Design and Analysis of Curved Electrode Configurations for Enhanced Sensitivity in 1-Axis MEMS Accelerometers

TL;DR

This study analytically and via simulation investigates curved electrode geometries in MEMS accelerometers, showing that certain shapes like biconvex improve sensitivity significantly without increasing device size.

Contribution

It introduces analytical models for various curved electrode profiles and validates them with simulations, identifying geometries that enhance sensitivity without size increase.

Findings

Biconvex electrodes offer the highest sensitivity improvement.

Sensitivity increases monotonically with arc length for certain shapes.

Concave and plano-concave designs can invert output polarity.

Abstract

This paper presents a comprehensive analytical and simulation-based study of curved electrode geometries for enhancing the sensitivity of MEMS capacitive accelerometers. Expressions for the capacitance between a planar movable electrode and six distinct fixed electrode profiles (biconvex, biconcave, concavo-convex, convexo-concave, plano-convex, and plano-concave) are derived, enabling direct calculation of differential gain and sensitivity as functions of electrode curvature and gap displacement. These analytical models are then rigorously validated using finite element simulations performed using COMSOL Multiphysics under identical bias and boundary conditions. The simulation results demonstrate agreement with the analytical results with a deviation of less than 7% in all configurations. The results also reveal that biconvex curved electrodes yield the greatest sensitivity improvement over the planar electrodes, with sensitivity monotonically increasing with arc length, while concave and plano-concave designs exhibit reduced performance. The concavo-convex and convexo-concave configurations furthermore introduce polarity inversion in the output voltage, offering additional design flexibility. Importantly, these sensitivity enhancements are achieved without any change in the overall volumetric dimensions of the device or the proofmass dimensions of the module for achieving higher-resolution inertial sensing.