This letter presents an approach for achieving high-dynamic-range bulk-mode piezoelectric MEMS resonators with suppressed Duffing nonlinearity via crystal orientation tuning. The demonstrated width-extensional (WE) mode resonator is designed and fabricated on a thin-film piezoelectric-on-silicon (TPoS) platform. By strategically tuning the crystal orientation of the bulky single-crystal silicon device body, Duffing nonlinearity is systematically mitigated, thereby enhancing the dynamic range of the resonator. Experimental characterization confirms that the resonator with a crystal orientation of 30° from the <100> direction exhibits the optimal dynamic range performance. When integrated into a closed-loop oscillation system, this optimized resonator achieves exceptional phase noise metrics of −124.25dBc/Hz at 1 kHz offset and −164.30dBc/Hz at 1MHz offset. Itprovides a novel solution for suppressing Duffing nonlinearity in piezoelectric MEMS resonators, addressing critical challenges in precision timing and frequency control applications.[2025-0110]

In this work, we present a newly proposed concept of engineering the anchor fins in a suspended Lamb wave resonator and using it as an important design parameter to optimize the resonator performance. The exploration was conducted on antisymmetric Lamb wave resonators based on 128° Y-cut lithium niobate (LN or LiNbO<inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${}_{\mathbf {3}}$</tex-math> </inline-formula>) and the deformation was adjusted by introducing different anchor fin designs that mount the suspended structure to the supporting substrate. The proposed design methodology was validated by the demonstration of first-order antisymmetric Lamb wave (A1-mode) resonators operating at 12GHz, and also third-order antisymmetric Lamb wave (A3-mode) resonators operating at 36GHz. It has been proved that special anchor fin design serves as an effective means to improve the power handling or linearity of Lamb wave resonators. Therefore, this work opens up a brand new design space for acoustic resonators based on a suspended thin-film structure, which makes it possible to realize further performance enhancement and optimization for Lamb wave resonators especially at centimeter wave (cmWave) and millimeter wave (mmWave) frequencies. [2025-0117]

The development of high-performance MEMS devices necessitates efficient design optimization and yield analysis in the presence of unavoidable process variations. The existing algorithms are not able to integrate the different requirements of optimization and uncertainty quantification (UQ), highlighting the necessity for a comprehensive algorithmic framework. To this end, a novel algorithm framework is proposed for the optimization and UQ of MEMS devices. The deep neural network (DNN) and polynomial chaos expansion (PCE) are respectively leveraged to build the global surrogate model for the nondominated sorting genetic algorithm II (NSGA-II) and the local surrogate model for the UQ analysis. The DNN is refined by the bi-fidelity data training scheme to reduce the training data demanded, and the active subspace dimension reduction (AS) is employed to fully utilize the information within the global surrogate model, thereby minimizing the computational resources needed for the local surrogate model construction. This algorithm is applied to a kind of MEMS thermal flow sensor. The proposed algorithmic simultaneously increases sensor sensitivity by 14.77% while reducing computational expenditure by 44.4% during the optimization phase. Besides, the yield analysis is also derived accurately, with computational cost reductions of 73.0%. These results indicate that this novel algorithm supplies an effective, accurate, and efficient technique for the optimization and yield analysis for MEMS devices. [2025-0231]

This paper presents a novel two-dimensional (2D) model based on partial differential equations (PDEs) to characterize the input-output response of the dual-axis micro thermal convective tilt (<inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\mu $</tex-math> </inline-formula>TCT) sensors. The proposed model enables highly efficient performance analysis of <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\mu $</tex-math> </inline-formula>TCT sensors with a computational time of 3 minutes. Based on the theoretical framework, the key parameters of the <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\mu $</tex-math> </inline-formula>TCT sensor, including the heater-detector distance, sensor length, film thickness, and cavity depth, are optimized to achieve high sensitivity and low power consumption. Furthermore, the model provides a reliable method for quantifying cross-axis crosstalk relative to the sensitivity of the sensitive axis. Accordingly, a <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\mu $</tex-math> </inline-formula>TCT sensor has been designed and fabricated, demonstrating bidirectional inclination detection over ±180° range with a power consumption of 3.29mW. Measured sensitivities along x-axis and y-axis were measured as 37.53 mV/° and 43.03 mV/°, respectively. Good agreement between experimental results and theoretical predictions validates the model’s accuracy and feasibility. With a normalized sensitivity of 13.07 mV/°/mW, the <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\mu $</tex-math> </inline-formula>TCT sensor surpasses the previous micro tilt sensors by >3 times, while maintaining the cross-axis crosstalk error (<inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\delta $</tex-math> </inline-formula>) below 5%. This study establishes the developed 2D model as an indispensable tool for the design and optimization of high-performance dual-axis <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\mu $</tex-math> </inline-formula>TCT sensors. The significant performance improvements highlight the model’s critical role in advancing micro tilt sensor technology for demanding applications such as high-precision attitude detection in industrial IoT systems. [2025-0218]

This work reports novel solidly mounted acoustic resonators based on the high-order guided shear vertical surface acoustic wave (SV-SAW) mode. The design utilizes a high-velocity supporting substrate to suppress the leakage of the main mode, while incorporating a bottom electrode to establish a vertical electric field for acoustic wave excitation. Using the finite element method (FEM), structural parameters were optimized to achieve a resonator with large electromechanical coupling coefficient (<inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${k}_{t}^{2}$</tex-math> </inline-formula>), low loss, and weak spurious modes simultaneously. The layout and dimensions of the bottom electrode are optimized to minimize bulk acoustic wave leakage. Subsequently, two specifications of resonators based on X-cut LiNbO<sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</sub>/patterned metal/SiO<sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub>/SiC substrate were fabricated via a secondary transfer process of piezoelectric single-crystal film. The fabricated resonators aiming at the C-band exhibit <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${k}_{t}^{2}$</tex-math> </inline-formula> ranging from 6% to 16%. Resonators operating at higher frequencies near 9 GHz were also fabricated, achieving <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$k_{t}^{2}$</tex-math> </inline-formula> value up to 16%. With further optimization, this novel high-order guided SV-SAW resonator will offer a promising filtering solution for 5G New Radio (NR) and future 6G.[2025-0194]

Low-frequency characteristics serve as primary metrics for evaluating the acoustic performance of modern microspeakers. However, conventional microspeakers exhibit flat low-frequency in-ear responses due to their inherent diaphragm displacement–sound pressure relationship, which deviates from ideal human auditory characteristics and fails to meet active noise cancellation requirements. To address this limitation, this study proposes a piezoelectric MEMS speaker that is based on ultrasonic modulation-demodulation principles. This is achieved by dynamic acoustic impedance, where the ultrasonic acoustic source and valve actuators work synchronously. The speaker achieves superior low-frequency performance, featuring a low-frequency sound pressure level of 134 dB at 20 Hz, following the IEC 711 standard. The total harmonic distortion remains below 2% in the 20 Hz to 1 kHz range. The speaker measures only 4 mm <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\times 9$</tex-math> </inline-formula> mm in size, and this size is possible due to its source–valve decoupling architecture. Combined with high performance and miniaturization, this technological breakthrough overcomes the performance limitations of conventional speakers and establishes a new paradigm for microspeakers.[2025-0163]

We present a high-density silicon neural microelectrode array featuring a three-dimensional architecture that vertically separates metal interconnects from recording sites. This design decouples the signal acquisition and transmission layers, overcoming the spatial constraints of conventional coplanar layouts. With interconnect linewidth and spacing scaled to <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$1~\mu $</tex-math> </inline-formula>m, the array integrates 256 recording channels within a 0.8mm<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> footprint, achieving a 41.8 % increase in channel density compared with coplanar counterparts. Circuit modeling shows that crosstalk between adjacent interconnects remains below <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$10~\mu $</tex-math> </inline-formula>V, while thermal simulations confirm that the operational temperature rise stays within safe limits, ensuring stability for chronic in vivo recordings. We optimized the high-density bonding interface by eliminating the pre-bonding ball-stud process, which simplifies the flip-chip bonding workflow and reduces both manufacturing complexity and cost while improving device consistency. This interface design also enhances system scalability, enabling higher-throughput integration. A 10-week chronic recording experiment evaluated key performance metrics including spike amplitude, firing rate, and signal-to-noise ratio (SNR). Results demonstrated that high-quality spontaneous neural signals were stably recorded throughout the 10-week period, validating the electrode’s reliability for long-term chronic recording. These findings confirm that the high-density silicon-based MEA enables high-resolution, long-term stable neural recording with significant potential for brain-computer interface and neural engineering applications.[2025-0134]

Complicated multi-failure mechanisms triggered by distinctive stresses in different harsh environments have become essential yet challenging in reliability research of microelectromechanical systems (MEMS). For this issue, accelerated degradation testing (ADT) is widely used as an efficient strategy to obtain the long-term reliability of MEMS devices in a short amount of time. Therefore, precise and robust monitoring of performance degradation in harsh environments is the foundation of ADTs. However, current electro-mechanical-coupling methods, especially on-chip piezoresistive sensors, exhibit sensitivity shifts and must be calibrated manually at different temperatures, making them costly and restricting the accuracy. Here, we propose a calibration-free, temperature-robust split accelerated degradation testing platform (S-ADTP), which can accurately evaluate the long-term reliability of 2D MEMS micromirrors without any thermal-induced sensitivity shifts. S-ADTP eliminates the error of the FOV detection caused by sensitivity shifts, contributing to higher and more temperature-robust accuracy than electro-mechanical-coupling methods. ADTs with single and multiple stresses are subsequently conducted. Experimental results reveal the distinct failure mechanisms associated with varying environmental conditions, indicating that the crack propagation is the primary failure mode in high-temperature environments, while the demagnetization of permanent magnets becomes dominant in temperature-humidity coupling environments. The work can provide a calibration-free and temperature-robust method without any thermal-induced sensitivity shifts in ADTs for 2D micromirrors, and distinguish the complicated multi-failure mechanisms triggered by distinctive environmental stresses.[2025-0062]