AVS 45th International Symposium
    The Science of Micro-Electro-Mechanical Systems Topical Conference Monday Sessions
       Session MM+PS-MoM

Invited Paper MM+PS-MoM3
Materials, Process, and Integration Issues in SiC MEMS

Monday, November 2, 1998, 9:00 am, Room 324/325

Session: MEMS Processing and Deep Si Etch Technology
Presenter: M. Mehregany, Case Western Reserve University
Correspondent: Click to Email
SiC MEMS technology holds great promise for applications which are characterized by presence of harsh environments (e.g., high temperatures, large number of vibrational cycles, erosive flows, and corrosive media). Historically, SiC research has focused on the materials and processes needed for high-temperature and high-power microelectronics. These devices require high-quality single crystal films and substrates, which lead most researchers to use 6H-SiC, since nearly defect-free wafers and epitaxial films are available. Unfortunately, high quality comes at a high price; 6H-SiC wafers are very expensive and are available only in small wafer diameters. Thus, applications for 6H-SiC devices are limited to areas which can absorb such high costs, such as (military) aircraft and spacecraft. Our work has been motivated by the necessity to develop a low-cost SiC MEMS technology to penetrate a much more diverse set of markets, including for example automotive. Additionally, we have been motivated to leverage off of the latest fabrication process technologies available from Si to push the SiC MEMS technology further, faster. As a result, we have pursued large-area substrates, i.e., 3C-SiC on Si. Unlike 6H-SiC, 3C-SiC is the only SiC polytype which can be heteroepitaxially grown on Si substrates. Heteroepitaxy on Si gives 3C-SiC a distinct advantage over 6H-SiC in terms of batch fabrication, since high quality, large-area Si substrates are readily available and comparatively very inexpensive. We have pursued the development of bulk and surface micromachining processes using 3C-SiC and poly-SiC, respectively. Heteroepitaxy of 3C-SiC on Si is attractive to MEMS not only for batch fabrication, but also for bulk micromachining. In fact, SiC is undoubtedly an excellent etch stop material for Si bulk micromachining, since Si anisotropic etchants such as KOH and EDP are impervious to SiC. We have used Si bulk micromachining techniques to fabricate a multitude of 3C-SiC structures, ranging from diaphragms for mechanical properties studies, pressure sensors, and optical transmission windows, to cantilever beams and torsional micromechanical structures. Bulk micromachining of 6H-SiC has been demonstrated by others, however the process is much more complicated and the dimensional control and etch stop capabilities are limited at this time. Unlike electronics applications which require high-quality single crystal films, MEMS is much more flexible in that structures can be fabricated from polycrystalline films. SiC MEMS is no exception. We have developed poly-SiC as a basic structural material for SiC MEMS. We have deposited APCVD poly-SiC films atop polysilicon and silicon dioxide sacrificial films on 4 inch diameter Si wafers. We have demonstrated SiC surface micromachining processes, and these have been used to fabricate the first SiC lateral resonant structures. These devices tested at temperatures up to 900C outperformed polysilicon resonators of like geometry with respect to high temperature capability. Of course, the surface micromachining technology using poly-SiC would be extendable to 6H- and 4H-SiC substrate technology, as well as integration with SiC electronics on these substrates. An overview of materials, process, and integration issues in SiC MEMS will be presented, including current device examples.