0 ) and isovector (δ 1 ) deformation lengths of the 2 + 1 states in 18,20,22 O have been extracted from the folding model analysis of the (p, p ) data. A specific N dependence of δ 0 and δ 1 has been established which can be linked to the neutron shell closure occurring at N approaching 16. The strongest isovector deformation was found for the 2 + 1 state in 20 O, with δ 1 about 2.5 times larger than δ 0 , which indicates a strong core polarization by the valence neutrons in 20 O. The ratios of the neutron/proton transition matrix elements (M n /M p ) determined for the 2 + 1 states in 18,20 O have been compared with those deduced from the mirror symmetry, using the measured B(E2) values of the 2 + 1 states in the proton-rich 18 Ne and 20 Mg nuclei, to discuss the isospin impurity in the 2 + 1 excitation of the A = 18, T = 1 and A = 20, T = 2 isobars.