Metrology

In today's global marketplace, productivity is the bulwark that keeps the flood of competitors at bay. To increase productivity, companies seek to lower overhead, such as labor and research and development (R&D) costs, while increasing high-quality yields. The combined result is increased profitability, regardless of whether the company makes aspirin or amplifiers.

One of the threads that ties together labor, R&D, production yields, and profitability, is test and measurement; supporting these test systems is the science of metrology. Automated metrology like machine vision systems reduce labor costs for quality inspection, while high-resolution profilometers, interferometers, atomic force microscopes, and similar systems enable the manufacture of complex systems like electronics and optical systems, where tolerances exceed human sensory abilities. The acceptance of internationally approved standard units means that production lines in Guam and Spain can produce the same high-quality product, giving manufacturers geographic flexibility. Automated metrology turns the one-off experiments of an R&D laboratory into efficient batch operations, while similar metrology systems feed production data back into manufacturing lines for real-time yield improvements. To most manufacturing quandaries, metrology is the answer.

Despite the critical and growing need for metrology, getting CFOs to sign off on new systems is still an uphill battle. Metrology is not viewed as a revenue generator but rather a cost center. The aging, short-term view is that metro-logy costs a lot but doesn't produce revenue by it-self. The good news is that, in recent years, industries have been becoming believers in quality as the best way to differentiate a company from competitors—and metrology as the engine that delivers it.

Although revenue trends and sales data on metrology used in the manufacture of optics is scarce, several high-point presentations at SPIE's recent 2005 Optifab conference (Rochester, NY) emphasized the connection between metrology and manufacture. "The new equipment that's being introduced for the manufacturing and measurement of aspheres is pretty exciting, maybe one of the most exciting developments in a number of years," exclaims Don Golini, president of QED Manufacturing Inc. (Rochester, NY) and Optifab steering committee member. "Most of the talks were about manufacturing improvement and applications oriented across materials, fabrication, design, and metrology," adds John Schoen, director of the Center for Optics Manufacturing at the University of Rochester (Rochester, NY).

As public education and acceptance of metrology business drivers improve, these systems continue to face technical challenges arising from the "need for speed" in manufacturing and new applications in micro- and nanostructures. "The benchmark has always been interferometry in one way or another," explains Bob Fischer, CEO of Optics1 (Westlake Village, CA), "but it takes a fair amount of time. People have defaulted to profilometry . . . but it doesn't sample as well as you'd like it to."

Combining throughput with high-resolution metrology is pushing metrology system designers to integrate multiple methods into a single system to deliver the breadth of data manufacturers need. (See "Measurement in Micro," page 17, and "Getting to the Heart of the Matter," page 20.)

Unfortunately, adding capability comes at a price. As dimensions shrink into the micro- and nanoscale, the line separating metrology's technical challenges from business challenges begins to fade. New imprint lithography methods that can produce features far smaller than those of optical lithography methods are languishing because of a lack of suitable, commercial metrology systems that can reproducibly measure what imprint lithography produces, slowing the development and acceptance of this important technology.

In the meantime, microchip makers and their equipment providers that use traditional lithographic and optical metrology systems are already bending under the burden of increasing equipment costs, as revenue growth lags behind the increasing cost of designing new manufacturing and metrology systems.

According to Dan Hutcheson, CEO of semiconductor analyst VLSI Research Inc. (Santa Clara, CA), the cost of microchip manufacturing, including metrology, is leading to an "R&D crisis." In a recent research note, Hutcheson says that semiconductor equipment makers will spend less on R&D than microchip makers for the first time in 30 years, pushing the burden of R&D back on chip makers who are also feeling pricing pressures.

"Chip sales have been growing at around 6% a year since 1995," Hutcheson writes, "but R&D is growing at a 12% CAGR. By 2020, RD&E spending rate would reach 40%. Clearly, this is unaffordable." Hutcheson predicts that—if equipment costs continue along the present path—either the rate of new product introduction must slow, which would lead to a downward spiral in the microchip industry, or "the industry must become more efficient in R&D."

Hutcheson offers his own answer, suggesting greater collaboration among system developers and better preparation for new lithography systems so that costs do not spike. With $51 billion at stake—based on 2004 sales of semiconductor equipment, including metrology and inspection systems—perhaps the carrot of success will overcome the fear of collaboration. oe