ISOTOPE-SEPARATION AND GROWTH MECHANISMS OF INTERMETALLIC PHASES - AN INVESTIGATION OF NICKEL SILICIDES BY SECONDARY ION MASS-SPECTROMETRY

Isotope separation resulting from the diffusion-controlled growth of reacted films has been proposed as a tool to elucidate the diffusion mechanism which controls the growth. A "reverse" isotope separation would be indicative of a grain boundary diffusion mechanism. Such an isotope separation has been studied by secondary ion mass spectrometry (SIMS) during the growth of nickel silicides (Ni2Si and NiSi) for different experimental growth conditions: silicon substrate orientation, temperature of reaction, and silicide thickness. In every case, the isotope separation resulting from the diffusion-controlled growth is below the accuracy of the SIMS measurements. This is consistent with the fact that isotope effects are usually very small. The difficulties of such studies and the multiple SIMS artifacts encountered at interfaces - matrix effects, variations in the sputtering yield, mass interferences, and sequential acquisition of data - are analyzed in detail. They are found to be quite important in the case of nickel silicides and may lead to erroneous conclusions. A model for isotope separation during layer growth, which had been previously proposed, is shown to provide overoptimistic estimates of the magnitude of the anticipated effect.