Microelectromechanical Devices for High-Order Passive Temperature Compensation and Their Design Methods
By using a highly doped silicon substrate, in-plane rotation, and geometric modification in the MEMS resonator, the frequency drift problem of the MEMS resonator under temperature changes was solved, second-order temperature compensation was achieved, and temperature stability was improved, approaching the performance of a quartz resonator.
Patent Information
- Application Number
- CN202580029100.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Priority Date
- 2024-04-26
- Filing Date
- 2025-04-25
- Publication Date
- 2026-11-17
AI Technical Summary
The frequency drift problem of existing MEMS resonators with temperature changes, especially the drift caused by the second-order temperature coefficient, makes them unable to replace quartz resonators in timing and frequency reference applications.
By applying highly doped silicon substrates, in-plane rotation, and geometric modification to MEMS resonators, and combining this with the finite element method, a second-order temperature-compensated MEMS resonator was designed. The dopant type and concentration, as well as geometric modification, are used to compensate for temperature-induced frequency drift.
It achieves temperature stability with frequency drift of less than 1ppm in the temperature range of -40℃ to 85℃, improves the temperature compensation effect of MEMS resonators, and approaches the performance of quartz resonators.
Smart Images

Figure CN123125166A_ABST
Abstract
Description
Cross-references to related applications
[0001] This patent application claims priority to U.S. Nonprovisional Application No. 18 / 647,923, filed April 26, 2024, which is a continuation-in-part of U.S. Application No. 17 / 973,896, filed October 26, 2022, and claims priority under 35 USC § 120, which in turn claims priority to U.S. Provisional Application No. 63 / 271,956, filed October 26, 2021. The contents of each of the aforementioned applications are incorporated herein by reference in their entirety. Background Technology
[0002] The rise of the Internet of Things (IoT) has spurred the development of numerous sensor-based devices used in wearables, smartphones, and remote sensing for industrial and consumer applications. Timing references are ubiquitous in these devices and help provide signals for tracking time, synchronizing events in digital integrated circuits (ICs), and processing signals. For such high-performance electronic applications, high-accuracy microelectromechanical system (MEMS) resonators are likely to be desirable. Summary of the Invention
[0003] The purpose of this disclosure is to realize an improved temperature-compensated MEMS device, such as a resonator. In particular, the purpose of this disclosure is to realize a resonator design that provides second-order temperature compensation. This disclosure also provides a method for designing second-order temperature-compensated micromechanical resonators for various purposes. Attached Figure Description
[0004] The features, aspects, and advantages of the technology disclosed in this invention can be better understood by referring to the detailed embodiments listed below, the appended claims, and the accompanying drawings. Those skilled in the art will understand that the features shown in these drawings are for illustrative purposes and variations may exist, including different and / or additional features and their arrangements.
[0005] Figure 1 This is a flowchart illustrating a general method for passively compensating both the first-order and second-order temperature coefficients of frequency.
[0006] Figure 2 A cross-sectional view of a passively compensated MEMS resonator device is shown.
[0007] Figure 3 This is a flowchart illustrating a method for using in-plane rotation to generate at least partially temperature-compensated MEMS resonator devices.
[0008] Figure 4A The two in-plane eigenmodes of the Lamé mode resonator are shown.
[0009] Figure 4B This shows the relationship between the crystal axis and silicon. <100> or <110> A graph showing the second-order temperature coefficient of the frequency of Lamé mode resonators with N-type or P-type dopant orientation versus dopant concentration.
[0010] Figure 5 This shows the first-order temperature coefficient of the frequency of a square resonator element operating in in-plane resonance mode, relative to silicon. <110> A diagram showing the in-plane rotation angle of the crystal axis.
[0011] Figure 6 This shows the first-order temperature coefficient of the frequency of a square resonator element operating in out-of-plane resonance mode, relative to silicon. <110> A diagram showing the in-plane rotation angle of the crystal axis.
[0012] Figure 7 This is a flowchart illustrating a method for using geometric modifications to produce at least a second-order temperature-compensated MEMS resonator devices.
[0013] Figure 8 The diagram shows two in-plane eigenmodes of both the Lamé mode resonator and the surface shear (FS) mode resonator.
[0014] Figure 9 This illustrates an exemplary geometry modification involving adding regions to resonator elements to compensate for the first-order temperature coefficient of the frequency.
[0015] Figure 10 An exemplary geometry modification involving subtracting a region from the resonator element to compensate for the first-order temperature coefficient of the frequency is shown.
[0016] Figure 11 Examples of Lamé resonance modes and modified Lamé resonance modes with altered geometry are shown.
[0017] Figure 12 Examples of FS resonance modes and modified FS resonance modes with altered geometry are shown.
[0018] Figure 13 Examples of wine glass resonance modes in a circular resonator element and modified wine glass resonance modes with altered geometry are shown.
[0019] Figure 14 Examples of wine glass resonance modes in an octagonal resonator element and modified wine glass resonance modes with altered geometry are shown.
[0020] Figure 15 Examples of both torsional resonance mode and saddle resonance mode, as well as modified torsional resonance mode and modified saddle resonance mode with altered geometry, are shown.
[0021] Figure 16 This illustrates the relationship between silicon and other materials. <110> An exemplary geometry modification of a Lamé-mode resonator element with its crystal axis rotated by 34°, minus the region to compensate for the first-order temperature coefficient of the frequency.
[0022] Figure 17 Showing relative to silicon <110> Examples of Lamé mode resonator elements with a crystal axis rotated by 34° and resonator elements with modified geometry.
[0023] Figure 18 A MEMS resonator device with an anchor decoupling region is shown.
[0024] Figure 19 A cross-sectional view of a MEMS resonator device with an anchor decoupler region is shown.
[0025] Figure 20 A top view showing various anchor decoupler area configurations.
[0026] Figure 21 This is a flowchart illustrating a method for using anchor decoupler regions to generate at least a second-order temperature-compensated MEMS resonator devices.
[0027] Figure 22 The diagram shows Δf and Q for various anchor decoupler region configurations. ANC The image.
[0028] Figure 23 This is a flowchart illustrating a method for producing at least a second-order temperature-compensated MEMS resonator devices by using modifications to the anchor decoupler region and geometry.
[0029] Figure 24 The Δf and Q values for various anchor decoupler region configurations and geometry modifications are shown. ANC The image.
[0030] These figures are used to illustrate exemplary embodiments, but those skilled in the art will understand that the technology disclosed herein is not limited to the arrangements shown in these figures. Detailed Implementation
[0031] I. Overview While quartz crystal oscillators have been fundamental to timing and frequency reference applications for the past century, the rapid development of sensor-based electronics has highlighted certain limitations of the technology, such as power consumption, robustness, size, and CMOS compatibility. Over the past two decades, silicon-based MEMS resonators have garnered significant attention due to their small size, low cost, and integration compatibility. However, MEMS resonators still cannot replace their quartz counterparts in many applications.
[0032] One drawback limiting the widespread adoption of MEMS resonators is their lack of temperature stability compared to quartz. Silicon MEMS resonators exhibit an inherent first-order temperature-induced frequency drift of approximately -30 ppm / °C, resulting in a temperature stability of approximately 3,750 ppm over an industrial operating temperature range of -40°C to 85°C. In contrast, AT-cut quartz resonators offer approximately 20 ppm of temperature stability over the same industrial operating temperature range. Several attempts have been made to overcome this temperature-induced frequency drift, such as methods involving the use of highly doped silicon substrates and composite materials. However, these methods are insufficient, as they only provide first-order compensation for the temperature-induced frequency drift.
[0033] The frequency change of a MEMS resonator relative to temperature is given by the following formula: Where f0 is the resonant frequency of the resonator at the reference temperature, ΔT is the deviation from the reference temperature, TCF1 is the first-order temperature coefficient of the frequency, and TCF2 is the second-order temperature coefficient of the frequency. For monocrystalline silicon, the value of TCF2 is typically -25 ppb / ℃. 2 Up to -80ppb / ℃ 2 Within a certain range, depending on the type and concentration of dopants, this can result in a temperature-induced frequency drift of approximately 200 ppm to 400 ppm over the industrial temperature range. While this second-order temperature-induced frequency drift is relatively small compared to the uncompensated temperature-induced frequency drift of silicon (approximately 3,750 ppm based on -30 ppm / °C), it is still significantly worse than the typical temperature-induced frequency drift that can be provided by AT-cut quartz crystals. Therefore, further reduction in the temperature-induced frequency drift of the fabricated silicon MEMS resonators is desired to enable their widespread adoption in the timing market.
[0034] While active temperature compensation techniques can be used to minimize temperature-induced frequency drift, such techniques can impose a significant burden on the system in terms of power consumption, circuit size, and circuit complexity. Therefore, it may be beneficial to find a method to address temperature-induced frequency drift in silicon MEMS resonators fabricated with passive compensation.
[0035] Various techniques have been previously used to passively compensate for temperature-induced frequency drift. For example, fabricating composite MEMS resonators using secondary materials such as silicon dioxide (SiO2) is a well-established technique. SiO2 is known to have a large positive TCF1 (approximately 85 ppm / °C), and due to its compatibility with silicon, it can be incorporated into silicon MEMS resonators to counteract the negative TCF1 of silicon. Within the industrial temperature range, this method has achieved temperature-induced frequency drift as low as sub-100 ppm. Within the same temperature range, for smaller devices operating at higher frequencies (such as thin-film bulk acoustic resonators), this technique has shown temperature-induced frequency drift as low as 3 ppm. However, the TCF2 effect of oxides (which is not well documented in the literature and is difficult to measure due to its small magnitude) and the loss in the resonator quality factor (Q) associated with thermoelastic damping and / or surface losses must be considered. Furthermore, this technique increases fabrication complexity and presents reliability issues due to film stress, leading to long-term aging problems.
[0036] Another example of existing passive compensation techniques involves fabricating MEMS resonators using highly doped silicon substrates. This is a well-established technique because the TCF1 for a particular resonant mode (including both shear and extended modes) varies with doping concentration (note that both the Lamé mode and the planar shear mode are examples of shear modes). For example, in an industrial temperature range of -40°C to 85°C, a highly doped silicon substrate can be used to limit the temperature-induced frequency drift of a MEMS resonator operating in Lamé mode to between 200 ppm and 400 ppm. However, even when the TCF1 becomes zero, a significant amount of temperature-induced frequency drift still exists due to the higher-order temperature coefficients of the frequency.
[0037] II. A General Method for Compensating Temperature-Induced Frequency Drift in Passive Compensating Silicon MEMS Resonators In the prior art discussed above, highly doped silicon substrates can help engineer TCF1 of Lamé-mode MEMS resonators to near zero. However, prior art is not efficient enough in compensating for high-order temperature-induced frequency drift in silicon MEMS resonators, hindering their competitiveness with quartz crystal resonators. Therefore, in the new techniques described herein, highly doped silicon substrates compensate for TCF2 rather than TCF1, and modifications to the resonator elements compensate for TCF1. Examples of such modifications to the resonator elements may include one or a combination of: (i) adding regions to the resonator elements; (ii) subtracting regions from the resonator elements; and (iii) changing the orientation of the resonator elements relative to the crystal axis of silicon. Using these techniques can provide at least second-order temperature compensation for MEMS resonator devices, thereby achieving enhanced intrinsic temperature stability (such as frequency drift of less than 1 ppm at temperatures between -40°C and 85°C) with appropriate design. Such techniques will be described in further detail below.
[0038] Figure 1 Flowchart 100 illustrates a general method for designing resonator elements for passively compensated MEMS resonator devices with temperature-induced frequency drift. The first stage of this method is to determine the type and concentration of dopant for which TCF2 equals zero for a resonator element defined by an initial geometry, an initial orientation relative to the crystal axis of silicon, and a set of associated eigenmodes. In this document, "eigenmode" refers to the natural vibration of the system such that all parts move together at the same frequency. This set of eigenmodes can be determined using finite element method (FEM) simulation software tools or any other software tools known now or developed in the future that can perform simulations.
[0039] At box 102, the method involves determining or otherwise defining an initial resonator geometry and an initial orientation relative to the crystal axis of silicon. This selection will then define the selection of eigenmodes supported by the resonator element. While a given resonator element can support many eigenmodes, it is not necessary to exhaustively test them, and the selection can be narrowed down by resonator design parameters such as the desired Q-factor and operating frequency range. This occurs at box 104, where a specific eigenmode is selected for analysis. Then, at box 106, multiple sets of parameters are determined or otherwise established, each of which defines a corresponding combination of (i) the type of dopant and (ii) the doping concentration.
[0040] At boxes 108 and 110, the method involves selecting a single set of parameters for analysis and calculating the TCF2 of that single set of parameters, respectively. The purpose of this process is to identify a single set of parameters that causes the TCF2 of a resonator element with a specific eigenmode to be approximately zero. As shown in box 112, if the TCF2 is approximately zero, the method proceeds to box 114. If not, the method returns to box 104 to identify an alternative set of parameters that causes the TCF2 to be approximately zero.
[0041] At box 114, once a set of parameters is identified that causes the TCF2 of the resonator element and the specific eigenmode to be approximately zero, the method involves calculating TCF1 for the type of dopant, dopant concentration, and resonator element with the specific eigenmode. At box 116, if the absolute value of TCF1 for the type of dopant, dopant concentration, and resonator element with the specific eigenmode is large, the method returns to box 104 to identify alternative types of dopant, dopant concentration, and specific eigenmode that cause TCF2 to be approximately zero. This process can be repeated until a type of dopant, dopant concentration, and resonator element with the specific eigenmode are identified that causes TCF2 to be approximately zero while also reducing the magnitude of TCF1; this is referred to herein as a “dense temperature compensated” (CTC) mode. If the combination of dopant concentration, dopant type, and resonator element with the specific eigenmode does not result in a CTC mode, the specific eigenmode is removed, and the method returns to box 104.
[0042] In some examples, the CTC mode corresponds to a mode in which the type of dopant, dopant concentration, and eigenmode of the resonator element cause TCF2 to be approximately zero, while also reducing the magnitude of TCF1 to below a predefined threshold. The predefined threshold for TCF1 leading to the CTC mode can be a fixed value sufficiently lower than the uncompensated value of TCF1. For example, since the uncompensated value of TCF1 in silicon is approximately 30 ppm / ℃, the predetermined threshold can be a lower value, such as 10 ppm / ℃, 5 ppm / ℃, or even 1 ppm / ℃. Alternatively, the TCF1 threshold leading to the CTC mode may not be a fixed value, but rather the value of TCF1 with the minimum magnitude calculated at box 114. In this way, the CTC mode corresponds to a mode with a smaller TCF1 magnitude than the non-CTC mode after TCF2 compensation. By reducing the magnitude of TCF1 in this way, the amount of subsequent modification used to compensate for TCF1 can also be reduced.
[0043] At box 118, once the CTC modes are identified, the next stage of the method is to apply modifications to the initial geometry and / or initial orientation of the resonator elements to slightly “distort” these CTC modes (introducing stray vibrations due to these modifications, resulting in a deviation from the pure mode shape). This distortion is achieved by altering the elastic constant of silicon (C... 11 C 12 and C 44 The contribution of the initial eigenmode to the initial eigenmode is used to modify the initial eigenmode. By applying modifications to the initial geometry and / or initial orientation of the resonator element to distort the CTC mode, the absolute value of TCF1 can be reduced, making it possible to design at least partially temperature-compensated MEMS resonator devices. In some implementations, fully second-order temperature-compensated MEMS resonator devices can be designed. A parameter sweep method using FEM is used to determine the type of modification that will result in at least partially temperature-compensated MEMS resonator devices.
[0044] In some examples, the methods shown in boxes 118 and 120 can be combined to apply in-plane rotation, which alters the initial orientation of the resonator element relative to the crystal axis of silicon. A sufficiently large set of different in-plane rotations should be tested to find an approximate rotation angle that minimizes the absolute value of TCF1.
[0045] In some examples, geometric modifications that alter the initial geometry of the resonator element can be applied in conjunction with the methods shown in boxes 122 and 124. These geometric modifications may include one or a combination of: (i) adding one or more regions to the resonator element, and (ii) subtracting one or more regions from the resonator element. The selection of geometric modifications utilizes a parametric sweep method, which requires establishing the selection of the geometry to be tested, systematically changing its dimensions and placement on the resonator element, and measuring the resulting TCF1. Therefore, for a given resonator element, multiple examples of geometric modifications that reduce TCF1 may exist.
[0046] In some examples, the methods shown in boxes 118 and 120 can be combined to apply in-plane rotations to the resonator elements before further modifying the geometry in boxes 122 and 124 using methods such as those described in the next section. Alternatively, in some examples, the methods shown in boxes 118 and 120 can be ignored based on resonator design requirements while still achieving the desired effect of reducing the absolute value of TCF1. Or, in some examples, the methods shown in boxes 122 and 124 can be ignored based on resonator design requirements while still achieving the desired effect of reducing the absolute value of TCF1. Finally, at box 126, the resulting design is a resonator design with at least a second-order temperature-compensated mode.
[0047] It should be understood that while the in-plane rotation and geometry modification techniques disclosed herein describe general methods for generating second-order temperature-compensated MEMS resonators, these techniques can be applied individually or in combination similarly. Furthermore, it should be understood that the techniques disclosed herein also describe a general method for reducing temperature-induced frequency drift in MEMS resonators without generating fully second-order temperature-compensated MEMS resonators.
[0048] Figure 2 Exemplary device configurations of second-order temperature-compensated MEMS resonators are shown. A first device configuration 200 illustrates a MEMS resonator device compatible with in-plane resonance modes. The MEMS resonator device includes at least partially temperature-compensated resonator element 202, at least one drive electrode 204 for in-plane electrostatic actuation of the resonator element 202, and at least one sensing electrode 206 for in-plane electrostatic sensing of the resonator element 202. The resonator element 202 is attached to a support structure 208 via one or more anchors 210. The anchors 210 may include side anchors and / or column anchors. A second device configuration 212 illustrates a MEMS resonator device compatible with out-of-plane resonance modes, having additional out-of-plane electrodes for out-of-plane actuation and sensing of the resonator element. The additional out-of-plane electrodes include at least one out-of-plane drive electrode 214 for out-of-plane actuation of the resonator element 202, and at least one out-of-plane sensing electrode 216 for out-of-plane sensing of the resonator element 202.
[0049] III. Frequency drift caused by temperature in a passively compensated square MEMS resonator via in-plane rotation As previously mentioned, different cuts of quartz produce different temperature-induced frequency drifts, with AT-cut quartz exhibiting the smallest drift, approximately 20 ppm, over the industrial temperature range. Similar methods can be used in silicon to reduce temperature-induced frequency drift, but instead of applying different cuts to silicon, alternative modifications can be applied in the form of in-plane rotation. In the literature, first-order temperature compensation has been achieved by applying specific combinations of doping concentrations and in-plane rotation to silicon. However, existing techniques are limited to controlling the first-order temperature coefficient of the frequency, while uncompensated higher-order TCFs result in significant temperature-induced frequency drift. In contrast, in the exemplary techniques described herein, high doping of the silicon substrate compensates for TCF2 rather than TCF1, while in-plane rotation of the resonator element compensates for TCF1.
[0050] a. Method Figure 3Flowchart 300 illustrates an exemplary method for designing a passive compensated silicon MEMS resonator device with in-plane rotation to reduce temperature-induced frequency drift. The method shown in Flowchart 300 is largely similar to that shown in Flowchart 100, except that it does not involve applying any geometric modifications that alter the geometry of the resonator element; instead, it involves applying in-plane rotation to the resonator element to reduce TCF1. As shown, at box 302, the method shown in Flowchart 300 involves defining the geometry of the resonator element and its initial orientation relative to the crystal axis of silicon. Here, since the method will ultimately involve applying in-plane rotation to the resonator element as described below, the resonator element geometry cannot be circular, as in-plane rotation has no effect on the first-order temperature coefficient of the frequency of a circular resonator element. Once the geometry and initial orientation of the resonator element are determined at box 302, the method continues in a manner similar to that shown in Flowchart 100, proceeding to box 318, where the CTC mode has been identified.
[0051] Once the CTC mode is identified, the next stage is to apply an in-plane rotation to the resonator element at box 318 to slightly “distort” the CTC mode (introducing stray vibrations due to these geometric modifications, resulting in a deviation from the pure mode shape). A sufficiently large set of different in-plane rotations should be tested to find the approximate rotation angle that minimizes the absolute value of TCF1 at box 320, where the method involves selecting the in-plane rotation that results in the minimum absolute value of TCF1 from the different applied in-plane rotations. Then, at box 322, after applying the in-plane rotation that minimizes the absolute value of TCF1, the resulting design is a resonator element design with at least partial temperature compensation.
[0052] b. Examples with in-plane resonant modes Figure 4A It shows that they can be combined Figure 3The method shown is an example of a square silicon resonator element 400. Resonator elements like this can resonate in many in-plane resonance modes, but only the Lamé mode is considered in this example. Furthermore, the only initial orientation considered relative to the crystal axis of silicon is the resonator element edge 402 with... <100> or <110> Those orientations that are aligned. The pure shear version of the Lamé resonance mode exists only at the edges of the silicon lattice. <100> and <110> In a square resonator element with aligned orientation. For the Lamé resonant mode, the associated mode extends in the same direction as the alignment direction of the resonator edges. Therefore, for in <100> direction or <110> The Lamé resonant mode structure is aligned in the direction, and the modes extend in the directions of... <100> (Lamé intrinsic modulus A) direction or <110> (Lamé eigenmode B) in the direction. Therefore, only two eigenmodes are considered.
[0053] Figure 4B Showing for Figure 4A The graph shows the TCF² values for different intrinsic modes versus the dopant concentrations for both N-type and P-type dopants. This allows us to determine whether there exists a combination of dopant concentrations and dopant types that results in a TCF² of approximately zero for a particular intrinsic mode. In silicon, the temperature coefficient of elasticity (TCE) for a given intrinsic mode, and the resulting TCF, depends on three independent elastic constants (Ci, Ci, Ci) in the mode extension direction. 11 C 12 and C 44 The dependence of temperature and doping on the mode extension. For example, for the mode extension along the... <110> The direction of the Lamé mode (Lamé eigenmode B), the value of TCE depends only on C. 44 The value of TCF2 is determined by its own temperature dependence and doping concentration dependence. The following shows the N-type and P-type doping concentrations required to reduce TCF2 to approximately zero for the considered eigenmode.
[0054] The values in Table 1 can be determined using FEM simulation software tools or any other software tools that can perform simulations, whether now known or developed in the future.
[0055] As shown in Table 1, the types of dopants, doping concentrations, and intrinsic modes that result in the smallest absolute value of TCF1 leading to the Lamé mode at TCF2≅0 are 2.3x10. 20 cm -3At a doping concentration of [value], the direction of mode extension is located [location]. <110> A P-type doped substrate in the direction of the dopant. By using the type, concentration, and eigenmode of the dopant, the resonator element will require less TCF1 compensation than other combinations of dopant type, concentration, and eigenmode. In this example, only a single CTC mode exists, where the dopant type, concentration, and eigenmode result in a relatively small TCF1 value, but multiple CTC modes may exist in other examples.
[0056] Using the CTC modes identified in Table 1 above, if the resonator element for plotting the identified CTC modes is rotated in different planes from 0° to 90° (relative to...) <110> Looking at the corresponding TCF1 values under the direction of rotation (34° and 55°), it can be seen that TCF1 has a minimum value of approximately 1.28 ppm / ℃. Figure 5 As shown. Due to the cubic lattice symmetry of silicon, this trend will repeat every 90°, so it may not be necessary to apply in-plane rotations outside this range. In-plane rotations of 0° and 90° correspond to mode extension directions as follows: <110> The direction of the Lamé mode (Lamé eigenmode B). Furthermore, when rotated within a 45° plane, the direction of mode extension is... <100> Orientation alignment results in a facet shear (FS) mode. At the mid-angle, the mode gradually transitions from the Lamé mode to the FS mode. Overall, while TCF1 cannot be compensated for in this case solely by rotation (since the lowest possible TCF1 achievable by rotation is approximately 1.28 ppm / °C), this may provide sufficient temperature compensation for some applications. Further examples disclosed herein will describe additional methods for further reducing TCF1.
[0057] c. Examples with out-of-plane resonant modes The method shown in flowchart 300 can also be implemented using resonator elements with out-of-plane resonant modes. In this example, both torsional and saddle modes are considered. Similar to the previous example, by plotting the TCF2 values versus dopant concentrations for both N-type and P-type dopant with different intrinsic modes, any combination of dopant concentration and dopant type that causes the TCF2 of the resonator element with a particular intrinsic mode to be approximately zero can be identified. Subsequently, the TCF1 for each identified combination of dopant concentration and dopant type that causes the TCF2 to be approximately zero can be determined to determine whether any CTC mode exists in which TCF1 is sufficiently reduced. The table below lists the CTC modes identified in this example for the resonator element with the considered particular intrinsic mode.
[0058] As mentioned earlier, once these CTC modes are identified, the next step is to apply in-plane rotation to the resonator elements so that these CTC modes are slightly “distorted” (due to the introduction of stray vibrations by these geometric modifications, resulting in a deviation from the pure mode shape).
[0059] Figure 6 The following diagram illustrates the resonator element relative to the CTC example identified in Table 2 above. <110> Graphs of TCF1 under various in-plane rotation angles. Figure 602 shows the effect of in-plane rotation on saddle-shaped intrinsic mode A. In this case, an in-plane rotation of approximately 10° results in a TCF1 of approximately 0 ppm / °C. Figure 604 shows the effect of in-plane rotation on torsional intrinsic mode A. In this case, an in-plane rotation of approximately 7.5° results in a TCF1 of approximately 0 ppm / °C. In both cases, this trend will repeat every 90° due to the cubic lattice symmetry of silicon. Here, unlike the exemplary resonator elements with in-plane resonant modes described above, these exemplary resonator elements with out-of-plane resonant modes can have their TCF1 fully compensated simply by using in-plane rotation.
[0060] IV. Passive Compensation for Temperature-Induced Frequency Drift in MEMS Resonators via Geometric Modification As previously discussed, high doping of the silicon substrate can compensate for second-order TCF2 rather than first-order TCF1, while modifications to the initial geometry of the resonator element can compensate for first-order TCF1. These geometric modifications can include adding regions to and / or subtracting regions from the resonator element. Using these techniques can provide at least second-order temperature compensation for MEMS resonator devices, thereby enabling enhanced intrinsic temperature stability (such as frequency drift of less than 1 ppm caused by temperatures between -40°C and 85°C) with appropriate design.
[0061] a. Method Figure 7 Flowchart 700 illustrates an exemplary method for designing a passively compensated MEMS resonator device with modified geometry. The method shown in flowchart 700 is largely similar to that shown in flowchart 100, except that the method shown in flowchart 700 does not involve applying any in-plane rotation to the resonator element, but only involves applying modifications to change the geometry of the resonator element to reduce the geometry of TCF1. As shown, at box 702, the method involves defining the initial resonator element geometry and orientation relative to the crystal axis of silicon. The method continues in a manner similar to that shown in flowchart 100 until box 722, where the CTC mode has been identified.
[0062] The next stage involves applying geometric modifications to the resonator elements to slightly “distort” these CTC modes (introducing stray vibrations due to these geometric modifications, resulting in a deviation from the pure mode shape). These geometric modifications can include adding regions to the resonator elements, subtracting regions from the resonator elements, or a combination of both. A parametric sweep method is used to select the geometric modifications that minimize the absolute value of TCF1. This method requires establishing the selection of the geometry to be tested, systematically changing its dimensions and placement on the resonator elements, and measuring the resulting TCF1. While the placement of the geometric modifications may differ between different implementations, in these cases, the added geometry is placed adjacent to or at least partially overlaps with the region of maximum displacement. Subtracted geometry can also be placed in this manner, or it can be placed along the axis of symmetry, depending on the resonator design requirements.
[0063] See again Figure 7 In flowchart 700, at block 722, geometric modifications such as those described above are applied until TCF1 is reduced to approximately zero at block 724. Then, at block 726, the resulting resonator design has at least a second-order temperature-compensated mode.
[0064] b. Example with initial square resonator element Figure 8 It shows that they can be combined Figure 7 The method shown considers an example of a square silicon resonator element. Resonator elements like this can resonate in many in-plane resonant modes, but... Figure 8 The examples shown only consider Lamé modes and planar shear modes. Pure shear versions of these resonant modes exist only at the edges of the silicon lattice. <100> and <110> In a square resonator element with aligned orientation. As mentioned above, for the Lamé resonant mode, the associated mode extends in the same direction as the alignment direction of the resonator edges. Therefore, for in <100> direction or <110> The Lamé resonant mode structure is aligned in the direction, and the modes extend in the directions of... <100> (Lamé intrinsic modulus A) direction or <110> (Lamé eigenmode B) direction. However, for the FS resonant mode, the associated mode extension direction is 45° relative to the alignment direction of the resonator edge. Therefore, for in <100> direction or <110> The FS resonant mode structure is aligned in the direction, and the directions of mode extension are respectively <110> (FS intrinsic modulus B) or <100> (FS intrinsic mode A). In these embodiments, the direction of mode extension is a convenient way to identify the intrinsic mode. In alternative embodiments, the direction of mode extension may not be limited, and alternative features may be used to identify the intrinsic mode.
[0065] Similar to the above combination Figure 4BThe example described above, by plotting the TCF2 values versus dopant concentrations for N-type and P-type dopants of different intrinsic modes, can determine whether there exists a combination of dopant concentrations and dopant types that causes the TCF2 of a particular intrinsic mode to be approximately zero. The following shows the TCF2 values for silicon lattices in... <100> direction and <110> The directional mode stretching reduces the TCF2 of the Lamé mode and FS mode to approximately zero N-type doping concentration and P-type doping concentration.
[0066] The values in Table 3 can be determined using FEM simulation software tools or any other software tools that can perform simulations, whether now known or developed in the future.
[0067] As shown in Table 3, the types of dopants, doping concentrations, and intrinsic modes that result in sufficiently small absolute values of TCF1 for both the Lamé and FS modes are within 2.3 x 10⁻⁶. 20 cm -3 and 2.34x10 20 cm -3 At a doping concentration of [value], the direction of mode extension is located [location]. <110> The substrate is P-type doped in the direction of the dopant. Therefore, in this example, there are two CTC modes, where the type of dopant, doping concentration, and intrinsic mode result in a relatively small TCF1 value. By using any of these combinations of dopant type, doping concentration, and intrinsic mode, the resonator element will require less TCF1 compensation than other combinations of dopant type, doping concentration, and intrinsic mode.
[0068] Once these CTC modes are identified, the next stage is to apply geometric modifications to the resonator elements to slightly “distort” these CTC modes (introducing stray vibrations due to these geometric modifications, resulting in a deviation from the pure mode shape). As shown in Table 3, for the identified CTC modes, the FS mode has a smaller TCF1 (approximately 2.93 ppm / ℃) than the Lamé mode (approximately 4.57 ppm / ℃). Therefore, the FS mode may require smaller geometric modifications to compensate for the TCF1.
[0069] Figure 9 and Figure 10 Examples of modifications to the geometry of the TCF1 that can be applied to Lamé-mode or FS-mode resonator elements are shown. Based on the discussion above in conjunction with Table 2, the Lamé-mode resonator elements are based on… <110> The directional pattern stretches (Lamé eigenmode B), and at 2.3 x 10 20 cm -3The doping concentration is 4.57 ppm / ℃ due to the presence of P-type dopant; the FS-mode resonator element is based on... <110> The mode stretching in the direction (FS eigenmode B), and at 2.34x10 20 cm -3 The doping concentration is 2.93 ppm / ℃, resulting in TCF1 with P-type dopant. Similar to the effect of in-plane rotation described above, adding or subtracting regions from a resonator element will similarly distort the eigenmodes. Therefore, after adding or subtracting regions from an FS-mode resonator element (in... Figure 9 The image shows FS mode resonator element 902, in Figure 10 The modified resonator elements shown are FS-mode resonator elements 1002, referred to herein as "modified FS-mode resonator elements," examples of which are shown in [the document]. Figure 9 The image shows a modified FS-mode resonator element 904. Figure 10 The image shows a modified FS-mode resonator element 1004. Similarly, after adding or subtracting regions from a Lamé-mode resonator (in...) Figure 9 The image shows Lamé mode resonator element 906, in Figure 10 The modified resonators shown are designated as Lamé-mode resonator elements 1006. These modified resonators are referred to herein as "modified Lamé-mode resonator elements," and examples of them are shown in [the document / examples]. Figure 10 The modified Lamé mode resonator element 908 and the modified Lamé mode resonator element 1008 are shown in the figure.
[0070] Figure 9 An example is shown involving modifications to the geometry of a region added to the geometry of a resonator element. As shown, both the FS-mode resonator element 902 and the Lamé-mode resonator element 906 have an initial region (A). O To compensate for TCF1, an additional region can be added to the resonator element, resulting in a new modified region (A) for the resonator element. M For the modified FS-mode resonator element 904, an additional region can be added near the corner of resonator element 902, as this is where the maximum displacement occurs, and therefore the geometric modification will have the greatest impact on TCF1. For the modified Lamé-mode resonator element 908, an additional region can be added near the side of the resonator element, as this is where the maximum displacement occurs, and therefore the geometric modification will have the greatest impact on TCF1. In either case, the size and / or position of the added region can be adjusted until TCF1 decreases to approximately zero. Figure 9As shown, in order to reduce the TCF1 of the modified FS-mode resonator element 904 to approximately zero, the corner region of the resonator element 902 can be extended until A. M / A O The ratio is 1.29. Similarly, in order to reduce the TCF1 of the modified Lamé mode resonator element 908 to approximately zero, the side region of the resonator element 906 can be extended up to A. M / A O The ratio is 5.8.
[0071] Figure 10 An example is shown involving modifications to the geometry of a region by subtracting it from the geometry of the resonator element. Similarly, as shown, both the FS-mode resonator element 1002 and the Lamé-mode resonator element 1006 have an initial region (A). O To compensate for TCF1, a region can be subtracted from the resonator element to form one or more cavities within the resonator element, resulting in a new modified region (A) of the resonator element. M The region subtracted from resonator element 1002 relative to the modified FS-mode resonator element 1004 can be a square region aligned with the midpoint of the side of the square resonator element 1002. Similarly, the region subtracted from resonator element 1006 relative to the modified Lamé-mode resonator element 1008 can be a square region offset from the midpoint of the side of the square resonator element 1006 and arranged along the diagonal of the square resonator element 1006. In either case, the size and / or position of the added region can be adjusted until TCF1 is reduced to approximately zero. Figure 10 As further shown, in order to reduce the TCF1 of the modified FS-mode resonator element 1004 to approximately zero, the size of the subtracted region can be selected such that A M / A O The ratio is 0.977. Similarly, in order to reduce the TCF1 of the modified Lamé mode resonator element 1008 to approximately zero, the size of the subtracted region can be chosen such that A M / A O The ratio is 0.956.
[0072] It should be understood that these geometric modifications that can be applied to compensate for TCF1 are not limited to the embodiments disclosed herein, but may also include embodiments that add or subtract any other shapes based on resonator design requirements to achieve the desired effect of reducing TCF1 to approximately zero. Other embodiments may include, in addition to Figure 9 and Figure 10Outside the region shown, additions or subtractions are made in different regions of the resonator. For structural subtraction, the subtracted region can be distributed differently on the resonator with more or fewer subtracted regions. Other embodiments may include both adding and subtracting regions in different regions of the resonator. Additionally, it should be understood that the ratio (A) of the modified structure to the original structure region required for TCF1 compensation... M / A O Modifications specific to these geometries will result in different ratios for different shapes being added or subtracted.
[0073] Figure 11 An exemplary resonant mode of the Lamé-mode resonator element 1102 is shown, along with changes in the resonant mode caused by modifications to the geometry of the resonator element 1102. For example, Figure 11 The diagram shows a second-order temperature-compensated resonator element 1104 obtained by subtracting the square region from the diagonal of the Lamé-mode resonator element 1102 and operating in the modified Lamé mode. Figure 11 Further shown is another resulting second-order temperature-compensated resonator element 1106 operating in modified Lamé mode after a rectangular region is added to the edge of the Lamé mode resonator element 1102. Figure 11 Further shown is another resulting second-order temperature-compensated resonator element 1108 operating in modified Lamé mode after adding a trapezoidal region to the edge of the Lamé mode resonator element 1102.
[0074] Figure 12 An exemplary resonant mode of the FS-mode resonator element 1202 is shown, along with changes in the resonant mode caused by modifications to its geometry. For example, Figure 12 The diagram shows a second-order temperature-compensated resonator element 1204 obtained by subtracting the square region from the FS-mode resonator element 1202 and operating in the modified FS-mode. Figure 12 Further shown is another second-order temperature-compensated resonator element 1206 that operates in modified FS mode after a generally square region is added to the corner of the FS mode resonator element 1202. Figure 12 Further illustrating another second-order temperature-compensated resonator element 1208 operating in modified FS mode after adding a generally elliptical region to the corner of the FS mode resonator element 1202, such that the resulting second-order temperature-compensated resonator element 1208 is generally x-shaped with circular endpoints.
[0075] c. Examples of resonator elements with various initial shapes To illustrate how the techniques described herein can be applied to resonator elements of various initial shapes and modifications to their geometries, in addition to square resonator elements, the following examples also involve... Figure 7 The method illustrated is applied to both circular and octagonal resonator elements. Similar to the previous example, by plotting the TCF2 values versus dopant concentrations for both N-type and P-type dopants of different intrinsic modes, any combination of dopant type and dopant concentration that causes the TCF2 of a particular intrinsic mode to be approximately zero can be identified. Subsequently, the TCF1 for each identified combination of dopant concentration and dopant type that causes the TCF2 to be approximately zero can be determined to determine whether any CTC mode exists in which TCF1 is sufficiently reduced. The table below lists the CTC modes identified in this example for circular, octagonal, and square resonator elements with the specific intrinsic modes considered.
[0076] As mentioned earlier, once these CTC modes are identified, the next step is to apply geometric modifications to the resonator elements so that these CTC modes are slightly “distorted” (due to the introduction of stray vibrations by these geometric modifications, resulting in a deviation from the pure mode shape).
[0077] Figure 13 and Figure 14 Examples are shown of geometries that can be applied to circular and octagonal resonator elements to produce second-order temperature-compensated MEMS resonators. For example, Figure 13 The circular resonator element 1302, operating in a wine glass mode, is shown before any geometric modifications. Figure 13 The second-order temperature-compensated resonator element 1304, obtained by subtracting a region from the circular resonator element 1302 and operating in a "modified wine glass mode", is further shown. Figure 13 Further shown is another resulting second-order temperature-compensated resonator element 1306 operating in a modified wine glass mode after a generally elliptical region is added to the circular resonator element 1302. Figure 13 Further shown is another resulting second-order temperature-compensated resonator element 1308, which operates in a modified wine glass mode after a generally square region is added to the circular resonator element 1302.
[0078] See next. Figure 14 This shows an octagonal resonator element 1402 operating in a wine glass pattern before any geometric modifications. Figure 14 The second-order temperature-compensated resonator element 1404, obtained by subtracting a region from the octagonal resonator element 1402 and operating in a "modified wine glass mode", is further shown. Figure 14 Further shown is another resulting second-order temperature-compensated resonator element 1406, which operates in a modified wine glass mode after adding generally circular regions to some sides of the octagonal resonator element 1402. Figure 14 Further shown is another resulting second-order temperature-compensated resonator element 1408, which operates in a modified wine glass mode after adding square regions to some sides of the octagonal resonator element 1402.
[0079] d. Examples of resonator elements that resonate in out-of-plane mode To illustrate how the techniques described herein can be applied to resonator elements with out-of-plane eigenmodes, the following examples involve... Figure 7 The method illustrated is applied to resonator elements with saddle-shaped and torsional resonant modes. Similar to the previous example, by plotting the TCF2 values versus dopant concentrations for both N-type and P-type dopant for different eigenmodes, any combination of dopant concentration and dopant type that causes the TCF2 of a particular eigenmode to be approximately zero can be identified. Subsequently, the TCF1 for each identified combination of dopant concentration and dopant type that causes the TCF2 to be approximately zero can be determined to determine whether any CTC mode exists in which TCF1 is sufficiently reduced. Table 2 above lists the CTC modes identified for a square resonator element with saddle-shaped and torsional resonant eigenmodes considered in this example.
[0080] As mentioned earlier, once these CTC modes are identified, the next stage is to apply geometric modifications to the resonator element to slightly “distort” these CTC modes (due to the introduction of stray vibrations from the pure mode shape caused by these geometric modifications). In this example, to reduce the first-order temperature coefficient of the frequency, the region added to the resonator element is located in the out-of-plane direction.
[0081] Figure 15 Examples of geometric modifications that can be applied to square resonator elements with out-of-plane resonant modes to produce second-order temperature-compensated MEMS resonators are shown. For example, Figure 15 The square resonator element 1502 is shown operating in torsion mode before any geometric modifications are applied to it. Figure 15 The second-order temperature-compensated resonator element 1504, obtained by operating in a "modified torsional mode" after adding a region to the resonator element 1502, is further illustrated. As shown, the region added to the resonator element 1502 to obtain the obtained resonator element 1504 is added out of plane relative to the unmodified resonator element 1502. That is, the added region includes a square area at the corner of the resonator element 1502 both above and below the plane of the unmodified resonator element 1502. Figure 15Further illustrated is another resulting second-order temperature-compensated resonator element operating in a modified torsional mode after subtracting a region from resonator element 1502. As shown, the subtracted region comprises a square area subtracted along the diagonal of resonator element 1502 near the center of resonator element 1502.
[0082] in addition, Figure 15 The square resonator element 1508 is shown in saddle mode before any geometric modifications are applied to it. Figure 15 The second-order temperature-compensated resonator element 1510, operating in a "modified saddle mode" after a region is added to the resonator element 1508, is further illustrated. As shown, the region added to the resonator element 1508 to obtain the resulting resonator element 1510 is added out of plane relative to the unmodified resonator element 1508. That is, the added region comprises a rectangular area at the edge of the resonator element 1508 above the plane of the unmodified resonator element 1508. Figure 15 Further shown is another resulting second-order temperature-compensated resonator element 1512 operating in a modified saddle mode after subtracting a region from resonator element 1508. As shown, the subtracted region comprises a rectangular area subtracted near the center of resonator element 1508 at a location aligned with the midpoint of the edge of resonator element 1508.
[0083] V. Temperature-induced passive compensation in square MEMS resonators via in-plane rotation and geometry modification. Frequency drift As mentioned above Figure 1 Based on resonator design requirements, it may be desirable to use both in-plane rotation and geometric modifications to passively compensate for temperature-induced frequency drift in MEMS resonators. For illustrative purposes, Figure 1 The method illustrated, involving both in-plane rotation of the resonator element and modification of the resonator element's geometry, has been applied... Figure 5 The Lamé-mode resonator element is shown after a 34° rotation. As shown in Table 1, this resonator element has a TCF1 of 4.57 ppm / ℃ before the in-plane rotation. After the 34° in-plane rotation, the TCF1 decreases to 1.28 ppm / ℃.
[0084] Once the CTC mode is identified and the initial orientation of the resonator element is changed by in-plane rotation, the next stage is to apply geometric modifications to the resonator element so that these CTC modes are slightly “distorted” (due to the introduction of stray vibrations by these geometric modifications, resulting in a deviation from the pure mode shape).
[0085] Figure 16An example is shown using a parametric sweep method to determine the modifications to the geometry of a second-order temperature-compensated MEMS resonator. Without geometric modifications, a 34° rotated Lamé-mode resonator element exhibits A... M / A O =1, while TCF1 is 1.28 ppm / ℃. For example... Figure 16 As shown in Figure 1602, the parameter sweep is illustrated, where the dimensions of the four square regions to be subtracted from the 34° rotated Lamé-mode resonator element vary, thereby reducing A. M / A O Until TCF1 in A M / A O =0.9951 is approximately zero. By applying an appropriate in-plane rotation before modifying the geometry, the extent of geometry modification required to reduce TCF1 to approximately zero can be significantly reduced. For example, in this example, after applying a 34° in-plane rotation, the amount of geometry modification required to compensate TCF1 via subtraction (in A) is... M / A O Characterization) is much smaller than that of the 0° / 90° Lamé mode resonator, where the amount of geometric modification required to compensate TCF1 is determined by A. M / A O =0.956 is characterized as described above and in Figure 10 As shown in the image.
[0086] like Figure 16 As further shown, Figure 1604 illustrates a parameter sweep where the dimensions of the four rectangular regions to be added to the 34° rotated Lamé-mode resonator element are varied, thereby increasing A. M / A O Until TCF1 in A M / A O =1.011 is approximately zero. Similarly, by applying an appropriate in-plane rotation before modifying the geometry, the extent of the geometry modification required to reduce TCF1 to approximately zero via addition can be significantly reduced. For example, in this example, the amount of geometry modification required to compensate for TCF1 after applying a 34° in-plane rotation (in A...) M / A O Characterization) is much smaller than that of the 0° / 90° Lamé mode resonator, where the amount of geometric modification required to compensate TCF1 is determined by A. M / A O =5.8 characterization, as described above and in Figure 9 As shown in the image.
[0087] Figure 17 Showing the use Figure 16The illustrated parameter scan identifies exemplary geometric modifications that can be applied to a 34° rotated Lamé-mode resonator to produce a second-order temperature-compensated MEMS resonator. For example, Figure 17 The Lamé mode resonator element 1702 is shown with a 34° rotation before the geometry was modified. Figure 17 Further shown is the subtraction of a square region near the center of resonator element 1702 to produce a result from A M / A O =0.9951 characterizes the modified resonator element region, thereby reducing TCF1 to approximately zero, resulting in a second-order temperature-compensated 34° rotated Lamé mode resonator element 1704. Figure 17 Further illustrated is the addition of a square region to the corner of resonator element 1702 to generate a shape formed by A. M / A O =1.011 characterizes the modified resonator element region, thereby reducing TCF1 to approximately zero, resulting in another second-order temperature-compensated 34° rotated Lamé mode resonator element 1706.
[0088] IV. Passive compensation for temperature-induced frequency drift in MEMS resonators via anchor decoupling region Based on the foregoing discussion, various methods may be desired to compensate for temperature-induced frequency drift in MEMS resonator devices. In addition to the methods described previously, these temperature-induced frequency drifts in a given MEMS resonator device can be compensated by forming an anchor decoupling region on the resonator element of that device. The resonator element can take any of a variety of shapes, such as square, rectangular, circular, elliptical, octagonal, hexagonal, triangular, etc.
[0089] As used herein, "anchor decoupling region" can refer to the region of a resonator element surrounding one or more anchors, wherein the anchor decoupling region includes one or more trench layers etched into the resonator element. The anchor decoupling region can play a role in reducing TCF1 to approximately zero and can be combined with other previously mentioned methods to form second-order compensated MEMS resonator devices.
[0090] Besides reducing TCF1 to approximately zero, the anchor decoupling region can also increase the Q-factor of the resonator element (compared to the same resonator element without the anchor decoupling region). For example, the presence of an anchor decoupling region on the resonator element can reduce the amount of energy transferred to the anchor during operation of the resonator element, thereby increasing the Q-factor of the resonator element. This can be particularly valuable because the Q-factor of anchor damping, Q0, is significantly increased. ANC This can have a significant impact on the overall Q-factor of the resonator elements. For example, the overall Q-factor of the resonator elements can be based on the Q-factor Q of the anchor damping.ANC Q factor of thermoelastic dissipation TED Q factor of Akhiezer damping AKH Q factor of extrusion membrane damping SFD And Q ANC It can be at least two orders of magnitude larger than other Q factors.
[0091] The anchor decoupler region can affect the TCF1 and Q factors to varying degrees, depending on the configuration of the anchor decoupler region. As used herein, "configuration" of the anchor decoupler region refers to a unique set of values for a number of variables that define the anchor decoupler region. These variables can take any of a variety of forms.
[0092] An anchor decoupler region variable can be the number of trench layers of the anchor surrounding the MEMS resonator device. In some implementations, the anchor decoupler region may include a single trench layer, while in other implementations, the anchor decoupler region may include two or more trench layers.
[0093] Another anchor decoupler region variable can be the number of trenches within each trench layer. In some implementations, a given trench layer may include a single trench that at least partially, but not completely, surrounds the anchor of the MEMS resonator device. Furthermore, in some implementations, a given trench layer may include multiple trenches, and each trench may at least partially surround the anchor of the MEMS resonator device.
[0094] Another variable in the anchor decoupler region can be the distance between the anchor and the trench layer of the MEMS resonator device. For example, a first trench layer may surround (e.g., partially surround) the anchor at a first distance (e.g., a first radial distance), and a second trench layer may surround (e.g., partially surround) the anchor at a second distance (e.g., a second radial distance). Relatedly, another variable in the anchor decoupler region can be the distance between two trench layers, for example, in an implementation where the anchor decoupler region includes multiple trench layers.
[0095] Another anchor decoupler region variable can be the trench width. For example, the trenches can be etched with various widths. In some examples, each trench can be etched with the same trench width. However, in other examples, some trenches can be etched with a first trench width, while other trenches can be etched with a second trench width, and so on. For example, the trenches in a first trench layer can have a first trench width, and the trenches in a second trench layer can have a second trench width. Various other examples may also exist.
[0096] Another variable for the anchor decoupler region can be the trench shape. The trench in the anchor decoupler region can take various shapes, including linear trenches, curved trenches, and other possible irregular shapes and combinations thereof (such as curved trenches with portions extending outward at various locations).
[0097] The anchor decoupler region can also be limited by other variables.
[0098] Based on the foregoing discussion, various anchor decoupler region configurations can exist, and each configuration can correspond to a unique set of values for the anchor decoupler region variables. The extent to which the anchor decoupler region (i) reduces TCF1 and (ii) increases the Q factor of the resonator element can vary, for example, based on the configuration of the anchor decoupler region. For instance, one anchor decoupler region configuration may reduce TCF1 to approximately zero but may only nominally increase the Q factor, while another anchor decoupler region configuration may have a limited effect on TCF1 but may significantly increase the Q factor. It should be understood that these examples are non-limiting, and the anchor decoupler region configuration can also affect TCF1 and the Q factor to various other degrees.
[0099] It may be desirable to identify anchor decoupler region configurations that simultaneously (i) reduce TCF1 to approximately zero and (ii) increase the Q factor, such that the Q factor satisfies a given threshold. The given threshold can take any of various forms and can be user-defined, such as, according to the foregoing discussion, such that the Q factor of the resonator element... ANC It is not a limiting factor for the overall Q factor of the resonator element. As described in more detail below, such anchor decoupler region configurations can be identified using FEM simulation software tools, and then this configuration can be formed on the resonator element.
[0100] Figure 18 An exemplary device configuration of a second-order temperature-compensated MEMS resonator device, including an anchor decoupling region, is shown in the figure, according to the foregoing discussion. Figure 18 The device includes a first MEMS resonator device 1800. The MEMS resonator device 1800 includes a resonator element 1802. The resonator element 1802 is attached to a substrate 1804 (or other support structure) via one or more anchors 1806, each anchor may be positioned at or near the center of the resonator element 1802, at or near the side of the resonator element 1802, or at any other location on the resonator element 1802. The resonator element 1802 also includes an anchor decoupling region 1808, which includes a trench layer 1810. Although not shown, the first MEMS resonator device 1800 may include other elements, such as one or more drive electrodes, one or more sensing electrodes, etc.
[0101] in addition, Figure 18 The device includes a second MEMS resonator device 1812, which includes features similar to those of the first MEMS resonator device 1800. Specifically, the second MEMS resonator device 1812 includes a resonator element 1814 that (i) is attached to a substrate 1816 via one or more anchors 1818, and (ii) includes an anchor decoupling region 1820 comprising a trench layer 1822. Similar to one or more anchors 1806, the one or more anchors 1818 may be positioned at or near the center of the resonator element 1814, at or near the side of the resonator element 1814, or at any other location on the resonator element 1814. Furthermore, although not shown, the second MEMS resonator device 1812 may include other elements such as one or more drive electrodes, one or more sensing electrodes, etc.
[0102] Figure 19 A cross-sectional view 1900 of a MEMS resonator device is shown, such as the previously described MEMS resonator device 1800 or MEMS resonator device 1812. Cross-sectional view 1900 may be similar in many respects to... Figure 2 The example shown is for simplicity, and only a portion of the elements shown in cross-sectional view 1900 are numbered. Specifically, cross-sectional view 1900 includes a resonator element 1902 attached to a substrate 1904 via an anchor 1906. A trench 1908 of a trench layer is also shown, which can be etched into the resonator element 1906 according to the foregoing discussion.
[0103] Figure 20 A top view is shown showing various possible anchor decoupler region configurations according to the foregoing discussion. For reference, a first top view 2000 is shown with anchor 2002 but without the anchor decoupler region.
[0104] A second top view 2004 is also shown, in which the anchor 2002 is now surrounded by an anchor decoupler region including a first trench layer 2006. As shown, the first trench layer 2006 includes four curved trenches surrounding the anchor 2002. Based on the foregoing discussion, the anchor decoupler region shown in the second top view 2004 can be configured in various ways according to the values selected for the anchor decoupler region variables described above.
[0105] A third top view 2008 is also shown, in which the anchor 2002 is now surrounded by an anchor decoupler region comprising a first trench layer 2010 and a second trench layer 2012. As shown, the first trench layer 2010 includes four trenches having a generally curved shape, but having portions extending outward from the anchor 2002 and toward the second trench layer 2012. On the other hand, the second trench layer 2012 includes four curved trenches that do not include portions extending outward from the anchor 2002. Based on the foregoing discussion, the anchor decoupler region shown in the third top view 2008 can be configured in various ways depending on the values selected for the anchor decoupler region variables described above.
[0106] Figure 21 Flowchart 2100 illustrates an exemplary method for designing a passively compensated MEMS resonator device, comprising forming an anchor decoupling region on the resonator element of the MEMS resonator device. The method shown in flowchart 2100 is largely similar to that shown in flowchart 100, except that instead of applying in-plane rotation to the resonator element or modifying its geometry, the method shown in flowchart 2100 involves forming an anchor decoupling region on the resonator element to reduce TCF1, etc. As shown, at box 2102, the method involves determining the initial resonator element geometry and orientation relative to the crystal axis of silicon. After identifying the CTC mode, the method continues in a manner similar to that shown in flowchart 100 until box 2118.
[0107] Boxes 2118 through 2122 can be repeated to determine the anchor decoupler region configuration according to the foregoing discussion, which (i) reduces the TCF1 of the resonator element to approximately zero and (ii) increases the Q factor of the resonator element to meet a threshold. Starting from box 2118, an anchor decoupler region configuration can be selected for testing, such as by selecting values for the previously described anchor decoupler region variables. The anchor decoupler region configuration can then be tested using FEM simulation software tools, etc., to determine how the anchor decoupler region configuration affects the TCF1 and Q factor of the resonator element. At box 2120, the method may include determining whether the TCF1 has been reduced to approximately zero, and at box 2122, the method may include determining whether the Q factor of the resonator element meets a threshold. If any or both of these conditions are not met, method 2100 can return to box 2118 and select another anchor decoupler region configuration for testing.
[0108] In some implementations, once it is determined that the selected anchor decoupler region configuration satisfies both conditions, method 2100 can proceed to block 2124, as described below. However, method 2100 can continue iterating through blocks 2118 to 2122 until a set of anchor decoupler region configurations is determined, wherein each anchor decoupler region configuration in the set (i) reduces TCF1 to approximately zero and (ii) increases the Q factor to satisfy the threshold. Once the set of anchor decoupler region configurations is determined, method 2100 can then select one configuration from the set (such as the best-satisfying anchor decoupler region configuration from the set) and proceed to block 2124 using the selected anchor decoupler region configuration.
[0109] At box 2124, based on the foregoing discussion, a selected anchor decoupling region configuration can be formed on the resonator element of the MEMS resonator device to obtain a MEMS resonator device with at least second-order temperature compensation, with the added benefit of an increased Q factor for the MEMS resonator device.
[0110] Figure 22 The TCF1 (denoted as Δf (ppm / ℃)) and Q are shown according to the various anchor decoupler region configurations discussed above. ANC The image. Figure 22 The anchor decoupler region configuration shown is (at least in part) limited by the number of trench layers and the trench width variable, although, based on the foregoing discussion, the anchor decoupler region configuration may also be limited by other variables.
[0111] Figure 23 Flowchart 2300 illustrates an exemplary method for designing a passively compensated MEMS resonator device. This method includes forming an anchor decoupling region on a resonator element of the MEMS resonator device and applying one or more geometric modifications to the resonator element (i.e., adding or removing one or more portions of the resonator element). The method shown in flowchart 2300 is largely similar to that shown in flowchart 2100, except that, in addition to selecting an anchor decoupling region configuration for testing, one or more geometric modifications can be selected to be tested together with the anchor decoupling region configuration. As shown, at box 2302, the method involves determining the initial resonator element geometry and orientation relative to the crystal axis of silicon. The method continues in a manner similar to that shown in flowchart 2100 until box 2318, where the CTC mode has been identified.
[0112] Boxes 2318 to 2324 can be repeated to determine a combination of anchor decoupler region configuration and one or more geometric modifications based on the foregoing discussion, which (i) reduces the TCF1 of the resonator element to approximately zero and (ii) increases the Q factor of the resonator element to meet a threshold.
[0113] Starting at box 2318, an anchor decoupler region configuration can be selected for testing, such as by selecting values for the previously described anchor decoupler region variables. Then, at box 2320, according to the foregoing discussion, one or more geometric modifications can be selected for testing. The combination of the selected anchor decoupler region configuration and the selected one or more geometric modifications can then be tested using FEM simulation software tools, etc., to determine how this combination affects the TCF1 and Q factor of the resonator element. According to the foregoing discussion, at box 2322, the method may include determining whether the TCF1 has decreased to approximately zero, and at box 2324, the method may include determining whether the Q factor of the resonator element meets a threshold. If any or both of these conditions are not met, method 2300 may return to boxes 2318 and / or 2320, and select another anchor decoupler region configuration and / or one or more geometric modifications for testing.
[0114] In some implementations, once it is determined that a combination of the selected anchor decoupler region configuration and one or more modifications of the selected geometry satisfies both conditions, method 2300 can proceed to block 2326, as described below. However, method 2300 can continue iterating through blocks 2318 to 2324 until a set of combinations of anchor decoupler region configurations and geometry modifications is determined, wherein each combination in this set (i) reduces TCF1 to approximately zero and (ii) increases the Q factor to satisfy a threshold. In some implementations, this may include exhaustively testing every possible combination of anchor decoupler region configurations and geometry modifications, while in other implementations, fewer than all possible combinations may be tested. Once the set of combinations of anchor decoupler region configurations and geometry modifications is determined, method 2300 can then proceed to selecting one combination of anchor decoupler region configurations and geometry modifications from this set (such as the combination that best satisfies the conditions from this set) and proceed to block 2326 using the selected combination.
[0115] At box 2326, as discussed above, a selected anchor decoupling region configuration can be formed on the resonator element of the MEMS resonator device, and one or more geometric modifications can be applied to the resonator element of the MEMS resonator device to obtain at least a second-order temperature-compensated MEMS resonator device, with the added benefit of an increased Q factor for the MEMS resonator device.
[0116] Figure 24 The TCF1 (denoted as Δf (ppm / ℃)) and Q are shown in the modified TCF1 and Q based on the various anchor decoupler region configurations and geometries discussed above. ANC The image. Figure 24 The anchor decoupler region configuration shown is (at least in part) defined by the number of trench layers and the trench width variable, although, based on the foregoing discussion, the anchor decoupler region configuration can also be defined by other variables. Furthermore, although... Figure 24 Only a few geometric modifications are shown in the figure, but any geometric modification can be tested during implementation.
[0117] Although not shown in the figure, forming an anchor decoupling region on the resonator element can be combined with applying rotation (in-plane or out-of-plane) to the resonator element to reduce TCF1 to approximately zero. For example, the TCF1 of the resonator element can be reduced to approximately zero using any of the following combinations: (i) forming an anchor decoupling region on the resonator element, (ii) applying a geometric modification to the resonator element, and / or (iii) applying rotation (in-plane or out-of-plane) to the resonator element.
[0118] Furthermore, the MEMS resonator devices described herein should be understood as having the capability to operate as dual-output MEMS resonator devices. Specifically, the MEMS resonator devices can be configured to operate in both in-plane and out-of-plane vibration modes.
[0119] VI. Conclusion It should be understood that the techniques described herein for reducing TCF2 or TCF1 to zero or approximately zero do not necessarily involve reducing TCF2 and TCF1 precisely to zero. Rather, it is sufficient to reduce TCF2 and / or TCF1 to as close to zero as possible. In some embodiments, this may involve reducing TCF2 to less than 1 ppb / ℃. 2 And / or reduce TCF1 to less than 1 ppm / ℃. In other embodiments, this may involve reducing TCF2 to less than 0.1 ppb / ℃. 2 And / or reduce TCF1 to less than 0.1 ppm / ℃. In other embodiments, this may involve reducing TCF2 to less than 0.01 ppb / ℃. 2 And / or reduce TCF1 to less than 0.01 ppm / ℃.
[0120] Furthermore, it should be understood that although the techniques disclosed herein are described in conjunction with square, circular, and octagonal resonators, these techniques can also be similarly applied to resonators of other shapes, such as rectangular, elliptical, or any other resonator shape. Additionally, while the techniques disclosed herein involving subtracting regions from resonator elements are described in conjunction with subtracting square regions, these techniques can also be similarly applied to subtracted regions of other shapes, such as rectangular, circular, or elliptical regions.
[0121] The techniques described herein can be used to design and fabricate microelectromechanical resonator devices. For example, once a MEMS resonator device is designed using these techniques, it can be fabricated using any semiconductor manufacturing technology now known or developed in the future; such a MEMS resonator device may include a support structure such as a silicon substrate, a resonator element capable of resonating at least partially in an in-plane resonant mode, at least one anchor coupling the resonator element to the support structure, at least one drive electrode, and at least one sensing electrode. In some examples, the device can be configured to operate as an oscillator.
[0122] Furthermore, when implementing the techniques described herein, the actual calculated value of the compensated TCF1, resulting from the formation of the anchor decoupling region, modifications to the geometry, and / or modifications to the orientation of the resonator elements, may differ slightly from the theoretical value collected from simulation tools. For example, extrapolation errors in the simulation model may cause the theoretical value of the compensated TCF1 to deviate from the actual value by a certain amount. Other inaccuracies in the simulation model or defects in the actual MEMS resonator device may also cause this difference. Therefore, when designing MEMS resonator devices using the techniques described herein, the extent to which the actual anchor decoupling region, geometric modifications, and / or rotations of the resonator elements applied during device fabrication may differ from the theoretical value by a certain amount. The magnitude of this amount can be determined in various ways, such as through experimentation, by changing the theoretical value by different amounts until the desired compensated TCF1 is achieved, or by employing any other applicable techniques.
[0123] While various aspects and embodiments have been disclosed herein, other aspects and embodiments will be apparent to those skilled in the art. The aspects and embodiments disclosed herein are for illustrative purposes only and are not intended to be limiting; the true scope and spirit are indicated by the following claims.
Claims
1. A MEMS resonator device, the MEMS resonator device comprising: Support structure; A resonator element, the resonator element being doped with at least one of an N-type dopant or a P-type dopant, wherein the doping concentration of the at least one of the N-type dopant or the P-type dopant causes a dense temperature compensation mode, wherein (i) the absolute value of the first-order temperature coefficient of the frequency of the resonator element decreases to a first value below a threshold, and (ii) the absolute value of the second-order temperature coefficient of the frequency of the resonator element decreases to approximately zero, and wherein an anchor decoupling region formed on the resonator element causes the absolute value of the first-order temperature coefficient of the frequency of the resonator element to further decrease to a second value less than the first value. and At least one anchor, which couples the resonator element to the support structure, wherein an anchor decoupling region is formed on the resonator element and at least partially surrounds the at least one anchor.
2. The MEMS resonator device of claim 1, wherein the anchor decoupling region includes at least one trench layer, the at least one trench layer including one or more trenches etched into the resonator element.
3. The MEMS resonator device of claim 1, wherein the threshold is a first threshold, and wherein the anchor decoupling region further reduces the amount of energy transferred to the at least one anchor during operation of the MEMS resonator device to satisfy a second threshold.
4. The MEMS resonator device of claim 1, wherein the MEMS resonator device is configured to operate as an oscillator.
5. The MEMS resonator device of claim 1, wherein the MEMS resonator device is configured to operate in an in-plane vibration mode and an out-of-plane vibration mode.
6. The MEMS resonator device of claim 1, wherein the MEMS resonator device is configured to operate in a single vibration mode including an in-plane vibration mode.
7. The MEMS resonator device of claim 1, wherein the MEMS resonator device is configured to operate in a single vibration mode including an out-of-plane vibration mode.
8. The MEMS resonator device of claim 1, wherein the geometric modification is applied to the resonator element to combine with the anchor decoupler region, causing the absolute value of the first-order temperature coefficient of the frequency of the resonator element to be further reduced to the second value.
9. The MEMS resonator device of claim 8, wherein applying the geometry modification includes adding one or more additional regions to the resonator element.
10. The MEMS resonator device of claim 8, wherein applying the geometry modification includes subtracting one or more regions from the resonator element.
11. The MEMS resonator device of claim 8, wherein applying the geometry modification includes adding one or more additional regions to the resonator element and subtracting one or more regions from the resonator element.
12. The MEMS resonator device of claim 1, wherein the in-plane rotation angle of the resonator element is combined with the anchor decoupling region, causing the absolute value of the first-order temperature coefficient of the frequency of the resonator element to be further reduced to the second value.
13. The MEMS resonator device according to claim 1, further comprising: At least one driving electrode, the at least one driving electrode being used for electrostatic actuation of the resonator element; and At least one sensing electrode is provided for electrostatic sensing of the resonator element.
14. A method for designing a MEMS resonator device with passive temperature-induced frequency drift compensation, the method comprising: For the resonator element of the MEMS resonator device, determine the doping concentration of at least one type of N-type or P-type dopant that causes a dense temperature compensation mode, in which (i) the absolute value of the first-order temperature coefficient of the frequency of the resonator element decreases to a first value below a threshold, and (ii) the absolute value of the second-order temperature coefficient of the frequency of the resonator element decreases to approximately zero. The resonator element is doped with at least one type of N-type or P-type dopant at the determined doping concentration; For the resonator element, an anchor decoupling region is determined to be formed on the resonator element, the anchor decoupling region causing the absolute value of the first-order temperature coefficient of the frequency of the resonator element to be further reduced to a second value less than the first value. as well as A defined anchor decoupling region is formed on the resonator element, the anchor decoupling region at least partially surrounding at least one anchor of the MEMS resonator device, the at least one anchor coupling the resonator element to the support structure of the MEMS resonator device.
15. The method of claim 14, wherein: For the resonator element, determining the anchor decoupling region includes: For the resonator element, one or more trench layers are defined, wherein each trench layer includes one or more trenches to be etched into the resonator element; and Forming the defined anchor decoupling region on the resonator element includes: The one or more trenches of the one or more trench layers are etched into the resonator element.
16. The method of claim 14, wherein the threshold is a first threshold, and wherein the anchor decoupler region further reduces the amount of energy transferred to the at least one anchor during operation of the MEMS resonator device to satisfy a second threshold.
17. The method of claim 14, wherein the MEMS resonator device is configured to operate as an oscillator.
18. The method of claim 14, wherein the MEMS resonator device is configured to operate in both in-plane and out-of-plane vibration modes.
19. The method of claim 14, wherein the MEMS resonator device is configured to operate in a single vibration mode including an in-plane vibration mode.
20. The method of claim 14, wherein the MEMS resonator device is configured to operate in a single vibration mode including out-of-plane vibration modes.
21. The method of claim 14, further comprising: For the resonator element, a geometric modification to be applied to the resonator element is determined to be combined with the anchor decoupler region, such that the absolute value of the first-order temperature coefficient of the frequency of the resonator element is further reduced to the second value. as well as The determined geometry modification is applied to the resonator element.
22. The method according to claim 21, wherein: For the resonator element, determining the modifications to be applied to the geometry of the resonator element includes: For the resonator element, determine one or more additional regions to be added to the resonator element; and Applying the determined geometry modifications to the resonator element includes: Add one or more additional regions to the resonator element.
23. The method according to claim 21, wherein: For the resonator element, determining the modifications to be applied to the geometry of the resonator element includes: For the resonator element, determine one or more regions to be subtracted from the resonator element; and Applying the determined geometry modifications to the resonator element includes: Subtract one or more regions from the resonator element.
24. The method of claim 21, wherein: For the resonator element, determining the modifications to be applied to the geometry of the resonator element includes: For the resonator element, determine one or more additional regions to be added to the resonator element; and For the resonator element, determine one or more regions to be subtracted from the resonator element; and Applying the determined geometry modifications to the resonator element includes: Add the one or more additional regions to the resonator element; and Subtract one or more regions from the resonator element.
25. The method of claim 14, further comprising: At least one driving electrode is formed on the MEMS resonator device, and the at least one driving electrode is used for electrostatic actuation of the resonator element; as well as At least one sensing electrode is formed on the MEMS resonator device, and the at least one sensing electrode is used for electrostatic sensing of the resonator element.
Citation Information
Patent Citations
Microelectromechanical Devices For Higher Order Passive Temperature Compensation and Methods of Designing Thereof
US20230131902A1
