Micro 3D printing of a functional MEMS accelerometer

Abstract

Microelectromechanical system (MEMS) devices, such as accelerometers, are widely used across industries, including the automotive, consumer electronics, and medical industries. MEMS are efficiently produced at very high volumes using large-scale semiconductor manufacturing techniques. However, these techniques are not viable for the cost-efficient manufacturing of specialized MEMS devices at low- and medium-scale volumes. Thus, applications that require custom-designed MEMS devices for markets with low- and medium-scale volumes of below 5000-10,000 components per year are extremely difficult to address efficiently. The 3D printing of MEMS devices could enable the efficient realization and production of MEMS devices at these low- and medium-scale volumes. However, current micro-3D printing technologies have limited capabilities for printing functional MEMS. Herein, we demonstrate a functional 3D-printed MEMS accelerometer using 3D printing by two-photon polymerization in combination with the deposition of a strain gauge transducer by metal evaporation. We characterized the responsivity, resonance frequency, and stability over time of the MEMS accelerometer. Our results demonstrate that the 3D printing of functional MEMS is a viable approach that could enable the efficient realization of a variety of custom-designed MEMS devices, addressing new application areas that are difficult or impossible to address using conventional MEMS manufacturing.

Keywords: Electrical and electronic engineering; Nanoscience and technology.

Fig. 1. 3D-printed accelerometer.

a SEM image of the 3D-printed accelerometer structure. b 3D schematic view of the cantilever cross-section before metal evaporation. c 3D schematic view of the cantilever cross-section after metal evaporation showing the shadow masking mechanism that enables the electrical isolation of the resistors. d SEM image of a lateral view of the top part of the device. e Close-up view of the T-shaped resistors on top of the cantilevers. The structures shown in the SEM images were coated with a thin sputtered Au–Pt layer to improve the SEM image quality.

Fig. 2. Mechanical characterization.

a Amplitude of oscillation of the MEMS accelerometer proof mass at different frequencies using a piezoshaker driving voltage of 7 V. The resonance frequency was measured at the maximum oscillation amplitude between 1.77 and 1.78 kHz for all devices instead of 1.58 kHz, as predicted by the model developed in COMSOL^®, using a Young´s modulus of 5.1 GPa as specified in the material datasheet. The measured resonance frequency corresponds to an actual Young´s modulus of 6.5 GPa. Fitting to a Lorentzian curve yields Q factors of 36 (accelerometer 1), 31.7 (accelerometer 2), and 35.7 (accelerometer 3). b Measured and simulated amplitude of oscillation of the proof mass at the different accelerations applied by the piezoshaker. The amplitude was measured at the resonance frequency, and the simulated force was multiplied by the average Q factor of 34 to match the behavior at resonance.

Fig. 3. Electrical characterization.

Relative resistance change (ΔR/R) measured at different frequencies and driving voltages of the piezoshaker on accelerometer 1 in (a), accelerometer 2 in (b), and accelerometer 3 in (c). The noticeable shift in resonance frequency is attributed to the increase in temperature in the polymer at large oscillation amplitudes. d Relative resistance changes of the strain gauge transducers as a function of the acceleration applied to the MEMS accelerometers computed off-resonance.

Fig. 4. Stability evaluation.

Measurements of relative resistance change and resonance frequency of the three devices over a period of 10 h using a constant piezoshaker driving voltage of 5 V_rms. a For each device, the resonance frequency remained stable within an interval of ±3.8 Hz from the average resonance frequency. b Relative resistance change ΔR/R, computed at the resonance frequency, shifting by approximately 20 ppm with respect to the initial value of 205 ppm for both accelerometer 2 and 3, while remaining stable at 193 ± 5 ppm for accelerometer 1.

Fig. 5. Measurement setup and piezoshaker calibration.

a Schematic of the setup used to measure the responsivity of the 3D-printed MEMS accelerometer. The lock-in amplifier was used to drive the piezo-shaker, extract the signal from the resistor through downmixing and demodulate the laser doppler vibrometer signal. b Acceleration applied by the piezoshaker was measured with the laser doppler vibrometer on top of the supporting pillars of the three devices under different driving voltages at a frequency of 1.775 kHz. The corresponding linear fits are shown. A variation of about ±10% between the different devices can be seen.