Role of Defects in the Nucleation of Whiskers Growing from Vapor

Abstract

The role of substrate surface defects in the nucleation of whiskers in the process of vapor deposition is a topic of long standing in the literature. In fact, the beginning was the diffusion-dislocation model of Sears [1] which was an attempt to relate the nucleation and one-dimensional growth of whiskers to the non-disappearing step created by a screw dislocation on a substrate surface. For some time this model was widely accepted because of a number of experimental data which supported the presence of defects in whiskers either directly or indirectly. However, the model had weaknesses which until recently had to be tolerated. The inconsistency of this model was demonstrated by Wagner [2] who substituted the diffusion- dislocation model by the diffusion-droplet model based on the vapor-liquid-solid (VLS) mechanism of whisker growth [3]. Wagner pointed to a good deal of conclusive evidence supporting the VLS mechanism. Many other facts could be mentioned, all pointing to the important role played by impurities in the nucleation of whiskers. Thus, in the reduction of iron halides the iron whiskers grow on magnesium segregations [4]. Selective nucleation of whiskers of iron oxides was observed on the surface of Armco iron doped by bismuth deposited through a mask [5]; the whiskers nucleated and grew in druses only at the impurity sites. The VLS growth of copper whiskers (reduction of halides by hydrogen), with drops of molten salt at the whisker tip, has been observed directly by in situ electron microscopy [6].