Monte Carlo simulations of via filling have been performed using the lattice model of crystal growth, which allows vacancy formation in the film during the growth. Adsorption, desorption and surface diffusion of adatoms are taken into account. Using a two-dimensional model, we examined the film growth from the initial surface with hollow parts (wedge-shaped and flat-bottomed holes), the aspect ratio of which ranged from 0.5 to 4. When a wedge-shaped hole is filled with deposited atoms, small voids appear in the film, which are aligned in the growth direction. In the case of the flat-bottomed hole, on the other hand, large voids with several hundred vacant sites appear in the middle of the hole. The voids are elongated in the growth direction as the aspect ratio becomes large. These voids have a similar structure to those of the experiment of filling vias and trenches in Cu interconnections from the solution without additives. The mechanism of the void formation is discussed in relation to the aspect ratio and the overpotential.
Initial deposition rate and appearance of deposited nickel on glass are measured to realize the role of a complexing agent in the electroless plating bath. We evaluated carboxylic acids, hydro carboxylic acids and amino acids as complexing agent. Among these complexing agents, carboxylic acids and hydro carboxylic acids indicated an acceleration effect and amino acids showed a retardation effect on the plating rate. Deposited nickel particles are agglomerated ranging from 10 to 30nm in the early deposition stages. Glycine among the complexing agents showed the smallest particle sizes in the initial deposition stages and nickel was uniformly deposited and distributed on glass. Therefore, this may lead the improvement of the adhesion strength of deposited nickel on glass.
Films of 6-dibutylamino-1, 3, 5-triazine-2, 4-dithiol (DB) were formed on deposited Cr substrates by the conventional vacuum deposition method. The deposited Cr substrates were exposed in air for various times. The adsorption of DB molecules was investigated using XPS to examine the influence on the oxide Cr surfaces. The results showed that C, N and S of DB molecules were bound to the metallic Cr surfaces, and formed disulfide bond. It is assumed that the adsorptions of DB molecules were occurred through the triazine ring to the surface, and formed the polymeric DB on the surfaces. From the XPS spectra, it was shown that DB molecules did not reacted to the Cr oxide surfaces. So it was thought that DB molecules physically react to the oxide surfaces as a structure of monomeric DB. Further, the roughness of DB films was investigated using AFM. It was shown that the DB films were changed into the smooth surfaces with increasing the thickness of the Cr oxide films. It was understood that there were influences in not only the chemical structure of DB molecules, but also the roughness of DB films.
The superelasticity of Ti-48.3, 50.0 and 51.5 at% Ni thin films was evaluated by measuring the stress-strain curves on loading and unloading. It was found that perfect superelasticity is obtainable above the reverse R phase transformation finish temperature for both the Ti-48.3 at% Ni and Ti-51.5 at% Ni thin films, while the Ti-50.0 at% Ni thin film does not show perfect superelasticity. The Ti-50.0 at% Ni thin film was found to be susceptible to plastic deformation because of a single phase. This plastic strain hinders the perfect shape recovery after unloading. In contrast, in the Ti-48.3 at% Ni thin film, crystallization from the amorphous state produces fine distribution of coherent Guinier-Preston zones. These precipitates are considered to impede plastic deformation during stress-induced martensitic transformation, resulting in a perfect shape recovery after unloading, that is, a perfect superelasticity. On the other hand, in the Ti-51.5 at% Ni thin film subjected to solution and aging treatments, fine distribution of Ti3Ni4 was observed and this fine precipitate is considered to be responsible for perfect superelasticity by suppression of plastic deformation.
Conditions that afford effective hydrophobicity to a dry process were examined by preliminarily treating coupling agents, VTES and VTMS, under various conditions and the correlation existing among oligomer formation, water consuming rate and hydrophobicity M value were considered. The main findings obtained are as follows: 1) Hydrophobicity M values increase with the preliminary treatment time of the coupling agent, and the time required for the stand up of M value was shortened by increasing initial water concentration, decreasing the initial solvent alcohol concentration, lowering of the pH value of aqueous phase and elevation of the temperature. 2) The behavior of the M value of TiO2 was greatly affected by oligomer formation in the solution of preliminary treatment. and 3) The existence of an intimate relationship between the half-life of VTES or VTMS monomer in the solution of preliminary treatment and hydrophobicity M value of TiO2 determined at the same preliminary treatment duration was found, thus the half-life was considered an important parameter for establishing conditions that afford effective hydrophobicity to TiO2.