| Summary: | 碩士 === 國立清華大學 === 材料科學工程學系 === 92 === In first part of our experiment, stress evolution and detail solid reaction process for Fe/Si system were provided. Two major curvature changes were found near 400 ℃ and 600 ℃, which is related to the transition of FeSi and β-FeSi2. FeSi forms at the interface between Fe film and Si substrate. The amorphous Si capping layer strongly affects the formation of FeSi phase. Large compressive stress was measured during the reaction, which is manly caused by the amorphorization of Fe film due to the diffusion of Si atoms. A small tensile stress was observed at the end of the FeSi transition (from 450 ℃ to 470 ℃) relating to the Fe to FeSi reaction. Nucleation-controlled growth mechanism was found for β-FeSi2. The nucleation site is at the interface between FeSi and Si substrate. The diffusion species is observed to be Si. The increase of diffusion distances cause the reaction retards above 600 ℃. No obvious stress change was found for the formation of β-FeSi2. However, the reason is too complex to give a detail study for it.
In second part, multilayer systems of Ni (60nm)/Fe (3nm)/Si, Ni (3nm)/Fe (3nm)/Si, and Ni (3nm)/Fe (60nm)/Si were used to synthesized tantalum silicide nanowires via vacuum annealing under 950 ℃ for 32 hr. For Ni (60nm)/Fe (3nm)/Si system and Ni (3nm)/Fe (3nm)/Si, TaSi2 wires with small amounts of Fe and Ni were observed. The growth direction is perpendicular to the (2-10) plane. Large concentration of Fe and Ni was found for Ni (3nm)/Fe (60nm)/Si. Owing to the absence of any extra points in the diffraction patterns, the Fe and Ni atoms should exist in a random form. Thus the Fe and Ni doped TaSi2 wires were acquired. The wires with high concentration of Fe (over 15%) induced by Fe dots system are also identified to be TaSi2. Length effect for every different condition was analyzed, and the wire length is highly affected by the Fe concentration. Cross-sectional TEM images of three different substrates show that Fe and Ni disilicides are the keys to the wire growth. The well-known Vapor-Solid (VS) mechanism is the possible candidate. Ta source is provided by Ta filament and the Si source is mainly from Fe and Ni disilicides due to the study of length effect. The melting points of FeSi2 and NiSi2 are 1200 ℃ and 993 ℃, which explains the longer wire obtained by the Ni rich system. Field Emission results show that the work function of wires with small amounts of Fe and Ni is about 4.7eV, which is similar to the bulk. A decrease of work function (2.39eV) was found for the Fe rich system. We also obtain the same result for the Fe dots system. Therefore, the adulteration of Fe atoms will reduce the work function of TaSi2 wire.
|
|---|
