Significant tuning of dispersive mode coupling in doubly clamped MEMS beam resonators using thermally induced buckling effect

Chao Li, Qian Liu, Kohei Uchida, Hua Li, Kazuhiko Hirakawa, Ya ZhangarXiv 2024 · DOI · arXiv

TL;DR

This paper demonstrates a novel method to precisely tune the dispersive mode coupling strength in MEMS beam resonators by leveraging thermally-induced buckling. By using an integrated heater to break the beam's symmetry, the researchers achieved a wide range of coupling coefficients—including a "zero-coupling" state—and established a direct theoretical link between mode coupling and mechanical nonlinearity.

What Problem Does This Paper Address?

In micro/nanoelectromechanical systems (M/NEMS), different vibration modes often interact. One common interaction is dispersive mode coupling, where the vibration amplitude of one mode (the "pump") shifts the resonance frequency of another (the "probe"). This is useful for advanced signal processing and high-sensitivity sensing.

However, two major hurdles remain:

  1. Weak and Fixed Strength: Dispersive coupling is typically weak, and its strength is usually "baked in" by the device's geometry during fabrication.
  2. Lack of Theory: While experimental observations exist, there hasn't been a clear quantitative model to predict how to tune this coupling or how it relates to the inherent nonlinearities of the resonator.

This paper tackles these issues by using heat to induce buckling, providing a "knob" to turn the coupling strength up, down, or even off.

Methods

The researchers fabricated a doubly-clamped MEMS beam made of Gallium Arsenide (GaAs). The device features:

  • Piezoelectric Actuation: Using the piezoelectric effect of GaAs to drive vibrations.
  • Integrated Heater: A thin NiCr layer deposited on the beam to generate thermal strain.
  • Optical Detection: A laser Doppler vibrometer (LDV) and a lock-in amplifier with a phase-locked loop (PLL) to measure frequency shifts in real-time.

To measure the coupling, they drove the 1st bending mode (probe) into self-sustained oscillation and then excited the 2nd bending mode (pump). By varying the DC voltage to the heater, they induced thermal expansion. When this expansion exceeds a critical point, the beam "buckles," changing its shape and internal tension, which in turn modifies the coupling coefficient ($\lambda_{1,2}$).

Datasets

As an experimental physics and engineering paper, it does not use standard machine learning datasets. Instead, the "data" consists of original experimental measurements from the GaAs MEMS device, including:

  • Frequency response spectra under varying heating powers ($0$ to $2.5\text{ mW}$).
  • Amplitude-frequency curves for the 1st and 2nd bending modes.
  • Calculated coupling coefficients derived from the observed frequency shifts.

Key Insights and Ideas

  • Symmetry Breaking as a Tool: Buckling is often seen as a failure mode in engineering. Here, it is used constructively. Buckling breaks the beam's symmetry, which fundamentally alters the vibration-induced tension. This change is what allows the coupling coefficient to transition from positive to negative.
  • The "Zero-Coupling" Sweet Spot: The researchers identified a specific state where the coupling coefficient reaches zero. This is highly valuable for multi-mode sensors where you want different modes to operate independently without "cross-talk."
  • The Nonlinearity Link: The authors derived a simple analytical relationship showing that the coupling coefficient is essentially an interaction between the Duffing nonlinearities of the two modes. This means engineers can predict coupling behavior simply by measuring the individual nonlinearities of the modes.
  • Internal Resonance Interference: They discovered a "sudden drop" in coupling when the frequencies of the two modes hit a 1:3 ratio. This suggests that internal resonance (a different type of coupling) can disrupt dispersive coupling, a nuance that future designers must account for.

Key Results

  • Wide Tunability: The coupling coefficient ($\lambda_{1,2}$) was successfully tuned from a positive value ($1.53 \times 10^{12}\text{ m}^{-2}$ at $0\text{ mW}$) down to zero, and then into negative values as heating increased.
  • Critical Buckling: The 1st bending mode reached its critical buckling point at a heating power of $1.33\text{ mW}$, resulting in a sharp reduction in the coupling coefficient.
  • Theoretical Validation: The developed Euler-Bernoulli model showed "nice agreement" with experimental data, confirming that the tunability stems from the breakdown of the beam's symmetric shape.
  • Ultra-Strong Potential: Theoretical analysis suggests that by reducing the initial center deflection of the beam ($x_0/h < 0.01$), the coupling strength could be boosted by over 100 times compared to the experimental results.

Limitations and Future Work

  • Model Scope: The current theoretical model focuses purely on dispersive coupling. It neglects the coherent energy transfer that happens during internal resonance (at integer frequency ratios). This explains why the model could not account for the "sudden drop" observed at the 1:3 frequency ratio.
  • Active vs. Passive Tuning: While thermal tuning is effective, it requires constant power. The authors suggest that lattice mismatch strain (built-in during material growth) could be a passive alternative for "pre-tuning" devices without an external heater.
  • Future Investigation: The authors call for more research into the "intrinsic relationship" between internal resonance and dispersive mode coupling to create a more unified theory of modal interactions.

Summary updated

Abstract

Dispersive mode coupling is a promising mechanism for the development of advanced micro/nanoelectromechanical devices. However, strong coupling strength remains a key challenge limiting the practical applications of dispersive mode coupling effect. Here, we experimentally demonstrate the significant tuning of the mode coupling coefficient of two flexural vibrational modes in a doubly-clamped MEMS beam resonator using thermally-induced buckling effect, which enables variable coupling strengths to be implemented for practical applications. Furthermore, a theoretical model is developed to describe the mode coupling coefficient, showing that the tunability is owing to the breakdown of the symmetric shape of the MEMS beam caused by buckling. Moreover, the theoretical model defines a simple relation between the coupling coefficient and the nonlinearity of the two coupled modes. These results provide valuable insight into physical mechanisms underlying dispersive mode coupling effect, as well as pave the way for the development of advanced MEMS devices with application-specified mode coupling strength.
BibTeX
@misc{li2024significant,
  author = {Chao Li and Qian Liu and Kohei Uchida and Hua Li and Kazuhiko Hirakawa and Ya Zhang},
  title = {{Significant tuning of dispersive mode coupling in doubly clamped MEMS beam resonators using thermally induced buckling effect}},
  year = {2024},
  eprint = {2405.00506},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2405.00506},
}
Figure