The interdiffusion and subsequent compound formation was investigated in thin film couples of Ti–Au, Ti–Pd, and Ti–Pt using transmission and glancing angle electron diffraction techniques. Aging temperatures up to 500°C and times up to 8 h were considered. For all three couples, intermetallic compound formation was observed in the as-deposited films indicating substantial diffusion during deposition (= 100–150°C). For the Ti–Au couples, the TiAu2, TiAu, and Ti3Au compounds formed during diffusion. For the Ti-Pd and Ti-Pt couples the isostructural compounds TiPd, TiPt and Ti3Pd, Ti3Pt were observed. After the formation of the Ti3Pd type compound diffusion ceased for temperatures <500°C indicating that this compound was acting as a diffusion barrier. In all cases Ti was found to be the fastest diffusing species, where it was observed that the diffusion occurred principally through the Au, Pd, or Pt grain boundaries. The effective diffusion coefficient was estimated to be about 105 times larger than expected for bulk diffusion. The role of diffusion in the formation of compounds and the role of film defects in determining the effective diffusion coefficients are discussed.

The growth of platinum silicide layers has been observed using Rutherford backscattering. By simultaneous observation of the relative movement of the Pt2Si and PtSi interfaces, it is possible to deduce that Si is the diffusing species with the reaction occurring at either the Pt2Si–Pt, PtSi–Pt2Si, or PtSi–Pt interfaces. Activation energies for the growth of the Pt2Si and PtSi layers are 1.6 eV. It is possible to interact Pt with Si by means of Ar ion bombardment; however, the reacted layer acts as a barrier to Si diffusion.

The internal stress can be of importance in the tungsten metallization process used for the fabrication of large−scale integrated circuits. The present work is an extension of an earlier study on the dependence of the stress in low−voltage triode sputtered tungsten films upon deposition conditions and substrate materials. As a function of film thickness, the stress was found to decrease with increasing thickness at various substrate temperatures. The effect of higher substrate temperatures is just to change from large compressive stress to smaller compressive stress and finally into tension. For example, the stress in a 5000−Å film decreases from 1.6×1010 dyn/cm2 in compression to 5×109 dyn/cm2 in tension as substrate temperature increases from 370 to 850 °C. Generally, no gross difference was found for films deposited on SiO2, Al2O3, or Si3N4 at higher substrate temperatures. As a function of deposition rate, the stress can be described in three regions. The stress was found to be small and relatively constant in the low−deposition−rate region. However, as the deposition rate increased to the transition region, a sharp stress increase was found. Finally, above a critical deposition rate, the large internal stress remains constant again. By measuring the stress and characterizing the microstructure of the film, it is concluded that the grain size is the dominant factor determining the stress. These experimental results are discussed in light of the grain boundary relaxation model of Finegan and Hoffman and annealing of disorder proposed by Klokholm and Berry.

An electron transmission study of the P-phase of NiAl was made in annealed and deformed samples with the purpose of looking for antiphase boundary (APB), networks, and superdislocations and to observe other configurations of the deformed state. No "grown in" APB networks, superdislocations, or APB trails were observed in annealed or deformed samples. It was concluded that in P-NiAI APB do not exist and the material deforms on slip systems which do not require the motion of superdislocations. Deformation appears intimately related to tangles of single dislocations, subgrains, tilt and twist subboundaries, distorted hexagonal networks, and dislocation loops.Eine Elektronen-Transmissionsuntersuchung der P -Phase von NiAl wurde in getemperten und verformten Proben durchgefiihrt, um Antiphasengrenzen (APB), Netzwerke und Superdislokationen zu finden, und um andere Konfigurationen des deformierten Zustands zu beobachten. I n getemperten oder vcrformten Proben wurden keine "grown in" APB Netzwerke, Superdislokationen oder APB-Spuren gefunden. Es wird geschlossen, daB in P-NiA1 keine Antiphasengrenzen existieren und das Material durch Gleitsysteme deformiert wird, die nicht die Bewegung von Superdislokationen erfordern. Die Deformation scheint wesentlich mit einem Gewirr von Einzelversetzungen, Kristalliten, Stufen-und Schraubenkleinkorngrenzen, gestorten hexagonalen Netzwerken und Versetzungsschleifen verbunden zu sein.

No abstract