The mighty mites of microelectronics

DEFINITION: MicroElectroMechanical Systems (MEMS) technology enables the creation of tiny machines that can work with microelectronics. MEMS also refers to the machines themselves, which add very small sensors and actuators that allow microchip-controlled systems to sense and control their environments.

“The very rich are different from you and me,” wrote F. Scott Fitzgerald. Similarly, very small things operate under different constraints than the machines and tools we’re accustomed to seeing and using in everyday life.



In the microscopic world, the standard principles of classical physics don’t always hold true. Factors such as weight, inertia and thermal mass become less important, while forces related to surface area, including friction and surface tension, take on much greater significance. Think about how an ant is able to carry something many times its own weight or a water bug can skate across the surface of a pond. It’s a different world down there.

We humans can take advantage of these differences through MEMS, in which tiny, sometimes microscopic mechanical elements are created on silicon chips using fabrication technologies similar to those used for making integrated circuits.

MEMS — also referred to as micromechanics or micromachines — can refer to both the technology and the specific devices, which range from a micrometer to a millimeter in size. Smaller devices are called NanoElectroMechanical Systems (NEMS) or simply nanotechnology. MEMS and NEMS operate using the principles of mechanics in tiny versions of the classical simple machines (inclined plane, wheel and axle, lever, pulley and screw). Today, MEMS refers to almost any miniaturized device, regardless of whether it’s based on silicon technology or traditional precision engineering.

Why Go Small?

Just as birds can fly but elephants can’t, MEMS devices can easily do things that are problematic for larger ones. For example, sensors and actuators are the most costly and least reliable parts of many human-scale machines and control systems. Large devices can’t move as quickly or as precisely as microscale machines; precision in large machines is expensive and often difficult to mass-produce.

In contrast, MEMS technology allows us to create complex electromechanical systems that can move, position, regulate, pump and filter, and to manufacture them in quantity using batch fabrication techniques originally developed for the semiconductor industry. This puts the cost and reliability of such small sensors and actuators on a par with those of electronic integrated circuits, enabling superior performance at much lower cost.

How to Make MEMS

MEMS are typically made on a silicon substrate using some of the same microfabrication technologies (indeed, sometimes the same machinery) designed to produce microprocessors and other electronic integrated circuits.

The micromechanical components are fabricated using micromachining processes that selectively remove parts of the silicon wafer or add new structural layers to form a variety of mechanical and electromechanical devices.

We speak of bulk micromachining, where the entire thickness of the silicon wafer is used, and surface micromachining, in which a thin layer of silicon is etched to make mechanical structures beneath it movable.

Three basic processes are used to make MEMS: deposition, photolithography and etching. Deposition adds thin films of material onto a substrate, using electroplating, physical and chemical vapor deposition, and sputtering (in which metal atoms are knocked off a target of pure metal with ions from a plasma and then deposited on the substrate). These deposited materials can be polymers or metals such as gold, nickel, aluminum, chromium, titanium, tungsten or silver.

A dust mite puts the size of MEMS devices into perspective.  Courtesy of Sandia National Laboratories, SUMMiTTM Technologies,

A dust mite puts the size of MEMS devices into perspective.
Courtesy of Sandia National Laboratories, SUMMiTTM Technologies,

Photolithography is used with both deposition and etching processes to transfer the pattern or blueprint of the device onto the substrate. The substrate is first coated with a photosensitive material, after which parts are covered up with a mask. It can then be selectively exposed to a radiation source (typically light). The exposed region will have different properties than unexposed areas and can then be removed or treated.

In wet etching, material is removed selectively by dipping it into a chemical solution that can dissolve it. In dry etching, also called reactive-ion etching, the substrate is put inside a reactor vessel, several gases are introduced, and a plasma is created by a radio frequency power source. Plasma ions are accelerated toward the surface of the material, where they react to form other gases.

In addition to this chemical process is a physical process similar to sputtering, in which high-energy ions knock atoms out of the substrate without a chemical reaction taking place. Balancing chemical and physical etching rates allows different effects, including sidewalls with different profiles, from rounded to vertical.

A newer etching process uses alternate injections of two different gases, creating etch-resistant polymers on sidewalls that concentrate removal on only horizontal surfaces. This allows deep etching, which can cut completely through the silicon substrate.

Some other processes — ones not derived from semiconductor technology — are used in MEMS fabrication. These include molding and electrical discharge machining.


Although the technology is still in its infancy, MEMS devices are found in many places, including ink-jet printers, large-screen TVs, airbag deployment systems, and pressure sensors used in car tires and some blood-pressure systems. MEMS are also used in the motion-sensing controller featured in Nintendo’s Wii video game system.

In telecommunications, MEMS has become nearly synonymous with the arrays of tiny tilting mirrors used for optical switching fabric.

With its need to reduce the size and weight of objects lifted into space, NASA uses MEMS for microgyroscopes, microthrusters, mass spectrometers and other devices.

See also: More on MEMS

Kay is a Computerworld contributing writer in Worcester, Mass. You can contact him at

Are there technologies or issues you’d like to learn about in QuickStudy? Send your ideas to

See additional Computerworld QuickStudies


Copyright © 2007 IDG Communications, Inc.

Shop Tech Products at Amazon