A microelectromechanical systems chip, sometimes called "lab on a chip"

In the past decade the miniaturization of personal technology has irreversibly changed the way we live our lives. Prior to the smartphone market boom that followed the Blackberry, the iPhone and other subsequent portable computing platforms, a common tech dream was to reduce the number of devices we carry to one. Making phone calls, using the internet, location and navigation services, taking pictures -- for the average user, having three separate devices to accomplish these tasks is a thing of the past. This dramatic technological change has largely come from a reduction in size of the various components that go into making these devices. This class of device is generally referred to as micro-electromechanical (MEM), and comprises devices that interact both electronically and mechanically on roughly micrometer-sized scales (1-100 micrometers, or 0.001-0.1 millimeters). Conceptually, devices on this size scale have been in the minds of ambitious inventors for the better part of the last century, but only since the invention of silicon manufacturing technologies for computer chips has MEMs manufacturing come to make economic sense. Now, as MEMs technology seeks to expand beyond the limitations of silicon-based manufacturing, manufacturers struggle with markets that are too small to support high production costs.

MEMs are on the forefront of what researchers are able to accomplish through assembling physical systems. As computers drove down the cost of manufacturing silicon chips for processors, researchers began to realize that they could use these microscopic manufacturing technologies to fabricate mechanical systems on increasingly small scales as well. Early uses included inkjet printers, which typically use a small piezoelectric mechanism to create a pressure which forces ink from the printer heads, were among the first to produce devices that could be classified as MEMs. Accelerometers, one of the most common modern uses of MEMs because of their ubiquitous usage in smartphones, were soon developed and manufactured for use in automobile airbag crash detection systems. These widespread uses have long been the exception rather than the rule, and the market for manufactured MEM devices has not previously been large enough to motivate significant development in manufacturing technology. Most MEM technologies have relied on silicon manufacturing because the technology and infrastructure was already there from the semiconductor industry's need for silicon microchips.

Synthetic detail of a standard cell through four layers of planarized copper interconnect, down to the polysilicon (pink), wells (greyish) and substrate (green) Today, it is rapidly becoming clear that the market for MEMs will soon be sufficient to drive MEM-specific manufacturing technology forward. In a recent MIT news article, Larry Hardesty focuses on issues that MEM manufacturers will face as market forces begin to shape this technological coming of age. In particular, many silicon-based MEM manufacturing technologies are not scalable and do not promote the sort of industrial innovation necessary to respond to market demand. Compared to the wide swath of potential uses for MEMs that researchers are working on, accelerometers are a very simple device, and most are created using outdated silicon etching technologies, with most of the costs arising from the material. Polymers, ceramics and metals are all cheaper and more effective materials for most MEM applications, but they are not used because the manufacturing infrastructure is not there as it is for silicon.

These technologies are likely to open up a whole world of MEM capabilities beyond the basic physical sensing and mechanical motion uses that are already widespread. Polymers are one promising example, and offer a wide variety of tunable material properties that would allow uses far beyond the capabilities of silicon. Metals are another, with radically lower raw material costs and a number of other potential benefits over traditional silicon systems. Both of these materials are much more conducive to microfluidics applications -- micro-scale fluid-based MEM systems. Developed manufacturing processes using these materials could lead to many new far-reaching MEM applications. One exciting example is in biological applications, from biosensing to drug delivery. For the moment, though, economies of scale make silicon almost universally appealing.

Now, as the economy emerges from the financial crisis and companies are once again interested in manufacturing, all of this is likely to change. Though MEMs experienced a serious decline in 2008-2009, in a report this past March, market research publisher Global Industry Analyst forecasted the MEM market to grow to $11.3 billion in the next five years. As the market grows I expect that consumers will want their devices to do even more, and manufacturing companies will soon yield to market pressure and invest in developing new and innovative manufacturing technologies.