The measurement of in-plane motion in microelectromechanical systems (MEMS) is a challenge for existing measurement techniques due to the small size of the moving devices and the low amplitude of motion. This paper studied the possibility of using images obtained using a scanning electron microscope (SEM) together with existing motion detection algorithms to characterize the motion of MEMS. SEM imaging has previously been used to detect motion in MEMS device. However, the differences in how SEM imaging and optical imaging capture motion, together with possible interference caused by electrical actuation, create doubts about how accurately motion could be detected in a SEM. In this work, it is shown that existing motion detection algorithms can be used to detect movement with an amplitude of 69 nm. In addition, the properties of SEM images, such as bright edges, complement these algorithms. Electrical actuation was found to cause error in the measurement, however, the error was limited to regions that were electrically connected to the actuating probes and minimal error could be detected in regions that were electrically insulated from the probes. These results show that an SEM is a powerful tool for characterizing low amplitude motion and electrical contacts in MEMS and allow for the detection of motion under 100 nm in amplitude.
Read full abstract