Atom and carrier concentration profiles in carbon-ion-implanted GaAs have been measured. Ion implantation of carbon is performed at 300 keV with dose of 1.0×1014 ions/cm2. Carbon concentration profile obtained by secondary ion mass spectrometry measurement is in good agreement with the profile obtained by Monte Carlo simulation. The implanted carbon does not diffuse markedly with annealing at 900° C because the diffusion coefficient is below 4×10-16 cm2/ s for the ion-implanted carbon. Therefore, a shallow carrier concentration profile is formed after annealing. Activation efficiency is 17% at the surface (depth less than 0.47 µ m). However, this efficiency is as low as 4% in deeper regions. The lower activation efficiency in deeper regions is due to the suppression of activation by the precipitation of carbon after the annealing.

Damage density measurement for ion-implanted silicon is studied, where the density is measured using Rutherford backscattering spectrometry (RBS) and a newly developed photoacoustic displacement (PAD) technique. Since correlation is obtained between the observed PAD and quantitative damage density in ion-implanted silicon at 150 keV at a dose of 2.0×1015 ions/cm2, damage density can be determined quantitatively from the PAD measurement at the surface. A damage density depth profile can also be obtained using the differential PAD measurement. The observed profile agrees well with the ones of the displaced atom density obtained by RBS measurement and Monte Carlo simulation.

Damage formed by low-dose Si+ implantation is studied. Variation of damage density with dose in Si+ implantation at 50 keV is measured at doses from 3.0×1010 to 1.0×1014 ions/cm2 by the photoacoustic displacement (PAD) technique. A close correlation has also been found between the PAD value and ion implantation dose in low-dose ion implantation. Ion implantation dose can be monitored down to 3.0×1010 ions/cm2 by this technique. This dose detection limit is much lower than that of other methods. The in-depth damage profile can also be measured by differentiating the observed PAD values with depth.

No abstract