Historically, bulk micromachining, in which a single-crystal substrate is formed into a micromechanical element, comes before surface micromachining. In surface micromachining, deposited layers over the substrate are machined into mechanical elements. However, the first surface micromachined device (Nathanson et al. 1967) was coincidentally the first actuated micromechanical device. In the early 1980s, when surface micromachining made startling advances as a consequence of utilizing materials and processes that had been well characterized by the integrated-circuits industry, once again the first surface micromachined device was demonstrated as an actuated microdevice (Howe and Muller 1986). Both of these microactuated devices relied on Coulombic forcing functions that could easily be activated by changing voltage levels.
An expanding horizon for actuated micromechanisms has helped to propel research and development on increasingly sophisticated devices and microsystems. Microactuators are becoming recognized as nearly indispensable elements for force-balanced systems, for the control of tiny movements, for self-adjusting systems to accommodate environmental and material-aging variations, and for precise measurement systems such as those employing resonators.
The economical production of microactuated elements integrated into sophisticated systems will open very large areas of opportunity in microsensors, microoptics, microfluidics, and microrobotics. The great potential of microactuated elements has become recognized, and research and development on microactuators in the United States is intense and growing.
In the United States, research at universities has played an important role in developing microactuation. In the mid-eighties, there were identifiable research projects on microactuated devices at the University of California at Berkeley, the University of Utah, Stanford University, and the Massachusetts Institute of Technology. In 1987, an NSF panel identified movable micromechanisms as being a very significant development in a publication called Small Machines, Large Opportunities. The NSF publication introduced the word "microdynamics," which was used later as a stand-alone title for a keynote lecture at Transducers '89 in Montreux, Switzerland (Muller 1990).
As the mid-nineties approach, a growing list of campuses, in addition to the universities named above, have targeted research on microactuated devices using diverse technologies and focusing on a variety of applications. These universities include: Case Western Reserve University, Cornell University, Georgia Institute of Technology, the University of California at Los Angeles, the University of Michigan, and the University of Wisconsin.
In industry and at the national laboratories, there has been lesser publication of microactuation ideas -- both because much work is still in progress and also because publication is certainly not an industrial priority. Microactuation does, however, play an important role as a force-balancing means in microaccelerometers as described by Analog Devices, Motorola, IC Sensors, and Lucas NovaSensor. The digital-mirror display device of Texas Instruments qualifies as an important application for microactuation, as does the work of small companies like Redwood Semiconductors, which is developing microvalves using fluidic phase-change principles (Zdeblick and Angell 1987).
Actuating forces that are under the most study at this time for applications to microdynamics include: electrostatic, resonant impulse transfer, fluidic phase change, magnetic, pneumatic, piezoelectric, ultrasonic, thermal bimorph, piezoelectric bimorph, and shape memory alloy.
In the United States at present, there is research on devices employing virtually all of these prime movers; however, the focus is not spread uniformly across the list. Electrostatic actuation is naturally in a favorable position if sufficient force can be gained from it for microactuation because electrostatic forcing can be so comfortably integrated with microelectronics. Thus, for example, two major MEMS for industry, the microaccelerometer of Analog Devices (Payne and Dinsmore 1991) and the modulated mirror array of Texas Instruments (Sampsell 1993a), employ electrostatic forcing either to balance accelerative forces (in the first case) or to move tiny mirrors in an array (in the second). Electrostatic drive has also been used to power a very tiny microgripper made from polycrystalline silicon (Kim et al. 1990) with a maximum opening to 20 mm.
To provide a snapshot of the broad activity in microactuation in the United States at present, a bibliography of recent papers describing trends in U.S. research on microactuation is included on pages 285-287. The listing includes papers on comb drive, resonators, rotating micromotors, friction and sticking, magnetic drive, ultrasonic drive, piezoelectric drive, and microactuated electrooptic devices. The first section provides a view of recent research on comb-drive devices, both surface micromachined (Fan and Crawford 1993; Tang, Lim, and Howe 1989) and using monocrystalline silicon (MacDonald 1992). Following the first demonstration of electrostatic comb drives by W.C. Tang and R.T. Howe (Tang, Nguyen, and Howe 1989), these drivers became a prime power source for microresonators. Comb drivers are used for two of the resonators in the next grouping of papers (Lee, Ljung, and Pisano 1990; Nguyen and Howe 1993) while the work of Boustra et al. (1992) at Michigan employs a piezoelectric driver for a resonating cantilever.
Continued research in the United States on rotating micromotors has produced high rpm devices (above 10,000 rpm), and showed ways to reduce frictional effects that have been found to dominate the mechanics for these devices. Activity on rotating micromotors has been strong at Case Western Reserve University. Two papers on microdynamical friction provide some picture of present research on its control. Four papers are cited that show the growing efforts on magnetic actuation, which has been an area of concentration at the University of Wisconsin and at Georgia Institute of Technology. A paper by Busch-Vishniac at the University of Texas contrasts microactuation with other prime movers for microdynamics (1992). The research on ultrasonic Lamb waves at the University of California has shown the possibilities for using this means for moving solid elements and also for pumping and mixing fluids; Moroney, White, and Howe are authors of a paper describing this research (1991). A growing focus on piezoelectricity derived from thin films is highlighted in three papers that are complemented by the Boustra paper (1992) cited in the group of papers on resonators. The last section of the bibliography contains two papers from Carnegie Mellon University and from Stanford University that highlight an emerging concentration of research on microelectrooptical devices.
Figure D.5. SEM view of a four-port electrostatically-driven lateral micromechanical resonator developed at the University of California at Berkeley (Nguyen and Howe 1992).