Mastro, Michael A. | Search NASA

DOE OSTI2024

Detecting defects that reduce breakdown voltage using machine learning and optical profilometry

Abstract Semiconductor wafer manufacturing relies on the precise control of various performance metrics to ensure the quality and reliability of integrated circuits. In particular, GaN has properties that are advantageous for high voltage and high frequency power devices; however, defects in the substrate growth and manufacturing are preventing vertical devices from performing optimally. This paper explores the application of machine learning techniques utilizing data obtained from optical profilometry as input variables to predict the probability of a wafer meeting performance metrics, specifically the breakdown voltage (V bk ). By incorporating machine learning techniques, it is possible to reliably predict performance metrics that cause devices to fail at low voltage. For diodes that fail at a higher (but still below theoretical) breakdown voltage, alternative inspection methods or a combination of several experimental techniques may be necessary.

42 ENGINEERING↗

NASA NTRS2006

Recent Results from Epitaxial Growth on Step Free 4H-SiC Mesas

This paper updates recent progress made in growth, characterization, and understanding of high quality homoepitaxial and heteroepitaxial films grown on step-free 4H-SiC mesas. First, we report initial achievement of step-free 4H-SiC surfaces with carbon-face surface polarity. Next, we will describe further observations of how step-free 4H-SiC thin lateral cantilever evolution is significantly impacted by crystal faceting behavior that imposes non-uniform film thickness on cantilever undersides. Finally, recent investigations of in-plane lattice constant mismatch strain relief mechanisms observed for heteroepitaxial growth of 3C-SiC as well as 2H-AlN/GaN heterofilms on step-free 4H-SiC mesas will be reviewed. In both cases, the complete elimination of atomic heterointerface steps on the mesa structure enables uniquely well-ordered misfit dislocation arrays to form near the heterointerfaces with remarkable lack of dislocations threading vertically into the heteroepilayers. In the case of 3C-SiC heterofilms, it has been proposed that dislocation half-loops nucleate at mesa edges and glide laterally along the step-free 3C/4H interfaces. In contrast, 3C-SiC and 2H-AlN/GaN heterofilms grown on 4H-SiC mesas with steps exhibit highly disordered interface misfit dislocation structure coupled with 100X greater density of dislocations threading through the thickness of the heteroepilayers. These results indicate that the presence of steps at the heteroepitaxial interface (i.e., on the initial heteroepitaxial nucleation surface) plays a highly important role in the defect structure, quality, and relaxation mechanisms of single-crystal heteroepitaxial films.

Neudeck, Philip G.↗