| Budget Amount *help |
¥11,900,000 (Direct Cost: ¥11,900,000)
Fiscal Year 2000: ¥5,600,000 (Direct Cost: ¥5,600,000)
Fiscal Year 1999: ¥6,300,000 (Direct Cost: ¥6,300,000)
|
|---|
| Research Abstract |
This study focused on the research and development of Selective Area MOVPE process, which may be a good candidate to produce monolithically integrated optical devices known as OEIC. During the growth of InGaAsP based compound semiconductor thin films by MOVPE, film growth will occur only on the crystal surface, but not on the dielectric mask which is partly deposited on the GaAs or InP substrate. This process is called selective growth and in the case of compound semiconductor thin films, the growth rate will be enhanced by the area ratio of growth region and mask region. This is the reason why we call it Area Selective process and this is also the reason for utilizing this process to fabricate OEIC. SA-MOVPE has a potential to produce OEIC by making a suitable mask design to fabricate it. However, the lack of kinetic information makes it quite difficult, so that many researchers are making try & error efforts to make the devices. In this study, we investigated the gas-phase and surfac
… More
e reactions which may control the SA-MOVPE process.
First, we employed CFD (Computer Fluid Dynamics) simulation to analyze the growth rate and composition profile in a commercial reactor from the inlet to the outlet. We have developed a reaction model that can explain the growth rate and composition profile in CFD calculation. The model proposed here could explain the growth rate and In/Ga ratio, which may explain the accuracy of our model. The As/P ratio could not well explained by this model, so there still remain a problem, but it is not so a serious issue, because most of the profile is determined by the group-Ill elements distribution.
The growth rate and composition non-uniformity is also informative to analyze the reaction chemistry. When we grow InGaP, InGaAs, or InGaAsP, the In/Ga ratio in the selectively grown region has non-uniformity. We could analyze the reactive sticking probabilities of In species from the growth rate non-uniformity of InP growth and the reactive sticking probabilities of Ga species from growth rate non-uniformity of GaAs growth. We could find that the non-uniformity of In/Ga ratio was originated from the difference of sticking probabilities between In and Ga species. These kinetic information can be used to design optimum mask pattern to develop OEIC without any experimental trials. Less
|
