Search

We demonstrate two novel dual-mode ovenized MEMS resonators, each with an in-chip device layer micro-oven that utilizes less than 30mW for resonator temperature control over variations in external temperature from −40°C to +60°C. The device layer micro-oven enables correction for ambient temperature variations and achieves a 1-week frequency stability for the output mode over temperature near 1.5 ppb for the Lame-mode resonator. The devices were built in the Epi-Seal fabrication process and take advantage of the exceptional long-term stability of MEMS resonators built in that process. These results exceed all prior reports for frequency stability over time and temperature for MEMS resonators and have the potential to impact the development of miniature, low-power time references. [2019-0054]

This paper presents a novel spring disk resonator gyroscope (SRG). The SRG consists of coupled concentric serpentine spring-like rings to reduce mode stiffness while maintaining a device footprint the same as that of existing standard DRGs: 600um in diameter. The $Q$ is increased by a factor of 2.5, and the ring down-time of 934ms is 4 times longer than that of a standard DRG.

This paper presents a novel wafer-level thin-film encapsulation process that allows both narrow and wide trenches, which are necessary for traditional structures such as comb-drives. Fully functional devices with trench widths up to 50 $\mu \text{m}$ are fabricated by employing a vapor phase XeF2 isotropic silicon etch to create large cavities and an epitaxial deposition seal to encapsulate the devices in an ultra-clean, high vacuum environment with no native oxide and humidity. In this paper, we demonstrate the robustness of the proposed fabrication process, as well as the inherent benefits of the high-temperature epitaxial encapsulation process: high quality factor, extreme stability, exceptional aging, and fatigue performance. [2017-0098]

We present unexpected results showing that thinning the bottom substrate of a resonant MEMS tuning fork resonator decreases anchor damping. We also present findings that the tuning fork experiences more anchor damping when mounting the die with silver paste. This is important for wearable devices where minimizing the volumetric footprint of sensors is paramount.

Silicon Microelectromechanical Systems (MEMS) resonators have broad commercial applications for timing and inertial sensing. However, the performance of MEMS resonators is constrained by dissipation mechanisms, some of which are easily detected and well-understood, but some of which have never been directly observed. In this work, we present measurements of the quality factor, Q, for a family of single crystal silicon Lamé-mode resonators as a function of temperature, from 80–300 K. By comparing these Q measurements on resonators with variations in design, dimensions, and anchors, we have been able to show that gas damping, thermoelastic dissipation, and anchor damping are not significant dissipation mechanisms for these resonators. The measured f · Q product for these devices approaches 2 × 1013, which is consistent with the expected range for Akhiezer damping, and the dependence of Q on temperature and geometry is consistent with expectations for Akhiezer damping. These results thus provide the first clear, direct detection of Akhiezer dissipation in a MEMS resonator, which is widely considered to be the ultimate limit to Q in silicon MEMS devices.

The aim of this study is to investigate the behavior of the thermal expansion coefficient (CTE) of silicon at different crystallographic orientations through studies of the quality factor, Q, over a wide range of temperatures. We investigate how the observed effects of CTE may deviate from the expected isotropic behavior due to irregularities of fabrication, or because of the influence of doping or other effects.

This work demonstrates, for the first time, a wafer-scale encapsulation process for hermetic sealing of MEMS resonant structures incorporating fully differential electrodes, in and out of the device plane, and large lateral transduction gaps. The polysilicon electrode and device layers were deposited using an epitaxial silicon reactor with Tetra-Ethyl-Ortho-Silicate (TEOS) as a spacer material. Large transduction gaps were fabricated using vapor-phase xenon difluoride (XeF 2 ) and hydrofluoric acid (vHF) etches, with silicon and silicon dioxide as sacrificial materials, respectively. Transduction gaps of 0.7μm up to 50μm were fabricated and the polysilicon devices were driven into a strongly nonlinear regime using the bottom, top, and in-plane capacitive electrodes.

In this study, a novel method of measuring the Quality Factor (Q) of a MEMS tuning fork resonator at low temperature enabled us to accurately pinpoint the energy dissipation resulting from anchor damping [1]. We used this new method to perform the first direct examination of the impact of design of the anchors and the die mounting structures on the strength of anchor damping. Our results show that the elastic energy can escape the resonator through the anchor(s), and still be retained within the die, and that energy stored in the die can return to the resonator. Thus, the measured Q of the resonator is increased, depending on mounting structures far from the resonating element. Moreover, our data shows that the number of anchors does not change the total energy dissipation through the anchors.

The aim of this study is to quantify the energy loss mechanisms for single anchored double-ended tuning forks (DETF) — particularly, gas damping, thermoelastic dissipation (TED) and anchor damping. We performed ring-down measurements of quality factor (Q) as a function of pressure to confirm the absence of air damping. We also measured Q over a wide temperature range including where the coefficient of thermal expansion (CTE) crosses zero, and consequently the effects from thermoelastic dissipation (TED) become negligible. This provides the first opportunity to directly quantify energy losses due to anchor damping and allows us to vary the mounting configuration to explore the impact of the mounting on the loss.

This paper presents a topology optimization approach for reducing thermo-elastic dissipation (TED) in MEMS resonators. This algorithm is applied to a clamped-clamped resonant beam to maximize the quality factor (Q). Optimal designs have a Q ten times higher than a solid beam and are 75% higher than previously optimized devices. Furthermore, new designs have intuitive topologies. Beams are fabricated in silicon wafers and experimental measurements of Q agree well with simulation.

We demonstrate how to identify regions of major thermo-elastic dissipation (TED) in MEMS resonators and reduce this energy loss by modifying the device geometry. To demonstrate this, various geometries of a disk resonating gyroscope (DRG) are used. Devices are fabricated and tested to show that the TED-limited quality factor (Q) can indeed be increased using geometric manipulation, as predicted by simulation.

There is significant interest in integration of multiple MEMS functionalities into a single compact device. Our group has developed a wafer-scale encapsulation process that provides an ultraclean, stable environment for operation of MEMS timing references, which has been commercialized by SiTime, Inc. In this paper, we discuss some of the issues associated with incorporation of inertial sensors into this encapsulation process, including design constraints, stiction, pressure, and other issues.

We present models and measurements of lightly-doped and heavily-doped dual ring/dual-bar and double-ended tuning fork (DETF) resonators. Both models and measurements indicate that the Q of these resonators is only slightly impacted by the doping level, despite being dominated by thermoelastic dissipation (TED), which has a strong dependence on doping-dependent material properties. We compare experimental measurements of Q over a range of temperatures with models for TED, anchor damping and squeeze film damping. We found that the Q of light and heavily-doped resonators can be accounted for by a combination of TED and anchor or squeeze film damping. These results indicate that it is possible to fully account for the damping mechanisms in MEMS resonators if temperature-dependent measurements of Q are compared with models of the important mechanisms including the temperature-dependent materials properties.