No abstract

A neon ion beam has been used to regrow epitaxially a -1700-A-thick amorphous surface layer in silicon on sapphire at low temperatures. The damaged layer was produced by implanting 80-keV silicon ions to a dose of 2)&10' ions/cm at room temperature. The channeling technique with 315-keV protons was used to investigate the depth distribution of the damage, and disorder depth profiles were extracted from the backscattering spectra using calculations based on multiplescattering theory. The epitaxial regrowth was quantitatively determined from the extracted profiles.Many of the parameters which influence the regrowth rate, such as dose, dose rate, target temperature, energy, and random or channeled direction for the annealing beam, were varied. The results were compared with energy deposition calculations which indicated strongly that the annealing rate depends on the energy deposited in elastic collisions by the annealing ion beam. A defect annealing model based on vacancy diffusion is discussed.

The influence of dose rate on the ion-beam-induced crystallization of amorphous layers in silicon has been investigated. The amorphous layers were produced by self-ion implantation both in bulk silicon and in silicon on sapphire. Subsequent recrystallization was induced at 200 to 400 °C by Ne, Si, Ar, and Kr ion beams of 300 keV energy passing through the amorphous layers. Rutherford backscattering/channeling measurements showed that the regrowth rate decreased with increasing dose rate. This behavior was more pronounced for heavy ions where high dose rates and/or low temperatures could reverse the recrystallization and induce further amorphous growth of the layer. In this new solid-phase growth regime, the amorphous/crystalline interface moved inwards into the crystal in a manner similar to an epitaxial process. An intermittent beam experiment yielded a time constant for the ion beam induced crystallization mechanism of the order of 0.3 s. The time constant and a scaling law for different ions support a model where the planar growth is caused by the accumulation of divacancies in the interface region.

The electron yield from SiOz films on Si was measured as a function of oxide thickness and ion energy for 150 -300 keV H+, He+, N+, Ne+, Ar+, Kr+, and Xe+ ions. For oxide films thicker than 200 A, the electron yield was found to be approximately independent of oxide thickness. The electron yield from the thick oxide increased as a function of the energy deposited D in electronic excitations with approximately the form D ' and not D as expected from theory. Close to the Si02/Si interface, unexpected variations in the electron yield with oxide thickness, dependent on ion mass and energy, were found. Experiments were performed to investigate whether or not the oxide surface was charged by the ion beam, but such effects could not be observed. However, a satisfactory interpretation of the data could be obtained with a model previously suggested for explaining the dependence of the electron yield on the angle of ion incidence. In this model, the positive charge left behind in the oxide by the liberated electrons within the electron cascades of individual ions, causes the probability of escape of electrons to decrease with increasing electron yield.

The secondary-electron yield from polycrystalline aluminum induced by protons, noble-gas ions, and aluminum ions has been measured as a function of ion energy in the range of 10 to 350 keV and as a function of angle of ion incidence in the interval of 0 ° to 70 °. The experimental data are compared with a theory of electron emission from solids proposed in our previous works. The agreement is good providing additional support for the theory.