IEEE SENSORS JOURNAL, VOL. 10, NO. 8, AUGUST 2010 1311 Toward Sensitivity Enhancement of MEMS Accelerometers Using Mechanical Amplification Mechanism Assaf Ya’akobovitz and Slava Krylov Abstract—We report on the novel architecture and operational principle of microelectromechanical accelerometers, which may lead to enhanced sensitivity achieved through mechanical amplifi- cation of a proof mass displacements. The integrated amplification mechanism serving also as a linear-to-angular motion transformer is realized as an eccentric elastic torsion link that transforms small out-of-plane motion of the proof mass into significantly larger motion of a tilting element whose displacements are sensed to extract acceleration. The design parameters as well as the amplification ratio were elaborated using a lumped model and numerical finite element simulations. The device was fabricated from silicon-on-insulator wafer and is distinguished by a robust single-layer architecture and simple fabrication process. The devices were operated using electrostatic and inertial actuation of the proof mass combined with the optical sensing. Theoretical and experimental results, which are in a good agreement with each other, indicate that the motion amplification scheme realized in the framework of the suggested architecture results in larger de- tectable displacements and could be efficiently used for sensitivity improvement of microaccelerometers. Index Terms—Accelerometer, amplification mechanism, micro- electromechanical systems (MEMS), silicon on insulator (SOI). I. INTRODUCTION T HE field of microelectromechanical systems (MEMS) sensors grew constantly, and an effort has being made toward making smaller and more precise sensors. In particular, MEMS accelerometers became a core part of many engineering systems, starting from consumer electronics, such as digital cameras, cellular phones, computer games, and airbag systems, and up to high-end devices for sophisticated navigation and guiding systems for defense applications. The increasing in- terest in inertial microsensors is explained by their unique and attractive features, including small-dimensions and low-weight, low-energy consumption, low-cost mass production and com- patibility with ICs environments. Though great deal of progress was achieved during the last years, improving the performance of such devices, especially for navigation applications, remains challenging. Manuscript received June 22, 2009; revised November 19, 2009; accepted December 20, 2009. Date of current version published May 21, 2010. The asso- ciate editor coordinating the review of this paper and approving it for publication was Dr. Patrick Ruther. A. Ya’akobovitz is with the Faculty of Engineering, Tel-Aviv University, Tel- Aviv 69978, Israel (e-mail: assafyaa@eng.tau.ac.il). S. Krylov is with the Department of Solid Mechanics, Materials and Systems, Tel-Aviv University, Tel-Aviv 69978, Israel (e-mail: vadis@eng.tau.ac.il). Digital Object Identifier 10.1109/JSEN.2009.2039751 The MEMS accelerometers usually consist of a proof mass suspended using an elastic element (spring). When the device is accelerated, an inertial force is applied to the proof mass, resulting in its deflection in the direction opposite to the ap- plied acceleration. The acceleration can be extracted either from the measuring of the stress in the suspension elements [1] or by registering the displacement of the proof mass [2]. First ac- celerometers to appear were fabricated using bulk microma- chining [1], [3]. At first, piezoresistive sensing was commonly used [1], [4], [5], but due to its relative simplicity, low-cost and high-accuracy, capacitive sensing has become increasingly pop- ular [2], [3], [6]. Over the years, the use of surface microma- chining has become common due to its compatibility with ICs technology as well as the ability to create small dimension ele- ments and small electrostatic gap in capacitive sensors [7], [8]. However, surface micromachined acceleroemters usually char- acterized by relatively thin layer, resulting in small proof mass, which is necessary for high sensitivity. Some studies reported the combination of both surface and bulk micromachining in order to benefit from the advantages of surface micromachining in addition to large proof mass, typically obtained in bulk mi- cromaching, usually at the expense of more intricate fabrica- tion process [9]. In recent years, the use of single-crystal silicon combined with silicon-on-insulator (SOI) technology and deep reactive-ion etching (DRIE)-based process became popular due to excellent mechanical properties of single-crystal silicon and high reliability and robustness of SOI-based MEMS devices. In addition, the use of SOI substrates with relatively thick and bulky handle layers is beneficial in inertial sensors where larger proof mass and consequently thicker layers are required for su- perior performance of the devices [10]. Very large variety of architectures and operational principles of microaccelerometers were reported, and the literature on the subject is voluminous [11]. The devices differ by the proof mass motion direction (in-plane and out-of-plane linear and tiling [2], [6]), suspension architecture (bending and torsional axes, magnetic and electrostatic levitation), main structural material (polysilicon [12], single-crystal silicon [2], [13], metal [14]), and sensing approaches, which include mainly capacitive, op- tical, and piezoresistive sensing [13], [15], but also thermal [16], magnetic [17], and tunneling [18], [19] sensing. Note that in order to be able to sustain environmental shocks and vibrations as well as to meet requirements of wider band- width, the natural frequency of the accelerometer has to be suf- ficiently high. On the other hand, sensitivity requirements dic- tate lower spring constants to achieve detectable deflections of 1530-437X/$26.00 © 2010 IEEE