Microrelay

Our goals in this project were to (1) develop a new design concept for a high reliability microrelay, (2) build a prototype, and (3) demonstrate high force relay closure in the prototype. During FY1999, we designed a microrelay to meet commercial specifications: 3 g (or 0.03 N) closure force and 30-mA actuation current at less than 0.5 V. Our microrelay not only occupies less than 1 mm{sup 3}--about 1% of the volume of the smallest commercial part--but also its fabrication takes advantage of semiconductor processing, which has the potential to automate microrelay production. Conventional relays are fabricated by assembling many discrete parts. The process includes a number of nonautomated assembly and inspection steps, which increase fabrication cost and limit possible size reductions. Microrelays based on electrostatic forces can be fabricated by thin-film techniques employed in the semiconductor industry; however, the voltages required to make reliable electrical contact in an electrostatic relay significantly increase the cost of the driver. Microrelays based on electromagnetic forces, on the other hand, provide reliable contacts at low voltage. Reliable metal-to-metal contacts require sufficient contact force to plastically deform contact surfaces at asperities-thereby increasing the contact area. On the other hand, contact metallurgy and the gaseous environment must be controlled to prevent contact welding, contamination, oxidation, and other effects that change contact resistance over time. A contact force of 3 g is commonly used with gold/gold-alloy contacts in a sealed relay (e.g., a reed relay). In this way, more than 10 million closures can be achieved with a resistance of less than 100 m{Omega}. Our prototype relay preserves the contact metallurgy of commercial relays. The fundamental innovation in the fabrication of our microrelay is the use of a 3-D lithographic process to create a ''winding'' around a discrete magnetic core. To achieve sufficient inductance to generate the desired contact force, we chose a discrete core of substantial cross section (about 0.3 mm{sup 2}). Because of the core thickness, thin films deposited on it cannot be patterned by conventional lithography but can be patterned by our 3-D process. The microrelay is formed on a single substrate so that critical core-to-armature distance can be precisely defined using a thin sacrificial layer. The initial separation of the core and armature is about 10 {micro}m. The issue of greatest significance to the performance of the relay is the dimensional precision of relay closure--the electrical contacts must touch when the armature and the core (which define the magnetic circuit) are separated by only 1 {micro}m. Defining a manufacturable process which can achieve this goal has been the triumph of this year's development effort. To define the design of our prototype microrelay, we performed both 1-D analytic and 3-D numerical modeling. Photos of the prototype in fabrication are shown in Figure 1. The prototype differed in a number of ways from the design due mainly to problems in fabricating the iron core. The prototype served its purpose, however, verifying the design concept and demonstrating the closure forces required