The Problem of Doping Wide Gap II-VI Compound Semiconductors and Its Solutions
- PDF / 377,082 Bytes
- 7 Pages / 420.48 x 639 pts Page_size
- 43 Downloads / 154 Views
THE PROBLEM OF DOPING WIDE GAP II-VI COMPOUND SEMICONDUCTORS AND ITS SOLUTIONS W.I. WANG Department of Electrical Engineering Columbia University, New York, NY 10027 ABSTRACT Wide gap II-VI compound semiconductors are difficult to be doped amphoterically. After more than thirty years of research in II-VI compound semiconductors, there does not even exist a satisfatory simultaneous explanation as to why ZnSe can be easily doped n-type while undoped ZnTe only exhibits p-type conductivity. In this paper we propose an explanation based on the III-V/II-VI analogy which for the first time can explain these phenomena, and provide solutions to the problem of doping II-VI compound semiconductors. The wide band gap II-VI compound semiconductors such as ZnSe and ZnTe have direct band gaps and in principle can yield very efficient blue/green lasers. Such visible lasers would have a tremendous impact on a wide range of applications including optical recording and display. There have been many efforts in this field for more than thirty years. Many excellent review articles [1,2] summarized the difficulties in this field: it is difficult to dope the materials both nand p-type. ZnSe can only be doped n-type easily, whereas undoped ZnTe exhibits p-type conductivity and has not been doped n-type. The difficulty in producing amphoteric doping has been attributed to compensation by native defects [3] such as Zn or Se vacancies during crystal growth: it is difficult to dope ZnSe p-type due to excess Se vacancies (which act as a donor to compensate the p-type dopant) and it is difficult to dope ZnTe n-type due to excess Zn vacancies (which act as an acceptor to compensate the n-type dopant). The problem with this widely accepted model is that there has never been any conclusive evidence of the existence of simple Zn or Se vacancies. Also, this model contradicts the fact that it is now relatively easy to grow very high purity ZnSe by molecular beam epitaxy (MBE) and metal-organic chemical vapor deposition (MOCVD) to the point that only band-edge excitons can be routinely observed without any defect-related structures [4-6]. If ZnSe were really plagued by excess Se vacancies as suggested for many years, it would have been impossible to grow extremely low defect ZnSe. After more than thirty years of research, there still does not exist a solution to this doping problem. In this paper we believe that we have been able to correctly
Mat. Res. Soc. Symp. Proc. Vol. 228. (-1992 Materials Research Society
320
explain the difficulty of the doping problem of II-VI compounds for the first time. Our explanation actually is based on our experience in GaAs and GaSb as will be explained later. In addition, we will provide solutions to this problem, and our solutions also come from III-V compounds. In the III-V compound family, undoped high purity GaAs is either very lightly n-type or p-type, depending on the residual background impurity being donors or acceptors [7]. However, undoped GaSb is always p-type [8]. It is wellknown that such a p-type con
Data Loading...
