The field of micromechanics is rapidly expanding in both the number of research groups and the number of materials being employed. Although this diversity is a strong indication of a healthy field, care must be taken to keep the focus on producing products and processes which can be transferred to the manufacturing facility. During the 1980s polysilicon was shown to contain a significant amount of process flexibility and economic potential. Although the demonstration of polysilicon sensors was noticed and in some cases pursued by industry, single-crystal silicon sensors continue to dominate the products used by the primary sensor customer, the automotive industry. A similar trend which also began in the 1980s was the use of the LIGA process for sensor fabrication. Once again, this field showed a significant amount of economic promise. However, for the resources being invested and the number of research groups pursuing this process, significant problems exist with respect to product manufacturability. Although LIGA remains an exciting field of research, new micromechanical processes or materials may greatly reduce the window of device profitability before the difficulties associated with LIGA can be brought under control. Oddly enough, it is the same material which polysilicon has failed to displace which may limit the LIGA process to only one application area, that area being magnetics. New deep anisotropic etching systems and new substrate suppliers for micromachining applications, along with the knowledge and experience industry already possesses, will maintain single crystal silicon as the sensor material of choice for the 1990s and beyond. This article reviews the material stability and processes associated with polysilicon, single crystal silicon, and electrodeposits. Emphasis will be placed on the inherent material structure and processes required to manufacture a profitable device.
|