In-Situ RHEED Observation of MOCVD-Gan Film Growth

  • PDF / 766,331 Bytes
  • 6 Pages / 612 x 792 pts (letter) Page_size
  • 29 Downloads / 221 Views

DOWNLOAD

REPORT


IN-SITU RHEED OBSERVATION OF MOCVD-GAN FILM GROWTH Masatomo Sumiya, Noritaka Ogusu, Kouhei Osada, and Shunro Fuke Department of Electrical &Electronic Engineering., Shizuoka University, 3-5-1 Johoku Hamamatsu, 432-8561, Japan ABSTRACT We developed the MOCVD apparatus equipped with RHEED system, which enable us to observe in-situ and real time RHEED for GaN film growth in ~100mTorr of pressure. We attempted to grow GaN film with this MOCVD chamber in 100mTorr. The in-situ RHEED was subsequently observed along the film deposition process in order to understand both the role of buffer layer and the mechanism of GaN film growth by MOCVD on highly lattice-mismatched substrate like sapphire. The results indicate that oxygen removed from the sapphire surface was observed during its cleaning in H2 flow at 1100ÂșC. The dependence of re-crystallization and evaporation of the buffer layer on the annealing ambient was also detected. Although the nitrogen was slightly deficient, HT-GaN film with smooth surface was obtained in 100mTorr by adding H2 gas and reducing total flow rate. In preliminary deposition, the RHEED oscillation-like was observed in MOCVD-GaN growth. Thus, our developing deposition system has a potential to understand the growth mechanism with atomic level. INTRODUCTION GaN and related materials have been applied not only to light emitting devices [1,2] but also to electronic devices. Although their application has been intensively developed, the fundamental understanding such as the growth mechanism of highly-lattice mismatched system seems to be left behind. GaN film is conventionally deposited on c-plane sapphire substrate through several processes in metalorganic chemical vapor deposition (MOCVD); (1) substrate treatment (H2 cleaning and/or nitridation), (2) deposition of low-temperature (LT) buffer layer, (3) ramping up the substrate temperature (annealing of LT buffer layer), and (4) deposition of high-temperature (HT) GaN film. Since the condition in each process can be independently controlled and it gives an influence on the subsequent process, it is very difficult to optimize many process parameters. For example, after the substrate nitridation, HT-GaN is likely to have N-face (-c) polarity. However, when LT-buffer layer on nitrided sapphire is deposited under Ga-rich condition, the film is likely to have Ga-face (+c) polarity. We have investigated GaN growth mechanism related to the polar structure with respect to the role of LT-buffer layer in MOCVD method [3]. It was found that the substrate nitridation, V/III ratio of LT-buffer deposition, and the annealing of LT-buffer layer were essential for controlling the polar structure. We have considered the method to observe how the film growth should take place in each MOCVD process with in-situ and atomic level. There are several in-situ measurement techniques of the growth such as the temperature oscillation [4], the reflection of laser light [5], and the elipsometry [6]. However, these methods cannot observe with atomic level. Recent work done by Fini e