Abstract: A method is described for characterizing defects on a test surface of a semiconductor wafer using a confocal-microscope-based automatic defect characterization (ADC) system. The surface to be tested and a reference surface are scanned using a confocal microscope to obtain three-dimensional images of the test and reference surfaces. The test and reference images are converted into sets of geometric constructs, or “primitives,” that are used to approximate features of the images. Next, the sets of test and reference primitives are compared to determine whether the set of test primitives is different from the set of reference primitives. If such a difference exists, then the difference data is used to generate defect parameters, which are then compared to a knowledge base of defect reference data. Based on this comparison, the ADC system characterizes the defect and estimates a degree of confidence in the characterization.

Abstract: A system for analyzing an object has two operating modes such as scanned imaging mode and stop scan spectral analysis mode. A beam scanner is optically connected to a laser to receive laser beams from the laser. Beam scanner also scans the beams if the system is in the scanned imaging mode. A lens (e.g., an objective lens) is optically coupled to the beam scanner to focus the beams received from the beam scanner. A sensor is optically coupled to the lens to receive the beams that reflect from the object. The sensor may detect characteristics of the beam such as color and intensity. A spectrometer is optically coupled to the objective lens. During stop scan spectral analysis mode, spectrometer generates wavelength spectrum data.