Microelectromechanical Systems Devices Research Articles

A Hybrid Mechanism and Data‐Driven Approach for Predicting Fatigue Life of MEMS Devices by Physics‐Informed Neural Networks

ABSTRACTOur research group has focused on the long‐term performance degradation of micro‐electro‐mechanical systems (MEMS) devices under mechanical, thermal, and electrical stresses. We previously used physics‐based models for performance prediction, but the diverse materials, structures, and environmental conditions of MEMS devices affect their long‐term stability. Traditional physics‐based models struggle to account for these factors, limiting practical application. To address this, we employ data‐driven methods to include more variables. This paper introduces a hybrid approach combining physical principles with data‐driven techniques, specifically physics‐informed neural networks (PINN), to model and analyze performance degradation in MEMS resonators. We compare this method with previous physics‐based and data‐driven approaches (support vector machine and random forest) under identical conditions. The hybrid method not only provides accurate predictions but also considers more operating conditions, enhancing practical application.

Impact of Deposition Temperature on Structural and Electrical Properties of Sputtered AlN/ Si (111) for CMOS compatible MEMS

Aluminium nitride (AlN) thin films are grown by reactive sputtering on Silicon (Si) with (111) orientation at low temperature (< 450 °C) for piezoelectric-based micro-electromechanical systems (MEMS). The grown AlN thin films were polycrystalline wurtzite hexagonal structure with a high texture coefficient for the (002) plane of orientation. The leakage current density by the current-voltage (I-V) measurement was found to be as low as 1.6 ×10-6Acm−2 in the grown films. The trapped charges are crucial in controlling these electrical characteristics, and low interface trap density (6.72 ×1010cm−2 eV) was observed for the sputtered AlN grown at a substrate temperature of 300 °C. This study on the structural and electrical properties of AlN/Si (111) at low substrate temperature is compatible with the complementary metal oxide semiconductor (CMOS) process for constructing piezoelectric-based MEMS devices for various applications.

A Polyimide Composite-Based Electromagnetic Cantilever Structure for Smart Grid Current Sensing.

Polyimides (PIs) have been extensively used in thin film and micro-electromechanical system (MEMS) processes based on their excellent thermal and mechanical stability and high glass transition temperature. This research explores the development of a novel multilayer and multifunctional polymer composite electro-piezomagnetic device that can function as an energy harvester or sensor for current-carrying wires or magnetic field sensing. The devices consist of four layers of composite materials with a polyimide matrix. The composites have various nanoparticles to alter the functionality of each layer. Nanoparticles of Ag were used to increase the electrical conductivity of polyimide and act as electrodes; lead zirconate titanate was used to make the piezoelectric composite layer; and either neodymium iron boron (NdFeB) or Terfenol-D was used to make the magnetic and magnetostrictive composite layer, which was used as the proof mass. A novel all-polymer multifunctional polyimide composite cantilever was developed to operate at low frequencies. This paper compares the performance of the different magnetic masses, shapes, and concentrations, as well as the development of an all-magnetostrictive device to detect voltage or current changes when coupled to the magnetic field from a current-carrying wire. The PI/PZT cantilever with the PI/NdFeB proof mass demonstrated higher voltage output compared to the PI/Terfenol-D proof mass device. However, the magnetostrictive composite film could be operated without a piezoelectric film based on the Villari effect, which consisted of a single PI-Terfenol-D film. The paper illustrates the potential to develop an all-polymer composite MEMS device capable of acting as a magnetic field or current sensor.

Neural networks based surrogate modeling for efficient uncertainty quantification and calibration of MEMS accelerometers

This paper addresses the computational challenges inherent in the stochastic characterization and uncertainty quantification of Micro-Electro-Mechanical Systems (MEMS) capacitive accelerometers. Traditional methods, such as Markov Chain Monte Carlo (MCMC) algorithms, are often constrained by the computational intensity required for high-fidelity (e.g., finite element) simulations. To overcome these limitations, we propose to use supervised learning-based surrogate models, specifically artificial neural networks, to effectively approximate the response of MEMS capacitive accelerometers. Our approach involves training the surrogate models with data derived from initial high-fidelity finite element analyses (FEA), providing rich datasets to be generated in an offline phase. The surrogate models replicate the FEA accuracy in predicting the behavior of the accelerometer under a wide range of fabrication parameters, thereby reducing the online computational cost without compromising accuracy. This enables extensive and efficient stochastic analyses of complex MEMS devices, offering a flexible framework for their characterization. A key application of our framework is demonstrated in estimating the sensitivity of an accelerometer, accounting for unknown mechanical offsets, over-etching, and thickness variations. We employ an MCMC approach to estimate the posterior distribution of the device’s unknown fabrication parameters, informed by its response to transient voltage signals. The integration of surrogate models for mapping fabrication parameters to device responses, and subsequently to sensitivity measures, greatly enhances both backward and forward uncertainty quantification, yielding accurate results while significantly improving the efficiency and effectiveness of the characterization process. This process allows for the reconstruction of device sensitivity using only voltage signals, without the need for direct mechanical acceleration stimuli.

Modeling Electronic Devices with a Casimir Cavity

The Casimir effect has been exploited in various MEMS (micro-electro-mechanical system) devices, especially to make sensitive force sensors and accelerometers. It has also been used to provide forces for a variety of purposes, for example, for the assembly of considerably small parts. Repulsive forces and torques have been produced using various configurations of media and materials. Just a few electronic devices have been explored that utilize the electrical properties of the Casimir effect. Recently, experimental results were presented that described the operation of an electronic device that employed a Casimir cavity attached to a standard MIM (metal–insulator–metal) structure. The DC (direct current) conductance of the novel MIM device was enhanced by the attached cavity and found to be directly proportional to the capacitance of the attached cavity. The phenomenological model proposed assumed that the cavity reduced the vacuum fluctuations, which resulted in a reduced injection of carriers. The analysis presented here indicates that the optical cavity actually enhances vacuum fluctuations, which would predict a current in the opposite direction from that observed. Further, the vacuum fluctuations near the electrode are shown to be approximately independent of the size of the optical cavity, in disagreement with the experimental data which show a dependence on the size. Thus, the proposed mechanism of operation does not appear correct. A more detailed theoretical analysis of these devices is needed, in particular, one that uses real material parameters and computes the vacuum fluctuations for the entire device. Such an analysis would reveal how these devices operate and might suggest design principles for a new genre of electronic devices that make use of vacuum fluctuations.

True Random Number Generator Based on Chaotic Oscillation of a Tunable Double-Well MEMS Resonator.

Chaotic systems have aroused interest across various scientific disciplines such as physics, biology, chemistry, and meteorology. The deterministic but unpredictable nature of a chaotic system is an ideal feature for random number generation. Microelectromechanical systems (MEMS) are a promising technology that effectively harnesses chaos, offering advantages such as a compact footprint, scalability, and low power consumption. This paper presents a true random number generator (TRNG) based on a double-well MEMS resonator integrated with an actuator and position sensor. The potential energy landscape of the proposed MEMS resonator is actively tunable with a direct current voltage. Experimental demonstrations of tunable bistability and chaotic resonance are reported in this paper. A chaotic time sequence is generated through piezoresistive sensing of the position of the MEMS resonator once it is driven into the chaotic regime. Subsequently, the randomness of the bit sequence, achieved by applying the exclusive or function to a digital chaotic sequence and its delayed differential is confirmed to meet the National Institute of Standards and Technology specifications. Moreover, the throughput and energy efficiency of the proposed MEMS-based TRNG can be adjusted from 50 kb s-1 and 0.44 pJ per bit at a low energy barrier to 167 kb s-1 and 6.74 pJ per bit at a high energy barrier by changing the MEMS device's potential well. The tunability of the proposed double-well MEMS resonator not only offers continuous adjustments in the energy efficiency of TNRG but also unveils vast and diverse research opportunities in analog computing, encryption, and secure communications.

Comparative Study of Piezoelectric MEMS for Enhanced Biosensors

Piezoelectric MEMS (Micro-Electro-Mechanical Systems) refers to devices that combine the principles of piezoelectricity with microfabrication techniques to create small-scale mechanical systems with integrated electronic components. Piezoelectric Microelectromechanical Systems (MEMS) for enhanced biosensors represent a sophisticated integration of two key technologies: piezoelectric materials and MEMS fabrication techniques. This combination results in highly sensitive, miniaturized devices that can detect and analyze biological molecules with precision. Piezoelectric MEMS devices offer unique advantages including high sensitivity, low power consumption, and miniaturization, making them ideal candidates for cutting-edge biosensor development. This article provides a comparative study of various Piezoelectric MEMS-based biosensors, highlighting their strengths and limitations.

Research on Adaptive Closed-Loop Control of Microelectromechanical System Gyroscopes under Temperature Disturbance.

Microelectromechanical System (MEMS) gyroscopes are inertial sensors used to measure angular velocity. Due to their small size and low power consumption, MEMS devices are widely employed in consumer electronics and the automotive industry. MEMS gyroscopes typically use closed-loop control systems, which often use PID controllers with fixed parameters. These classical PID controllers require a trade-off between overshoot and rise time. However, temperature variations can cause changes in the gyroscope's parameters, which in turn affect the PID controller's performance. To address this issue, this paper proposes an adaptive PID controller that adjusts its parameters in response to temperature-induced changes in the gyroscope's characteristics, based on the error value. A closed-loop control system using the adaptive PID was developed in Simulink and compared with a classical PID controller. The results demonstrate that the adaptive PID controller effectively tracked the changes in the gyroscope's parameters, reducing overshoot by 96% while maintaining a similar rise time. During gyroscope startup, the adaptive PID controller achieves faster stabilization with a 0.036 s settling time, outperforming the 0.06 s of the conventional PID controller.

Electrohydrodynamic-Jet-Printed SnO2-TiO2-Composite-Based Microelectromechanical Systems Sensor with Enhanced Ethanol Detection.

Ethanol sensors have found extensive applications across various industries, including the chemical, environmental, transportation, and healthcare sectors. With increasing demands for enhanced performance and reduced energy consumption, there is a growing need for developing new ethanol sensors. Micro-electromechanical system (MEMS) devices offer promising prospects in gas sensor applications due to their compact size, low power requirements, and seamless integration capabilities. In this study, SnO2-TiO2 nanocomposites with varying molar ratios of SnO2 and TiO2 were synthesized via ball milling and then printed on MEMS chips for ethanol sensing using electrohydrodynamic (EHD) printing. The study indicates that the two metal oxides dispersed evenly, resulting in a well-formed gas-sensitive film. The SnO2-TiO2 composite exhibits the best performance at a molar ratio of 1:1, with a response value of 25.6 to 50 ppm ethanol at 288 °C. This value is 7.2 times and 1.8 times higher than that of single SnO2 and TiO2 gas sensors, respectively. The enhanced gas sensitivity can be attributed to the increased surface reactive oxygen species and optimized material resistance resulting from the chemical and electronic effects of the composite.

Supercapacitors on hairs with quantum capacitance-dominant extraordinary capacitance

Graphene-based planar microsupercapacitors are promising energy storage devices for microelectromechanical systems. However, the capacitance and energy density of such graphene-based microsupercapacitors cannot achieve further promotion because of the lack of strategies that permit tailoring both chemical and geometric structures of graphene at the nanoscale. In this study, supercapacitors on hairs with quantum capacitance-dominant extraordinary capacitance are found. Simulations reveal that the record high capacitance and energy density are attributed to the enhanced quantum capacitance effect. The graphene electrodes are prepared by femtosecond laser direct writing-assisted reduction and N-doping of graphene oxides on the human hair surface. Due to the distinct quantum capacitance dominance, the microsupercapacitors exhibit extremely high capacitance (∼3000 mF/cm2) and energy density (∼417 μWh/cm2). We hope that this work will promote the development of microsupercapacitors with high capacitance and energy density.

A novel evaluation method of the aging performance of MEMS flow sensor

The development of Micro Electro Mechanical Systems (MEMS) flow sensor towards high level market applications generates various challenges, in particular also on the reliable functionality. With the advancement of reliability engineering technology, aging phenomenon of MEMS devices has been widely concerned. As a result, an aging evaluation method is essential. The accelerated aging testing (AAT) is the most widely used in traditional aging methods. However, its performance is limited by highly economic and time cost. In this paper, a novel aging effect model is proposed, in which a comprehensive approach that integrates AAT, lifetime distribution modeling, and either Finite Element Modeling Simulation (FEMS) or fatigue simulation (FS) as fundamental is explored. However, the difference from the conventional approach was that the AAT results is imported in FS, to confirm fatigue analysis, while FS predictions are instrumental in analyzing degradation or fatigue phenomena and estimation lifetime. In this way, aging performance is illustrated detailed with the lower aging cost and high efficiency. Meanwhile, the results of proposed FS are verified by a thermal cycle (TC) AAT. Specifically, the resistor degradation mechanism, the characteristic parameter degradation is calculated. Moreover, the lifetime evaluation was acquired by the Arrhenius model. Finally, the proposed aging performance evaluation method can be applied to both discrete devices and modules. Compared with the traditional aging method the high aging cost can be eliminated, and the proposed aging evaluation strategy can be used in various temperature conditions.

Summary of MEMS Capacitive Accelerometer

Micro-Electro-Mechanical System (MEMS) accelerometer (MEMS accelerometer), as an important MEMS device, has been widely used in the field of high-precision electronic technology and consumer electronic products. At present, the technology of MEMS accelerometer has been mature, and the sensor based on capacitance principle, piezoelectric effect, piezoresistive effect and other principles has been widely used in many aspects of national life and production and sustainable development. Combined with and according to all kinds of references, this paper briefly describes the shape composition and sensing principle of the MEMS capacitive accelerometer, enumerates the design and error optimization research of each research group for the sensor, and explains it with the relevant research data.

4H silicon carbide bulk acoustic wave gyroscope with ultra-high Q-factor for on-chip inertial navigation

Inertial navigation on a chip has long been constrained by the noise and stability issues of micromechanical Coriolis gyroscopes, as silicon, the dominant material for microelectromechanical system devices, has reached the physical limits of its material properties. To address these challenges, this study explores silicon carbide, specifically its monocrystalline 4H polytype, as a substrate to improve gyroscope performance due to its low phonon Akhiezer dissipation and its isotropic hexagonal crystal lattice. We report on low-noise electrostatic acoustic resonant gyroscopes with mechanical quality factors exceeding several millions, fabricated on bonded 4H silicon carbide-on-insulator wafers. These gyroscopes operate using megahertz frequency bulk acoustic wave modes for large open-loop bandwidth and are tuned electrostatically using capacitive transducers created by wafer-level deep reactive ion etching. Experimental results show these gyroscopes achieve superior performance under various conditions and demonstrate higher quality factors at increased temperatures, enabling enhanced performance in an ovenized or high-temperature stabilized configuration.

An ultra-deep TSV technique enabled by the dual catalysis-based electroless plating of combined barrier and seed layers

Silicon interposers embedded with ultra-deep through-silicon vias (TSVs) are in great demand for the heterogeneous integration and packaging of opto-electronic chiplets and microelectromechanical systems (MEMS) devices. Considering the cost-effective and reliable manufacturing of ultra-deep TSVs, the formation of continuous barrier and seed layers remains a crucial challenge to solve. Herein, we present a novel dual catalysis-based electroless plating (ELP) technique by tailoring polyimide (PI) liner surfaces to fabricate dense combined Ni barrier/seed layers in ultra-deep TSVs. In additional to the conventional acid catalysis procedure, a prior catalytic step in an alkaline environment is proposed to hydrolyze the PI surface into a polyamide acid (PAA) interfacial layer, resulting in additional catalysts and the formation of a dense Ni layer that can function as both a barrier layer and a seed layer, particularly at the bottom of the deep TSV. TSVs with depths larger than 500 μm and no voids are successfully fabricated in this study. The fabrication process involves low costs and temperatures. For a fabricated 530-μm-deep TSV with a diameter of 70 μm, the measured depletion capacitance and leakage current are approximately 1.3 pF and 1.7 pA at 20 V, respectively, indicating good electrical properties. The proposed fabrication strategy can provide a cost-effective and feasible solution to the challenge of manufacturing ultra-deep TSVs for modern 3D heterogeneous integration and packaging applications.

Low-thermal-budget electrically active thick polysilicon for CMOS-First MEMS-last integration

Low-thermal-budget, electrically active, and thick polysilicon films are necessary for building a microelectromechanical system (MEMS) on top of a complementary metal oxide semiconductor (CMOS). However, the formation of these polysilicon films is a challenge in this field. Herein, for the first time, the development of in situ phosphorus-doped silicon films deposited under ultrahigh-vacuum conditions (~10−9 Torr) using electron-beam evaporation (UHVEE) is reported. This process results in electrically active, fully crystallized, low-stress, smooth, and thick polysilicon films with low thermal budgets. The crystallographic, mechanical, and electrical properties of phosphorus-doped UHVEE polysilicon films are studied. These films are compared with intrinsic and boron-doped UHVEE silicon films. Raman spectroscopy, X-ray diffraction (XRD), transmission electron microscopy (TEM) and atomic force microscopy (AFM) are used for crystallographic and surface morphological investigations. Wafer curvature, cantilever deflection profile and resonance frequency measurements are employed to study the mechanical properties of the specimens. Moreover, resistivity measurements are conducted to investigate the electrical properties of the films. Highly vertical, high-aspect-ratio micromachining of UHVEE polysilicon has been developed. A comb-drive structure is designed, simulated, fabricated, and characterized as an actuator and inertial sensor comprising 20-μm-thick in situ phosphorus-doped UHVEE films at a temperature less than 500 °C. The results demonstrate for the first time that UHVEE polysilicon uniquely allows the realization of mechanically and electrically functional MEMS devices with low thermal budgets.

Development of Environmental Monitoring Sensor for Steel Corrosion in Concrete based on Micro-Electro-Mechanical Systems (MEMS)

Many bridges located near marine environments usually have high risk to lose their structural performance due to corrosion of reinforcing bars with chloride attack. While a huge amount of budget to maintain may be consumed for the structural maintenance methodologies, an efficient monitoring system for Reinforced Concrete (RC) structures in the corrosive environment is strongly demanded. Thus, structural health monitoring in real situations is significant for bridge asset management. The alternative way for preventing severe damage propagation is proactively monitoring the environmental condition containing chloride ion and its effect on the durability of concrete structures, for which numerous methods have been conventionally suggested such as dry gauze method to capture chlorides in air. In this study, Micro Energy Harvester (MEH) based on Micro-Electro-Mechanical Systems (MEMS) is technically applied to develop a sensor to identify the corrosion environment with chlorides in the exposure condition. MEMS is a rapidly evolving device, which enables to monitor the changes in natural frequency and amplitude at nodal locations of the targeting structure due to vibration-based electrostatic power generation induced by the corrosive environment. The purpose of this research is to develop a system for detecting the high corrosion-prone areas and finding the suitable metal sheet material for various chloride ion conditions, using the metal sheet as a cantilever structure for the MEMS device. 4 types of metal, aluminium, copper, iron and zinc are considered for the parametric experiments. Several scenarios are considered including chloride ion concentration, temperatures, and wet and dry conditions to compare the dominant frequency changes of each metal sheet and determine their sensitivity to corrosion environment. From the preliminary results, iron and zinc have a significant drop in frequency with the elapsed exposure period, which indicates that they are highly sensitive to the external chloride ion environment. These results offer an innovative and proactive approach to structural health monitoring for further developed applications. It enables prompt maintenance and early detection of potential issues for corrosion of steel bars in concrete, ensuring the longevity and safety of RC bridge structures.

A new fabrication method for enhancing the yield of linear micromirror arrays assisted by temporary anchors

As one of the most common spatial light modulators, linear micromirror arrays (MMAs) based on microelectromechanical system (MEMS) processes are currently utilized in many fields. However, two crucial challenges exist in the fabrication of such devices: the adhesion of silicon microstructures caused by anodic bonding and the destruction of the suspended silicon film due to residual stress. To solve these issues, an innovative processing method assisted by temporary anchors is presented. This approach effectively reduces the span of silicon microstructures and improves the Euler buckling limit of the silicon film. Importantly, these temporary anchors are strategically placed within the primary etching areas, enabling easy removal without additional processing steps. As a result, we successfully achieved wafer-level, high-yield manufacturing of linear MMAs with a filling factor as high as 95.1%. Demonstrating superior capabilities to those of original MMAs, our enhanced version boasts a total of 60 linear micromirror elements, each featuring a length-to-width ratio of 52.6, and the entire optical aperture measures 5 mm × 6 mm. The linear MMA exhibits an optical deflection angle of 20.4° at 110 Vdc while maintaining exceptional deflection flatness and uniformity. This study offers a viable approach for the design and fabrication of thin-film MEMS devices with high yields, and the proposed MMA is promising as a replacement for digital micromirror devices (DMDs, by TI Corp.) in fields such as spectral imaging and optical communication.

Optimization Design of CMUT Sensors with Broadband and High Sensitivity Characteristics Based on the Genetic Algorithm.

In this study, we propose a method for optimizing the design of CMUT sensors using genetic algorithms. Existing CMUT sensors face frequency response and sensitivity limitations, necessitating optimization to enhance their sensing performance. Traditional optimization methods are often intricate and time-consuming and may fail to yield the optimal solution. Genetic algorithms, which simulate the biological evolution process, offer advantages in global optimization and efficiency, making them widely utilized in the optimization design of Microelectromechanical Systems (MEMS) devices. Based on the theoretical framework and finite element model of CMUT sensors, we propose a CMUT array element optimization design method based on genetic algorithms. The optimization and validation results demonstrate that we have successfully designed a broadband CMUT array element consisting of four microelements with a 1-2 MHz frequency range. Compared with a randomly arranged array element, the optimized array shows a 63.9% increase in bandwidth and a 7.5% increase in average sensitivity within the passband. Moreover, the sensitivity variance within the passband is reduced by 50.2%. Our proposed method effectively optimizes the design of high sensitivity CMUT sensors with the desired bandwidth, thereby offering significant reference value for the optimization design of CMUT sensors.

Dynamic imaging of micro-vibrations with an ultra-wide bandwidth and a femtometer noise using switchable pulsed laser interferometry

Imaging the complex dynamics of micro-vibrations plays a fundamental role in the investigation of microelectromechanical systems (MEMS). However, it remains a challenge for achieving both a wide bandwidth and a low noise due to the high photodetector noise and electromagnetic interference at GHz frequencies. Here, we propose a pulsed laser interferometry system with an adaptable switch to image GHz vibrations based on stroboscopic mixing, while measuring lower-frequency vibrations based on the homodyne scheme. The noise power spectral density is shown in both regions from DC to 10 GHz with an average noise down to 30.8 fm/√Hz at GHz frequencies, which holds the highest resolution to the best of our knowledge. Vibrational amplitude and phase mappings of a kHz comb-drive resonator, a GHz piezoelectric transducer, and a GHz film bulk acoustic resonator are presented with animated visualizations and k-space analysis, paving a new paradigm for the first time to image and analyze various MEMS devices of a bandwidth spanning 10 orders of magnitude.

Electrical actuation of single-crystal diamond MEMS resonators at high temperatures

Achieving efficient low-voltage actuation of microelectromechanical system (MEMS) resonators in high-temperature environments poses a difficult topic due to the thermal interference and the risk of high-temperature failure. In this work, the single-crystal diamond (SCD) resonators fabricated through the ion implantation-assisted lift-off (IAL) technique exhibit a SCD-on-SCD cantilever structure. We propose an electrical actuation system based on the electrostatic effect specifically designed for SCD MEMS resonators with a low radio-frequency amplitude of ∼ 100 mV. The SCD resonators demonstrate stable and efficient actuation across a wide temperature range, from room temperature to 500 °C. Importantly, the actuation voltage exhibits little impact on the resonance frequency and the Q factor of the resonator. The SCD resonator showcases exceptional thermal stability in resonance frequency, with a low temperature coefficient of frequency (TCF) below −12 ppm/°C up to 500 °C. The developed actuation scheme holds tremendous potential as a robust platform for realizing SCD MEMS devices, particularly in applications requiring high integration at high temperatures.