The effect of cascade mixing on profile broadening during secondary-ion mass spectrometry (SIMS) analysis has been thoroughly investigated for ${\mathrm{Al}}_x{\mathrm{Ga}}_{1-x}\mathrm{A}\mathrm{s}/\mathrm{G}\mathrm{a}\mathrm{A}\mathrm{s}$ structures of five different compositions ($x=0.1,$ 0.3, 0.5, 0.73, or 1) and layers with varying thicknesses (from one monolayer to 1000 Å). The SIMS analyses were performed using primary sputtering ions of ${}^{20}{\mathrm{Ne}}^{+},$ ${}^{40}{\mathrm{Ar}}^{+},$ ${}^{84}{\mathrm{Kr}}^{+},$ and ${}^{136}{\mathrm{Xe}}^{+}$ with an impact energy (E) ranging from 1.8 to 13.2 keV and an angle of incidence, with respect to the surface normal (θ), from 62° to 35°. Within the experimental accuracy, the decay length of the trailing edge was found to be proportional to $E^{1/2}\cos \theta$ where the proportionality constant displays a relatively weak dependence on primary ion mass. However, the leading edge is strongly affected by the extension of the collision cascade as demonstrated by a comparison of the results for the different ions at a given energy. As long as the cascade is fully developed before reaching an interface no dependence on the sample depth is obtained for the profile broadening. Furthermore, the decay length for the trailing edges is extracted in the dilute limit and no effect of the marker thickness or the x value is revealed. A numerical treatment of the profile broadening within a diffusional model, where the diffusion coefficient is assumed to be proportional to the energy deposited in elastic collisions, gives a surprisingly good agreement with the experimental data.

• Received 6 May 1999

DOI:https://doi.org/10.1103/PhysRevB.60.14302