Dimpling—A new manifestation of ion-produced lattice damage

Publisher Website
Google Scholar
Abstract
Bombardment of thin (1–10 μ) single crystal targets with energetic ion beams has been found to result in macroscopic distortion of the thin film in the bombarded region. This effect, which has been euphemistically termed a ‘dimple’, is readily observed with the naked eye even at relatively low particle fluence. A useful first-order model has been developed which interprets the dimpling as an expansion of the bombarded region. For very thin samples, this expansion can be accommodated by bowing of the crystal out of the original crystal plane. For this simple model, the fractional expansion is proportional to (δ/d)2 where δ is the maximum displacement from the original crystal plane and d is the diameter of the bombarded area. This measurement allows expansioh to be determined with a sensitivity comparable to or better than the most sensitive existing methods. For silicon about 3 μ thick bombarded by 1.8 MeV He ions, the expansion increases essentially linearly with fluence at the lowest fluence (below about 1015 to 1014 cm−2). In this region about 0.001 atomic volumes are added per incident ion. As the fluence is increased, the apparent expansion begins to increase more rapidly than linearly but approaches a saturation value at the highest fluence (about 1018 cm−2). The effect of particle flux, incident energy, and bombardment temperature is discussed as well as some preliminary results on C ion produced dimples in Si and the behavior of Ge samples. Thirty-minute isochronal annealing of silicon samples irradiated with He ions at room temperature shows reverse annealing for temperatures up to about 200–300°C. The dimple begins to anneal at 300°C and disappears after annealing to 600–700°C. After irradiation, the dimpled region absorbs light more strongly in the visible and near IR region. This disappears with annealing before the dimple itself completely disappears and is felt to be largely due to scattering from defect clusters.