As the volume of consumer electronics increases, semiconductor fabrication plants are manufacturing larger and larger wafers to handle the demand (e.g., 300 mm substrates). The escalating value of these larger wafers is driving the industry to employ more advanced imaging technologies for quality control. Inspecting the raw material substrate for flaws before processing and detecting defects during processing is critical to keeping costs down. New and improved inspection techniques save the semiconductor manufacturing industry hundreds of millions of dollars each year.
Material quality is important in all semiconductor applications. High material quality at the start of wafer processing is critical to reducing waste. Dislocation defects in the material can be very detrimental to a semiconductor device. Users can now image these defects by passing shortwave infrared light through a semiconductor wafer and imaging it with an uncooled InGaAs camera. The light diffracts ever so slightly at these large defects, allowing the user to image the location and determine the number of these defects in the wafer. The quality control inspectors are then able to prevent poor quality wafers from being processed. Figure 2 shows how the wafer is transparent to SWIR light (left) versus visible light (right) where the wafer is opaque.
Many semiconductor devices are processed using both sides of the wafer. Alignment difficulties may occur because the wafer is opaque to visible light. Since the wafers become transparent at longer wavelengths, SWIR cameras are able to see through them, while using standard glass optic microscopes. The ability to quickly inspect the alignment of the backside pattern to the front side is a noted benefit to using SWIR imaging. This is in addition to being able to observe processing defects in the semiconductor that can be otherwise obscured by the metal layers on the top surface of the integrated circuit. Another advantage is that InGaAs-SWIR cameras utilize glass optics instead of more expensive Germanium or other OR-optimized optics, making them simple to integrate into existing wafer processing fabs.
Silicon, GaAs, InP and other semiconductor materials are not only being used in the microelectronics industry but also in the optoelectronics industry. Silicon waveguides are being utilized in telecommunication applications and future work will include the microelectronics integrated with the optical control. These waveguides direct light through the material to various points but the waveguides do have losses. InGaAs cameras can image the light being used in the waveguides, typically 1.3 and 1.5μm, easily allowing the user to image the losses as well as the total output. Mating the waveguides with light emitters by viewing through the waveguide permits active, automated alignment, improving yield and driving down costs. SWIR cameras and photodiode arrays, developed with indium gallium arsenide technology from SUI, Goodrich Corporation (Fig. 3) are becoming increasingly important diagnostic tools for both the telecom and semiconductor manufacturing markets.
InGaAs cameras over the last couple of years have become more sophisticated with more resolution, sensitivity, and greater capability (processing power) in a smaller, lighter form factor. This has allowed the semiconductor industry the ability to capitalize on this new technology and to integrate it into its testing, inspection, and quality control systems. As the cameras become more integrated in this industry they will find even more applications.
This article was written by Martin H. Ettenberg, Ph.D., Director of Imaging Products, and Douglas S. Malchow, Manager of Commercial Business Development, at Sensors Unlimited Inc. (Princeton, NJ). For more information, contact Mr. Ettenberg at