Transformation of Gold in N-Type Silicon from a New Deep Level to the Gold Acceptor Level
- PDF / 518,099 Bytes
- 6 Pages / 420.48 x 639 pts Page_size
- 35 Downloads / 185 Views
TRANSFORMATION OF GOLD IN N-TYPE SILICON FROM A NEW DEEP LEVEL TO THE GOLD ACCEPTOR LEVEL EINAR 0. SVEINBJORNSSON AND OLOF ENGSTROM Department of Solid State Electronics, Chalmers University of Technology, S-41296 Goteborg, Sweden. ABSTRACT Using deep level transient spectroscopy (DLTS) on gold-doped n-type Czochralski (CZ) and float zone (FZ) silicon we observe a new gold-related acceptor level (G) with an activation 2 energy AE= 0.19 eV and an electron capture cross section an = 1.10-17 cm . The center anneals out at a temperature of 250*C, simultaneously as the gold acceptor concentration increases. Annealing at temperatures below 250"C does not reverse this process. However, etching a few microns off the sample surface using HF:HNO 3 based etch reforms the G center and the gold acceptor concentration decreases accordingly. From DLTS depth-profiling we determine that the new center is only found at depths less than 5 gIm, and in the same region we observe neutralization of phosphorus dopants and a reduction in the gold acceptor concentration. We propose that in-diffusion of hydrogen during the etching process is responsible for the three observed transitions, i.e. neutralization of both phosphorus donors and gold acceptors and formation of the G center. We suggest that there are (at least) two possible Au-H complex centers, one which is electrically inactive and another which gives rise to an acceptor level (AE = 0.19 eV) in the bandgap of n-type silicon. The electrically active center anneals out at 250°C while the electrically inactive one is more stable and has been observed earlier in remote plasma hydrogenation experiments performed at 150-350*C. INTRODUCTION Gold is one of the best known impurities in silicon and has been studied intensively for decades. The deep acceptor and donor levels are well documented in the literature although many of the published results are controversial [1]. The levels are believed to be two charge states of the same defect, normally assigned to substitutional gold [2,3]. Gold is also known to form complexes with other metal impurities in silicon. These complexes are formed at low temperature, typically at 200-300°C, and break up between 300-500"C [4-6]. Hydrogen passivation of the gold acceptor and donor levels was first reported by Pearton et al [7,8] where the hydrogen was induced by remote plasma hydrogenation or electrolytic charging. The levels were passivated to depths less than 10 g.tm and were only partially regenerated by a subsequent vacuum anneal at 400"C. In this paper we report studies on a gold related deep center which can be transformed irreversibly into a gold acceptor by annealing at 250"C. The center appears after wet chemical etching and is only found at depths less than 5 p.m. In the same region we observe passivation of the phosphorus dopants and gold acceptors. The phenomenon has a close resemblance to effects reported on hydrogen passivation and we propose that the new center is an electrically active gold-hydrogen complex. EXPERIMENTAL DETAILS
The sam
Data Loading...
