Point Defects in SiC
- PDF / 594,065 Bytes
- 12 Pages / 612 x 792 pts (letter) Page_size
- 27 Downloads / 245 Views
1069-D03-01
Point Defects in SiC Ádám Gali1, Michel Bockstedte2, Ngyen Tien Son3, and Erik Janzén3 1 Department of Atomic Physics, Budapest University of Technology and Economics, Budafoki ut 8, Budapest, H-1111, Hungary 2 Universität Erlangen-Nürnberg, Staudtstr. 7/B2, Erlangen, D-91058, Germany 3 Linköping University, Linköping, S-581 83, Sweden ABSTRACT Tight control of defects is pivotal for semiconductor technology. However, even the basic defects are not entirely understood in silicon carbide. In the recent years significant advances have been reached in the identification of defects by combining the experimental tools like electron paramagnetic resonance and photoluminescence with ab initio calculations. We summarize these results and their consequences in silicon carbide based technology. We show recent methodological developments making possible the accurate calculation of absorption and emission signals of defects. INTRODUCTION Point defects in semiconductors are pivotal for tailoring the electronic properties of materials according to needs. In silicon carbide (SiC), substitutional nitrogen or phosphorus are used for n-type doping while aluminum or boron for p-type doping. Deep defect centers compensate shallow impurities and produce semi-insulating regions in the semiconductor. These deep defect centers may be introduced un-intentionally during the technological process of doping that conteracts the efforts of the tight control of electronic conductivity of SiC. Recently, intrinsic deep defect centers have been started to apply to produce semi-insulating (SI) SIC substrate that can efficiently compensate nitrogen or boron impurities in the sample. The role of the intrinsic defects is crucial in both cases. However, even the basic intrinsic point defects are not entirely understood in SiC. The SiC electronic device problem is still a “defect engineering” problem. The first step is to identify the defects in order to control them. Computational physics and experiments are complementary tools to attack the same problem in defect physics: computational physics can determine lots of properties of a single defect (correlation of “signals”) but it cannot follow the complex processes occurring in the experiments and the accuracy of the obtained numbers are usually an order of magnitude higher than those coming from the experimental spectrum. In experiments one can characterize the defect accurately (sharp spectrum) but due to the difficulties of making correlations measurements (either due to the difficulties of sample preparation for two different experimental techniques or the complex processes occurring during the measurement) the interpretation of the signal is not straightforward and hinders the correct identification of the defect center. However, the combination of the two approaches can be very fruitful, at least, to disregard the inappropriate models for defect centers, and more: it can lead to the identification of defects. In the next sections we will summarize the computational methodologies that ha
Data Loading...
