A Three-Axis CMOS-MEMS Accelerometer Structure With Vertically Integrated Fully Differential Sensing Electrodes

This study presents a novel CMOS-microelectromechanical systems (MEMS) three-axis accelerometer design using Taiwan Semiconductor Manufacturing Company 0.18-μm one-poly-Si six-metal/dielectric CMOS process. The multilayer metal and dielectric stacking features of the CMOS process were exploited to vertically integrate the in-plane and out-of-plane capacitive sensing electrodes. Thus, the three-axis sensing electrodes can be integrated on a single proof mass to reduce the footprint of the accelerometer. Moreover, the fully differential gap-closing sensing electrodes among all three axes are implemented to increase the sensitivities and decrease the noise. The in-plane and out-of-plane sensing gaps are respectively defined by the minimum metal line width and the thickness of one metal layer by means of the metal wet-etching post-CMOS process. Thus, the capacitive sensitivities are further improved. The fully differential gap-closing sensing electrodes also bring the advantage of reduced cross talks between all three axes. As a result, the footprint of the presented three-axis accelerometer structure is only 400 × 400 μm². Compared with existing commercial or CMOS-MEMS studies, the size is significantly reduced. The measured sensitivities (nonlinearities) are 14.7 mV/G (3.2%) for the X-axis, 15.4 mV/G (1.4%) for the Y-axis, and 14.6 mV/G (2.8%) for the Z-axis.

I. Introduction

The microelectromechanical systems (MEMS) accelerometer is an important device for motion sensing and has found various applications in automobile, industry, and consumer electronics [1]. The three-axis accelerometer (-, -, and -axes) has been extensively exploited to detect the linear motion of an object in space. The three-axis MEMS accelerometer has also been frequently used to detect the inclined angles of bulky structures in the industrial field. Other industrial applications using multiaxis MEMS accelerometers are vibration and motion monitoring [2], robot positioning and control [3], etc.