• PDF / 298,743 Bytes
• 6 Pages / 416.88 x 635.4 pts Page_size
• 26 Downloads / 286 Views

DOWNLOAD

REPORT

---

ABSTRACT Recent advances in the preparation of gallium nitride (GaN) and related compounds have made possible the production of blue semiconductor laser. Conventional preparation involves growing GaN thin films on lattice-mismatching sapphire using metal-organic chemical vapor deposition (MOCVD). In this article, we describe an alternative method to produce a latticematching strained layer in GaAs for subsequent GaN growth by plasma immersion ion implantation (PM) followed by rapid thermal annealing. Our novel approach uses broad ion impact energy distribution and multiple implant voltages to form a spread-out nitrogen depth profile and an amorphous surface layer. This approach circumvents the retained dose and low nitrogen content problems associated with ion beam implantation at fix energy. Based on our Raman study, the resulting structure after PIMI and rapid thermal annealing is strained and contains some GaN possibly in crystal form. INTRODUCTION The wide band-gap semiconductor based on the nitride compounds of the Ill-V group has attracted much attention because of their potential applications in optoelectronics in the visible spectrum [1] and high temperature electronic devices [2]. At present, the majority of the commercial GaN-based devices use sapphire as the substrates. Although the inherent 15% lattice mismatch with GaN makes sapphire an unlikely substrate for high quality crystal growth, it has been found that one can grow a sequence of ultra-thin GaN buffer layers at low temperatures to confine the cracks to a narrow region [1]. On the surface of this "transition" layer, high quality GaN can be grown at normal temperature for device applications. Using this idea, it is possible that good quality GaN growth can be achieved if a surface that contains a thin layer of polycrystalline GaN can be prepared. Nitrogen ion implantation into a Ga-containing surface followed by an annealing process is a possible approach that can result in the formation of a GaN buffer layer. Pertaining to the formation of GaN via ion implantation, several groups so far have reported results on their exploratory experiments [2-7]. From their TEM results, Lin et al. showed that in GaAs, the heavily N÷-implanted surface contained nano-size crystals of cubicand hexagonal-GaN after annealing at 850TC [3]. Kuriyama et al. recently also demonstrated that N÷-implanted GaAs exhibits photoluminescence in the blue region, thus further confirming the prospect of producing functional GaN by ion implantation [4]. The group led by Sealy [5,6] adopted an alternative approach. They implanted Ga into Si3 N4 and found from Auger measurement that when the starting layer was nitrogen-rich, the final material 309

Mat. Res. Soc. Syrnp. Proc. Vol. 618 © 2000 Materials Research Society

contained GaN. The basic rationale associated with these experiments is that the Ga-N bond strength is stronger than that of Ga-As or Si-N and it should be energetically favorable for the formation of GaN if the two constituent atoms are available and in adequate proxi