Silicide formation and the interaction of metals with polycrystalline Si
<p>The main factors affecting solid-phase Si-metal interactions are reported in this work. The influence of the orientation of the Si substrates and the presence of impurities in metal films and at the Si-metal interface on the formation of nickel and chromium silicides have been demonstrated....
| Summary: | <p>The main factors affecting solid-phase Si-metal interactions are reported in this work. The influence of the orientation of the Si substrates and the presence of impurities in metal films and at the Si-metal interface on the formation of nickel and chromium silicides have been demonstrated. We have observed that the formation and kinetic rate of growth of nickel silicides is strongly dependent on the orientation and crystallinity of the Si substrates; a fact which, up to date, has never been seriously investigated in silicide formation. Impurity contaminations in the Cr film and at the Si-Cr interface are the most dominant influencing factors in the formation and kinetic rate of growth of CrSi<sub>2</sub>. The potentiality and use of silicides as a diffusion barrier in metallization on silicon devices were also investigated.</p>
<p>Two phases, Ni<sub>2</sub>Si and NiSi, form simultaneously in two distinct sublayers in the reaction of Ni with amorphous Si, while only the former phase was observed on other substrates. On (111) oriented Si substrates the growth rate is about 2 to 3 times less than that on <100> or polycrystalline Si. Transmission electron micrographs establish-·that silicide layers grown on different substrates have different microcrystalline structures. The concept of grain-boundary diffusion is speculated to be an important factor in silicide formation.</p>
<p>The composition and kinetic rate of CrSi<sub>2</sub> formation are not influenced by the underlying Si substrate. While the orientation of the Si substrate does not affect the formation of CrSi<sub>2</sub> , the purity of the Cr film and the state of Si-Cr interface become the predominant factors in the reaction process. With an interposed layer of Pd<sub>2</sub>Si between the Cr film and the Si substrate, CrSi<sub>2</sub> starts to form at a much lower temperature (400°C) relative to the Si-Cr system. However, the growth rate of CrSi<sub>2</sub> is observed to be independent of the thickness of the Pd2Si layer. For both Si-Cr and Si-Pd<sub>2</sub>Si-Cr samples, the growth rate is linear with time with an activation energy of 1.7 ± 0.1 ev.</p>
<p>A tracer technique using radioactive <sup>31</sup>Si (T<sub>1/2</sub> = 2.26 h) was used to study the formation of CrSi<sub>2</sub> on Pd<sub>2</sub>Si. It is established from this experiment that the growth of CrSi<sub>2</sub> takes place partly by transport of Si directly from the Si substrate and partly by breaking Pd<sub>2</sub>Si bonds, making free Si atoms available for the growth process.</p>
<p>The role of CrSi<sub>2</sub> in Pd-Al metallization on Si was studied. It is established that a thin CrSi<sub>2</sub> layer can be used as a diffusion barrier to prevent Al from interacting with Pd<sub>2</sub>Si in the Pd-Al metallization on Si.</p>
<p>As a generalization of what has been observed for polycrystalline-Si-Al interaction, the reactions between polycrystalline Si (poly Si) and other metals were studied. The metals investigated include Ni, Cr, Pd, Ag and Au. For Ni, Cr and Pd, annealing results in silicide formation, at temperatures similar to those observed on single crystal Si substrates. For Al, Ag and Au, which form simple eutectics with Si annealing results in erosion of the poly Si layer and growth of Si crystallites in the metal films.</p>
<p>Backscattering spectrometry with 2.0 and 2.3 MeV <sup>4</sup>He ions was the main analytical tool used in all our investigations. Other experimental techniques include the Read camera glancing angle x-ray diffraction, scanning electron, optical and transmission electron microscopy. Details of these analytical techniques are given in Chapter II.</p>
|
|---|