SiC MOSFET has been moving from planar to trench-type structures to achieve better tradeoff between breakdown voltage and on-resistance [1]. In the semiconductor industry, SiC power devices are already in volume production and the next wafer size of 200mm will be available to the market within years [2]. With every increment in the wafer size and device complexity, the process control of SiC power devices will become more challenging and critical. One example is the measurement of deep trenches which requires fast and accurate in-line monitoring of the trench depth, width, angle, and other geometries depending on the actual device. The accuracy, precision and speed of the structural measurement will ensure better process control hence better yield.
High accuracy, high throughput, and non-destructive optical critical dimension (OCD) metrology has become mainstream in semiconductor development. These measurements are based on scatterometry to derive shape, dimension and composition information from the scattering patterns observed in light that has interacted with the sample [3]. In this work, we will demonstrate the capability of OCD metrology solutions for the measurement of critical control parameters in SiC power devices, ie, trench depth, width (CD), side wall angle, and gate oxide thickness, etc.
The OCD measurements are performed with a metrology system equipped with advanced oblique spectroscopic ellipsometry (SE) and normal-incidence reflectometry (TE/TM). The true multi-variant metrological modeling (such as real structure construction, optical constants extraction, sensitivity analysis, next-generation real-time regression, measurement optimization, etc) was achieved using proprietary software for fast and accurate calculations. The measurement results show that the proposed OCD technology has great sensitivity to SiC trench profile features with low inter-parameter correlations, clear design of experiment (DOE) wafer discriminations and consistent wafer maps.
In summary, using an OCD metrology solution to detect within-wafer and wafer-to-wafer process variation matches accurately with sub-nanometer process changes to provide superior process control of SiC power devices. It has full capability to measure complex and critical SiC devices with high accuracy, precision and HVM production worthy throughput.