[14a-A410-1~10] 15.6 Group IV Compound Semiconductors (SiC)

▲ [14a-A410-1] Optimization of furnace structure for 300 mm SiC solution growth by machine learning

〇Wancheng YU¹, Can Zhu¹, Shunta Harada^1,2, Tagawa Tagawa^1,2, Toru Ujihara^1,2,3 (1.IMaSS, Nagoya Univ, 2.Nagoya Univ., 3.GaN-OIL, AIST)

Keywords:machine learning, SiC, solution growth

300 mm larger scale SiC TSSG furnace is designed in this paper for the next generation SiC growth. However, much time is required to find out the optimized structure as up to 12 parameters should be considered to design a new furnace. Therefore, we utilize machine learning to accelerate the simulation process. Under the guidance of machine learning, Si-Cr solvent with temperature difference at seed surface of 0.4K and temperature difference at crucible sidewall and bottom can be achieved.Except for furnace design, machine learning is also powerful in the optimization of growth receipt. The growth of 6-inch SiC crystals in our group is in progress under the guidance of machine learning, high quality crystal with a large thickness can be expected in a short time.