You cannot register for this webinar

Requirements for crystalline defect analysis have increased with the introduction of compound semiconductors in electronics applications. Characterization of crystalline defects during epitaxial growth optimization, followed by defectivity control in the fab is required to sustain device yield at profitable level. In particular, the presence of threading dislocations (TDs) in GaN has been found to have negative effects on the performance of optoelectronic devices. TDs act as centers of non-radiative recombination, leading to current leakage and a decrease in device performance and lifetime. Recognizing the significance of these concerns, there is a need within the semiconductor industry to quickly and comprehensively investigate the presence and type of TDs over large areas (>100 μm²). To address this, we propose a unique method: a semi-automated, non-destructive, high-throughput, and quantitative crystalline defect metrology system for GaN. Our approach is based on the utilization of electron channeling contrast imaging (ECCI). ECCI is a Scanning Electron Microscopy technique, which offers the capability to visualize, classify, and quantify threading dislocations (TDs) based on electron diffraction. Visualization of TDs is achieved by orienting the specimen under specific channeling conditions and detecting the backscattered electrons using appropriate imaging parameters. Determining the dislocation types (a-, c-, a+c type) is also possible by combining data acquired at different diffraction conditions. The identification and quantification of TDs can be performed efficiently, assisted by AI. In addition, denoising techniques on ECCI datasets leads to faster data acquisition compared to the current defect metrology techniques used in the semiconductor industry.