Stopping Power and Energy Straggling of Channeled He-Ions in GaN

Abstract

GaN epitaxial layers are usually grown on sapphire substrates. To avoid disastrous effect of the large lattice mismatch a thin polycrystalline nucleation layer is grown at 500• C followed by the deposition of thick GaN template at much higher temperature. Remnants of the nucleation layer were visualized by transmission electron microscopy as defect agglomeration at the GaN/sapphire interface and provide a very useful depth marker for the measurement of channeled ions stopping power. Random and aligned spectra of He ions incident at energies ranging from 1.7 to 3.7 MeV have been measured and evaluated using the Monte Carlo simulation code McChasy. Impact parameter dependent stopping power has been calculated for channeling direction and its parameters have been adjusted according to experimental data. For virgin, i.e. as grown, samples, the ratio of channeled to random stopping power is constant and amounts to 0.7 in the energy range studied. Defects produced by ion implantation largely influence the stopping power. For channeled ions the variety of possible trajectories leads to different energy loss at a given depth, thus resulting in much larger energy straggling than that for the random path. Beam energy distributions at different depths have been calculated using the McChasy code. They are significantly broader than those predicted by the Bohr formula for random direction.

Monte Carlo simulations of ion channeling in crystals containing dislocations and randomly displaced atoms

“…However, its exact value may vary depending on the structure. 21 Possible uncertainty of the depth conversion does not exceed 30 nm as evaluated at the depth of 100 nm and 200 nm during MC simulations of 1.7 MeV He ions channeling in ZnO using the most extreme mismatch of the value of α (α = 1 and α = 0 taken instead of α = 0.5).…”

Further improved technique for channeled stopping power measurements

“…In the light of such a small difference, it seems unlikely that the discrepancy between the measurements reported in the literature and our new results is due to a difference in beam divergence. Recent measurements of the random and channeled stopping powers for He in GaN found an even smaller reduction of only 30% in the energy range 1.7-3.7 MeV [8]. Similarly, a reduction of only (31 ± 5)% of the random stopping power was found at 3.3 MeV using transmission measurements but using much thicker crystals (3.6 lm) [9].…”