A high-sensitivity and low-noise MEMS accelerometer was developed using multi-layer metal structures composed of multiple metal layers. The accelerometer achieves 1 μG level resolution that has been a challenging task with conventional MEMS technology. The team previously proposed a method of using gold material to downscale the proof mass size of MEMS accelerometers to less than one-tenth. In this work, they have increased the proof mass per area by employing multiple layers of gold for the MEMS structures, thereby increasing the sensitivity over 100 times and reducing noise less than one-tenth when compared with conventional accelerometers.
In the design of accelerometers, there is a tradeoff between the size reduction and the noise reduction because the mechanical noise dominated by the Brownian noise is inversely proportional to the mass of the moving electrode, called proof mass. For capacitive accelerometers, the sensitivity is generally proportional to the accelerometer size, and thus there is also a tradeoff between the size reduction and the sensitivity increase. Since high-resolution accelerometers require low-noise and high-sensitivity performance, it has been difficult for conventional silicon-based MEMS accelerometers to detect 1 μG level input acceleration.
Multi-layer metal structures were employed to the proof mass and spring components and a low-noise and high-sensitivity MEMS accelerometer was developed. The Brownian noise, being inversely proportional to the proof mass, was reduced by increasing the mass per area with the use of multiple layers of gold for the proof mass structure. The 4-mm-square chip area was used by reducing the warpage of the proof mass, which enabled an increase in the capacitance sensitivity of the accelerometer.
The new accelerometer achieved sensitivity of 100 times or more and the noise one-tenth or less as compared with conventional accelerometers with the same size. Accordingly, the accelerometer could have potential to detect input acceleration as low as 1 μG. The fabrication process utilized semiconductor microfabrication process and electroplating and could implement the developed MEMS structures on an integrated circuit chip.
For more information, contact Assistant Professor Daisuke Yamane at