Achieving large stable vertical displacement in surface-micromachined microelectromechanical systems (MEMS)

This thesis describes electrostatic actuation techniques and mechanical design features for realizing large planar analog vertical travel in an electrostatically actuated diffractive mid-infrared optical device, which is robust, both to manufacture, and against pull-in during use. This device, called the Polychromator, is fabricated by polysilicon surface-micromachining and consists of many parallel elements, each 20 microns wide and a centimeter in length. Typically, achieving such a large travel would require prohibitively large gaps and actuation voltages. In order to reduce the actuation voltage and achieve greater travel before pull-in, thinner beams are used which exploit stress stiffening. This, in turn, creates a number of stress control hazards because tensile stress in one layer can induce buckling in a lower layer. These issues have been solved with detailed attention to supports and their compliance. A multi-layer nonlinear spring has been incorporated to make the device robust against pull-in. The electromechanical behavior of the device is simulated using Energy Methods, Finite Difference Methods, and the MEMCAD software. Excellent agreement between MEMCAD simulation and experimental measurements for this structure are reported. Each detail, stress control, support structure, and pull-in, must be addressed in order to achieve the combined effects of large travel, robustness against pull-in, and optically flat beams. Controlled planar actuation over a large vertical range at low applied voltages is obtained by combining a robust electromechanical design with a manufacturable surface micromachining process. Covering a centimeter square area, the 512 grating elements achieved 3.8 microns vertical displacement at 72 Volts with a 0.5 Volt standard deviation, indicating uniform performance across the device. The development of the device has led to innovations in position control, fabrication processes, device design, and device testing. Thesis Supervisors: Stephen D. Senturia Weller Professor of Electrical Engineering Rajeev Ram Associate Professor of Electrical Engineering and Computer Science Thesis Reader: Jeffrey H. Lang Professor of Electrical Engineering and Computer Science