The manufacturing of Integrated Circuits (ICs) on semiconductor wafers involves hundreds of complex and expensive process steps. Defects can occur during any of several steps such as etch, resist removal, particle contamination, incomplete process, process variations, dislocations, scratches, cracks, etc. Defects range in size from submicron to visually-detectable “macro” defects that may be as large as several inches long and span multiple dies. While many of these defects are merely “nuisance defects”, some eventually become “killer defects” that result in the die failing final electrical tests, reducing the yield of the semiconductor manufacturing process.
All defect detection tools consist of two main components: (1) Hardware to acquire an image of the wafer surface, and (2) image processing algorithms that analyze the acquired image for defects. Concerning the hardware for acquiring an image, almost all of the existing defect detection tools for macro defects use optics. Optical inspection systems for detecting particles and defects on semiconductor wafers may be divided into two broad classes: (i) laser scanners that detect defects by light scattering, and (ii) optical imaging systems. Traditionally, these tools are configured as specialized stand-alone inspection systems designed to provide sensitivity to extremely small defects and particles, and are thus complex in design, expensive, and have low throughput.
In order to maintain high yield and guarantee the success of high-volume semiconductor manufacturing, it is crucial to detect defects early in the process, to understand their causes, and to take steps to eliminate them. SC Solutions has developed an in-line macro-defect detection tool that satisfies this critical need to rapidly detect macro defects on whole wafers using inexpensive hardware that can be readily integrated into the existing wafer processing equipment. The main idea is the use of inexpensive imaging hardware combined with smart software algorithms that compensate for the limitations of the hardware.