Time to take the road less traveled?

Over the past several decades, new and powerful capabilities that allow manufacturing at the micro and nano scales have emerged. These developments have been driven largely by the microelectronics revolution and fueled by major scientific advances in materials science, solid-state physics, optics and computation—to name a few.

The enormous market for integrated circuits (ICs) has enabled cooperative development of reliable materials, tools and processes that improve at an exponential pace. In recent years, it has led to many semiconductor manufacturers and foundries jointly developing processing technologies in order to reduce manufacturing costs. We consider this to be a unique paradigm for manufacturing, and propose to define this as the “semiconductor-manufacturing paradigm.” In what follows, we outline the characteristics of this paradigm, and its benefits and limitations.

The paradigm has several characteristics. The first is an ever-increasing economy of scale for manufacturing realized by the increased volume of material processed in each fabrication facility (fab). This scale is reached by increasing substrate (wafer) size in each generation of technology and increasing fab production rates (wafers per month).

The second characteristic is standardization of the manufacturing process, which enables creation of semiconductor foundries that manufacture ICs of different designs for a diverse array of customers. This saves a foundry from having to make capital investments for individual customers. In the same way, fabless IC companies can provide unique IC-enabled products to diverse markets without the constraint of the capital investment required to create a fab.

The last unique characteristic of the semiconductor-manufacturing paradigm is the highly coordinated approach taken by those in industry to develop a collective technology plan—called the “semiconductor road map.” The IC industry comes together on a regular basis to define the next generation of semiconductor technology. This road-mapping has enabled materials and equipment suppliers to sharply focus their developmental efforts and allowed IC designers to anticipate the impact of next-generation technology on their markets and products. This coordination of technology development has also enabled industry to speak with a highly effective voice that drives government investment in research.

While the semiconductor-manufacturing paradigm has been driven by the integrated circuit, other products have, to varying degrees, leveraged this manufacturing paradigm. Notable examples are inkjet printers, inertial sensors, microphones, micromirror arrays and solar panels.

However, with the exception of these examples, the impact of the semiconductor-manufacturing paradigm beyond the IC industry has been limited. If a new micro or nano device requires a unique manufacturing process—one different from the standard IC process—the capital and developmental costs of semiconductor manufacturing make manufacturing these devices economically viable only if the market size is very large.

In the MEMS field, one often hears the phrase “one product, one process, one package,” which speaks to the historic difficulty of achieving standardized processes and packaging technologies. Today, if the one product has a market too small to justify a large capital investment in a dedicated fab line, the product is not likely to come to market.

As a result, new manufacturing paradigms are needed to address opportunities that either do not follow the economic map of semiconductor manufacturing or are fundamentally incompatible, such as fabrication on flexible substrates. We have already seen new manufacturing paradigms emerge in the microfluidics and 3-D printing industries.

Other paradigms are evolving for diamond-tip-based nanomanufacturing—under a new program launched by DARPA (Defense Advanced Research Projects Agency)—and for novel self-assembly methods used to create nano- and micro-structured surfaces.

It is instructive to step back from these examples and ask, “What are the unique attributes of the new manufacturing paradigm, and what coordinated efforts accelerate its development?” First, we must shift from a focus on integrated processes to development of robust, transferable unit processes. (Alissa Fitzgerald made an interesting case for this viewpoint in her column in the January/February 2011 issue “MEMS makers: Machinists who work in silicon.”)

Second, the industry must develop tools that cost an order of magnitude less than those currently available. This may be attainable with a new approach—one that embraces the fact that manufacturing throughput or other performance specifications for this class of tools can be different than tools optimized for IC manufacturing.

Lastly, the attributes of this new micro/nano manufacturing industry should look much more like the machine tool industry than the semiconductor industry. This begs the question of whether organizational principles of the semiconductor industry should be used as models.

If a manufacturing paradigm like this were to emerge, it would have the potential to enable distributed manufacturing—as compared to global consolidation into particular regions as is the case with state-of-the-art semiconductor manufacturing. Many argue that innovation in products that exploit emerging technologies is greatly enhanced when manufacturing expertise is colocated with design and innovation expertise. In this respect, a distributed manufacturing model would ensure that micro- and nano-enabled product innovation would not be concentrated in particular areas.

A collective recognition of the need for new manufacturing paradigms at the micro and nano scales should help define the roles stakeholders at universities, government research agencies and companies can play in this paradigm. The recently announced Micro-Nano Manufacturing conference (see below) is one example of working toward this goal.

To paraphrase the poet Robert Frost, perhaps now is the time that we need to look toward “the road not taken”— thus far! µ

Editor’s Note: The “Technologies for Future Micro-Nano Manufacturing” conference will take place August 8-10, 2011, at the Silverado Resort, Napa, Calif. Visit www.mfg2011.org or e-mail info@mfg2011.org.

About the authors: Dr. Martin A. Schmidt (left) is Professor of Electrical Engineering and Computer Science and Associate Provost at the Massachusetts Institute of Technology. Telephone: (617) 253-7817. E-mail: schmidt@mtl.mit.edu. Dr. Thomas Kenny (right) is professor of mechanical engineering at Stanford University. Telephone: (415) 725-3805. E-mail: kenny@cdr.stanford.edu.