Nickel silicide is a most up-to-date self-aligned silicide ͑salicide͒ technology for nanoscale complementary metal-oxidesemiconductor field-effect transistors. However, an unintended oxidation of nickel silicide happened only on As-doped substrate. This abnormal oxidation phenomenon occurred only when the annealing temperature was higher than 613°C ͑sublimation point of As͒. The main reason for the oxidation is believed to the thermal energy that induces the diffusion of Ni from the nickel silicide to the substrate direction. Due to the oxidation, nickel silicide on As-doped substrate showed poor thermal stability contrasted to BF 2 -doped substrate.Salicide ͑Self-aligned silicide͒ technologies have been widely used in high-speed metal-oxide semiconductor field-effect transistors ͑MOSFETs͒ to increase the drive current by reducing the sheet and contact resistance of source/drain ͑S/D͒ and gate. Titanium silicide (TiSi 2 ) and cobalt silicide (CoSi 2 ) have been used sequentially for salicide technology. TiSi 2 replaced CoSi 2 mainly due to its strong linewidth dependency below 0.25 m complementary ͑C͒MOS technology. 1 CoSi 2 has also revealed some problems such as oxide contamination, large silicon consumption, and drastic increase of sheet resistance in nanoscale MOSFETs. 2,3 Therefore, new salicide technology is needed for nanoscale CMOSFETs and nickel silicide is a strong candidate for the replacement of CoSi 2 . 4,5 Using NiSi technology, low resistivity ͑ϳ14 ⍀-cm͒ can be obtained with reduced silicon consumption. It also has a very thin layer after the formation of silicide. Therefore, NiSi is adequate for deep submicrometer integrated circuits that need ultrashallow junctions. 6 However, despite these advantages, poor thermal stability hinders the application of NiSi to nano CMOS technologies which need hightemperature postsilicidation processes. Therefore improvement in thermal stability is a key issue for developing NiSi technology suitable for nanoscale CMOSFETs. In this article, the thermal stability of NiSi was studied with different dopant types, i.e., BF 2 and As, and especially abnormal oxidation of As was characterized in depth before and after postsilicidation annealing. ExperimentalBF 2 and As-doped substrates were used for this experiment and the key processes are as follows. BF 2 ͑20 keV, 4 ϫ 10 15 cm Ϫ2 ) and As ͑50 keV, 5 ϫ 10 15 cm Ϫ2 ) ion implantation for p ϩ /n and n ϩ /p junctions was followed by rapid thermal annealing ͑RTA͒ activation for 10 s at 1050°C. Then, Ni ͑200 Å͒ was deposited using ion beam sputter ͑IBS͒ at a base pressure about 7 ϫ 10 Ϫ7 Torr after a dilute HF dip to remove the native oxide. Next, rapid thermal process ͑RTP͒ was applied to form NiSi at 500°C, 30 s. Unreacted Ni was etched selectively in a chemical solution of HCl ϩ H 2 O 2 ϩ H 2 O ͑4:1:1͒ at room temperature. To test the thermal stability of NiSi, the samples were furnace annealed between 600 and 800°C for 30 min with N 2 ambient after the formation of NiSi.X-ray photoelectron spectroscopy ͑XPS͒ analysis was ...

No abstract

For the analysis of plasma flow velocity, the Mach probe is widely used. Depending on the existence of superthermal electrons particularly at energies higher than the probe biasing voltage, collected ion saturation current can be reduced. This phenomenon can cause considerable errors in flow velocity estimation. The effect of superthermal electrons has been investigated in the steady-state linear divertor plasma simulator MAP-II. The directional measurement of electron energy distribution function (EEDF) suggests the energy of superthermal electrons and their generation mechanisms. The directional dependence of superthermal electrons is clearly observed from the angular distribution of ion saturation current measured using the Mach probe and a directional Langmuir probe (DLP). The error in flow velocity estimation under the conditions of a weakly magnetized plasma (a p = i $ 1) is calculated.

No abstract

No abstract