Transportation of Na and Li in Hydrothermally Grown ZnO
- PDF / 179,162 Bytes
- 6 Pages / 595 x 842 pts (A4) Page_size
- 72 Downloads / 252 Views
1201-H02-05
Transportation of Na and Li in Hydrothermally Grown ZnO Pekka T. Neuvonen1, Lasse Vines1, Klaus M. Johansen1, Anders Hallén2, Bengt G. Svensson1 and Andrej Yu. Kuznetsov1 1 Centre for Material Science and Nanotechnology, Department of Physics, University of Olso, P.O. Box 1048 Blindern, N-0316 Oslo, Norway 2 Royal Institute of Technology, School of ICT, Dept. of Microelectronics and Applied Physics, P.O. Box Electrum 229, SE-164 40 Kista, Sweden ABSTRACT Secondary ion mass spectrometry has been applied to study the transportation of Na and Li in hydrothermally grown ZnO. A dose of 1015 cm-2 of Na+ was implanted into ZnO to act as a diffusion source. A clear trap limited diffusion is observed at temperatures above 550 °C . From these profiles, an activation energy for the transport of Na of ~1.7 eV has been extracted. The prefactor for the diffusion constant and the solid solubility of Na cannot be deduced independently from the present data but their product estimated to be ~ 3 × 1016 cm -1s -1 . A dissociation energy of ~2.4 eV is extracted for the trapped Na. The measured Na and Li profiles show that Li and Na compete for the same traps and interact in a way that Li is depleted from Na-rich regions. This is attributed to a lower formation energy of Na-on-zinc-site than that for Li-on-zinc-site defects and the zinc vacancy is considered as a major trap for migrating Na and Li atoms. Consequently, the diffusivity of Li is difficult to extract accurately from the present data, but in its interstitial configuration Li is indeed highly mobile having a diffusivity in excess of 10-11 cm2s-1 at 500 oC . INTRODUCTION Zinc oxide (ZnO) is a wide (~3.4 eV) and direct energy-bandgap semiconductor with high exciton binding energy (~60 meV) [1], which makes it highly desirable material for optoelectronic applications [2, 3]. In addition, large scale manufacturing of ZnO wafers with good crystal quality has been accomplished. ZnO has a tendency to be natively n-type. The origin of the native n-type conductivity still remains a topic of discussion. Previously, it has been suggested to be zinc interstitials (Zni) or oxygen vacancies (Vo) [4-6], but recent theoretical estimates [7] show, that the formation energies of these native defects are too high, or the energy level lies too deep in the bandgap, for these defects to be of main responsibility for the n-type conductivity. The role of hydrogen [8] and other impurities, like Al and Ga [9], has been widely studied, but unambiguous evidence for any of these individual impurities as being the only cause for the n-type conductivity, has not been presented. Partly due to this native behavior, the p-type conductivity has proven difficult to achieve. For instance, elements exhibiting shallow enough acceptor levels in the bandgap are scarce, or the solubility of the dopant is too low to overcome the native n-type conductivity. In addition, several acceptor-like impurities have a tendency to change configuration when the Fermi-level position (EF) varies and become donor-like a
Data Loading...
