MEMS and method for producing same

Abstract

A MEMS comprises a substrate having a cavity in the substrate. The MEMS comprises a fluid displacement device having an element that is suspended in the cavity, is arranged in the cavity parallel to a substrate plane, is movable in-plane and is designed to interact with a fluid in the cavity. The fluid displacement device has an actuator device that is designed to deflect main sides of the movable element in-plane alternately along a first direction so as to displace a first fluid volume and along an opposing second direction so as to displace a second fluid volume. At least one of the main sides of the movable element is curved in an unactuated state. The displaced first fluid volume and the displaced second fluid volume are substantially identical.

Description

[0001] MEMS and method of manufacturing the same

[0002] Description

[0003] The present invention relates to a MEMS configured to displace a fluid and, for this purpose, having a main surface that is curved in an unactuated state. The present invention further relates to a method for manufacturing such a MEMS. Embodiments of the present invention relate to reducing strain stiffening in the fin plane while simultaneously increasing stiffness in the actuation plane or XZ plane.

[0004] The loudspeaker's objective is to translate sound reproduction of any complex signal as linearly as possible. Currently, only passive reduction of strain stiffening is possible in MEMS loudspeakers. Active reduction of strain stiffening, as in MEMS resonators, which uses the electrical signal for compensation, is not possible without additional signal processing, as the component must transmit a broadband signal as linearly as possible.

[0005] A passive reduction of strain stiffening can be achieved by varying the beam thickness [1], but this has the disadvantage that the trench width must be varied, which requires varying etch rates and can lead to under-etching of fins. It is also possible to achieve a passive reduction using curved elements along the beam, so-called arch beams [2, 3]. The disadvantage of this is that the displaced volume is not the same in the forward and backward directions, which causes buckling and even harmonic distortion. It is also possible to implement a passive reduction using a spring-loaded suspension of the beam. However, this has the disadvantage that the suspension takes up a lot of space, although to maximize the sound pressure level (SPL), as many fins as possible should be present on the chip.For the implementation of air-displacing elements in electrostatic microloudspeakers in multilayer systems, straight fins are used, as proposed in [4, 5]. These exhibit a warping in the center of the fin at higher frequencies. This warping reduces the displaced volume and thus the maximum achievable sound pressure level. This has led to a previous effort to develop fin dimensions for which this warping is less pronounced. Therefore, there is a need for MEMS and their manufacturing processes that can generate high pressures, such as sound pressure levels, without interference.

[0006] The object of the present invention is therefore to provide a MEMS and a method for producing the same which makes it possible to generate high pressures in a moving fluid and at the same time provide a high linearity with respect to an input signal.

[0007] This problem is solved by the subject matter of the independent patent claims.

[0008] According to one embodiment, a MEMS comprises a substrate with a cavity in the substrate. The MEMS comprises a fluid displacement device with an element suspended in the cavity and suspended parallel to a substrate plane in the cavity and movable in-plane, which element is designed to interact with a fluid in the cavity. The fluid displacement device comprises an actuator device designed to deflect main sides of the movable element alternately along a first direction to displace a first fluid volume and along an opposite second direction to displace a second fluid volume in-plane. At least one of the main sides of the movable element is curved in an unactuated state. The displaced first fluid volume and the displaced second fluid volume have a substantially equal volume size.

[0009] A method for producing such a MEMS comprises providing a substrate with a cavity in the substrate and arranging a fluid displacement device with an element suspended in the cavity and arranged parallel to a substrate plane in the cavity and movable in-plane, which element is designed to interact with a fluid in the cavity. Furthermore, an actuator device of the fluid displacement device is provided, which is designed to deflect main sides of the movable element alternately along a first direction for displacing a first fluid volume and along an opposite second direction for displacing a second fluid volume in-plane. The method is carried out such that at least one of the main sides of the movable element is curved in an unactuated state.Furthermore, the method is carried out such that the displaced first fluid volume and the displaced second fluid volume have essentially the same volume size. The curved main side in the unactuated state enables high dimensional stability of the element during actuation and movement, so that high deflections and thus high sound pressures can be achieved. At the same time, the high dimensional stability enables small deformations and thus high linearity, on the one hand. By maintaining the essentially equal volume sizes, it also reduces the total harmonic distortion (THD), which enables a high linear signal quality. The high dimensional stability can be particularly advantageous at comparatively high sound pressures, where deformation or yielding of the fin shape would flatten the force / sound pressure characteristic curve.

[0010] According to one embodiment, a curvature of the main side parallel to the substrate plane is inversion-symmetric, approximately relative to a center point of the movable element. This enables simple and precise control of the movable element. Alternatively, it is also possible to deviate from an inversion-symmetric design. For this purpose, a pre-deflection or bias is preferably applied via the actuator device in order to compensate for potentially differing volume displacements via an offset / bias.

[0011] According to one embodiment, the movable element is suspended in-plane on both sides of the substrate and possibly symmetrically, at least with regard to the magnitude of the acting forces. The two-sided suspension increases the rigidity toward the lid and thus reduces the risk of pull-in.

[0012] According to one embodiment, the curvature of the main side essentially follows a sine or cosine function. This enables both high stability perpendicular to the substrate plane and consistent force distribution in the movable element during deformation.

[0013] According to one embodiment, a path of the movable element parallel to the substrate plane from a first end to a second end of the movable element has at least two sign changes of a derivative of a curvature path of the curvature, for example, by providing a higher number of arcs, bends, or kinks. This enables further stabilization along a direction perpendicular to the substrate plane. According to one embodiment, the main side of the movable element is continuously curved in the unactuated state. This makes it possible to avoid force peaks in a material of the movable element.

[0014] According to one embodiment, the main side of the movable element is formed in sections that are straight in the unactuated state. Straight sections allow for the precise design or configuration of the deflection along the movable element, for example, to realize coupling points with minimal tilting when using a comb drive.

[0015] According to one embodiment, a path of the movable element parallel to the substrate plane is configured to have at least two arcs of curvature that have a different curvature amplitude along a deflection direction of the movable element. This enables advantageous adaptation of the movable element to a particular application environment.

[0016] According to one embodiment, the MEMS comprises a plurality of adjacently arranged movable elements whose curved main sides face each other or have main curvatures facing away from each other. This allows for the adjustment of high or, alternatively, low fluidic damping, depending on the requirements of the application environment.

[0017] According to one embodiment, a plurality of adjacently arranged movable elements are arranged such that they have a main curvature that is the same or opposite to one another.

[0018] According to one embodiment, the MEMS is designed such that a path of the movable element parallel to the substrate plane has an overhang from a first end to a second end. This enables the provision of comparatively large amounts of material of the movable element, which is available for expansion during actuation.

[0019] According to one embodiment, the actuator device comprises a stator electrode arrangement with at least a first stator electrode and a second stator electrode, which are arranged on the substrate or are part of the substrate. The actuator device is configured to deflect the movable element in the first direction based on an electrostatic force between the movable element and the first stator electrode and to deflect the movable element in the second direction based on an electrostatic force between the movable element and the second stator electrode. This enables the advantageous generation of symmetrical forces along both directions, which is advantageous for maintaining substantially equal fluid volumes.

[0020] According to one embodiment, the movable element is arranged centrally opposite the first stator electrode and the second stator electrode. This allows forces to be generated by both stator electrodes, which enables precise control.

[0021] According to one embodiment, the movable element is arranged opposite the first stator electrode and the second stator electrode in every operating position. This enables precise control of the movable element without any dead points that would be detrimental to linearity.

[0022] According to one embodiment, an electrode geometry of a stator electrode of the actuator device has a curvature that is adapted to a curvature of the movable element, for example, in the actuated or unactuated state. This enables a high quality of the electrostatically generated force due to the adapted shape.

[0023] According to one embodiment, the actuator device comprises a first stator electrode arrangement and a second stator electrode arrangement opposite and perpendicular to the substrate plane, wherein the movable element is arranged between the first stator electrode arrangement and the second stator electrode arrangement. This enables the generation of electrostatic forces at opposing regions of the movable element, which allows for the avoidance or deliberate generation of torsion of the movable element.

[0024] According to one embodiment, the movable element is substantially uncurved in the unactuated state along a direction perpendicular to the substrate plane. This enables high linearity in the generated fluid pressure. According to one embodiment, the MEMS has at least one opening in the substrate configured for fluid flow. This opening is arranged along a direction parallel to a surface normal arranged perpendicular to the substrate plane and has an opening geometry adapted to a curvature of the movable element, possibly in the actuated state or in the unactuated state. This enables the deliberate adjustment or avoidance of flow resistances in the MEMS.

[0025] Further advantageous embodiments are the subject of dependent patent claims.

[0026] Preferred embodiments of the present invention are explained below with reference to the accompanying drawings, in which:

[0027] Fig. 1 is a schematic plan view of at least parts of a MEMS according to an embodiment of the first aspect;

[0028] Fig. 2a is a schematic side sectional view of a MEMS according to an embodiment of the first aspect, illustrating an embodiment of the suspension of a movable element;

[0029] Fig. 2b is a schematic side sectional view of part of the view of Fig. 2a in a rotated representation;

[0030] Fig. 3a is a schematic plan view of a part of a MEMS according to an embodiment of the first aspect, which can be used, for example, as a MEMS according to Fig. 1 and / or according to Fig. 2a;

[0031] Fig. 3b is a schematic plan view of parts of a MEMS according to an embodiment of the first aspect, having an alternative configuration of the movable element compared to the MEMS of Fig. 3a;

[0032] Fig. 3c is a schematic plan view of the MEMS according to Fig. 3a for discussing a further embodiment;

[0033] Fig. 3d is a schematic top view of a MEMS according to an embodiment that has additional mechanical advantages compared to Fig. 3c; Fig. 3e is a schematic top view of a MEMS according to an embodiment, with a rotated direction of movement;

[0034] Fig. 3f is a schematic plan view of a MEMS that implements the improvement explained in connection with Fig. 3d on the MEMS of Fig. 3e;

[0035] Fig. 4a-c show schematic plan views of MEMS according to embodiments of the first aspect with differently shaped movable elements;

[0036] Fig. 5a-b show schematic plan views of MEMS according to embodiments of the first aspect with differently shaped movable elements each with varying curvature amplitude;

[0037] Fig.6a-b are schematic representations for discussing an advantageous curvature of a movable element in embodiments of the first aspect;

[0038] Fig. 7a is a schematic perspective view of a multi-corrugated fin according to an embodiment of the first aspect arranged in edge regions on a solid substrate;

[0039] Fig. 7b shows a positioning of the fin corresponding to Fig. 7a;

[0040] Fig. 8a is a schematic perspective view of a part of a MEMS according to an embodiment of the first aspect with a cap layer;

[0041] Fig. 8b is a schematic plan view of a part of a MEMS according to an embodiment of the first aspect, in which four movable elements are arranged by way of example;

[0042] Fig. 9a-c show schematic views of a portion of a MEMS according to an embodiment of the first aspect, in which the actuator device comprises a stator electrode arrangement; Fig. 10 shows a schematic perspective view of at least a portion of a MEMS according to an embodiment of the first aspect, in which a curved shape of the stator electrodes is implemented;

[0043] Fig. 1 1 a is a schematic plan view of possible courses or geometries of movable elements of a MEMS according to an embodiment of the first aspect;

[0044] Fig. 1 1 b is a schematic plan view of at least a part of a MEMS according to an embodiment of the first aspect, in which the movable element can have an overhang when extending along an x-direction;

[0045] Fig. 12a is a schematic plan view of parts of a MEMS according to an embodiment of the first aspect, having a plurality of adjacently arranged movable elements;

[0046] Fig. 12b is a schematic perspective view of the MEMS of Fig. 12a;

[0047] Fig. 13a is a schematic plan view of at least parts of a MEMS according to an embodiment of the first aspect, in which a plurality of at least two movable elements arranged adjacent to one another are also provided, in which main curvatures oriented opposite to one another are also implemented, in such a way that the main curvatures face away from one another;

[0048] Fig. 13b is a schematic perspective view of the MEMS of Fig. 13a;

[0049] Fig. 14 is a schematic plan view of at least parts of a MEMS according to an embodiment of the first aspect, in which a plurality of movable elements are arranged adjacent to one another;

[0050] Fig. 15 is a schematic top view of at least parts of a MEMS according to an embodiment of the first aspect, in which a plurality of at least two movable elements are arranged on the substrate; Fig. 16 is a schematic flow diagram of a method 1600 according to an embodiment of the first aspect;

[0051] Fig. 17 is an exemplary schematic graph of a sound pressure level curve that can be generated in embodiments over the frequency to show the low frequency range and the middle frequency range to illustrate the first and second aspects of embodiments described herein;

[0052] Fig. 18 is a schematic plan view of a MEMS according to an embodiment of the second aspect;

[0053] Fig. 19 is a schematic plan view of a MEMS 190 according to an embodiment of the second aspect, in which a boundary structure has a curved side;

[0054] Fig. 20 is a schematic plan view of at least parts of a MEMS according to an embodiment of the second aspect, in which a boundary structure has a planar side, unlike in the MEMS;

[0055] Fig. 21 is a schematic plan view of at least parts of a MEMS according to an embodiment of the second aspect, which has an active element as a limiting structure;

[0056] Fig. 22 again shows the representation of the MEMS from Fig. 15 for discussing an advantageous embodiment according to the second aspect; and

[0057] Fig. 23 is a schematic flow diagram of a method for manufacturing a MEMS according to the second aspect of embodiments discussed herein.

[0058] Before exemplary embodiments of the present invention are explained in more detail below with reference to the drawings, it is pointed out that identical, functionally equivalent or equivalent elements, objects and / or structures are provided with the same reference numerals in the different figures, so that the description of these elements shown in different exemplary embodiments is interchangeable or can be applied to one another. The exemplary embodiments described below are described in connection with a large number of details. However, exemplary embodiments can also be implemented without these detailed features. Furthermore, for the sake of clarity, exemplary embodiments are described using block diagrams as a replacement for a detailed representation.Furthermore, details and / or features of individual embodiments can be easily combined with one another, unless explicitly described otherwise.

[0059] The following embodiments relate to micromechanical components (MEMS components). The MEMS components described herein can be multilayer structures. Such MEMS can be obtained, for example, by processing semiconductor materials at the wafer level, which can also involve a combination of multiple wafers and / or the deposition of layers at wafer levels. Some of the embodiments described herein address MEMS levels. A MEMS level, such as a substrate level, is understood to be a plane that is not necessarily two-dimensional or non-curved and extends essentially parallel to a processed wafer, for example parallel to a main side of the wafer or of the subsequent MEMS. A plane direction of such a plane or layer level can be understood as a direction within this plane, which can also be referred to by the term "in-plane."A direction perpendicular to this, i.e. perpendicular to a plane direction, such as the substrate plane, can be simply referred to as thickness direction or stacking direction, whereby the term thickness does not impose any limitation in the sense of an orientation of this direction in space.

[0060] In this regard, it is understood that terms used herein such as “length,” “width,” “height,” “top,” “bottom,” “left,” “right,” and the like are used merely to illustrate exemplary embodiments described herein, since their position in space can be changed as desired and the exemplary embodiments are therefore not to be restricted thereto.

[0061] Some of the embodiments described herein are described in connection with a loudspeaker configuration or loudspeaker function of a corresponding MEMS component. It is understood that these embodiments, with the exception of the alternative or additional function of sensory evaluation of the MEMS component or the movement or position of movable elements thereof, can be transferred to a microphone configuration or microphone function of the MEMS component, so that such microphones represent further embodiments of the present invention without limitation. A frequency range of MEMS described herein is not limited to the frequency range audible to humans, but can easily also be used as an alternative or additionally in the higher frequency range, such as the ultrasound range.

[0062] Embodiments of the present invention relate to moving a fluid within a volume or displacing a fluid volume. Starting with a loudspeaker configuration, such an implementation can also be used for a pump function, for example, using additional valves. MEMS pumps thus represent further embodiments of the present invention. In the case of a loudspeaker, a fluid displaced according to embodiments can, for example, be gaseous, such as air. However, a sound pressure level or a loudspeaker function, as well as a pump function, can also be easily implemented in a liquid fluid.

[0063] Embodiments of the present invention relate to compensating or adjusting influences acting on a movable element during the forward movement on the one hand and the return movement on the other, in order to avoid or minimize harmonic distortions in the obtained output signal. The problem solved by the invention exists in different frequency ranges, wherein a first aspect of the embodiments described herein has particular advantages with low attenuations, such as in the low frequency range down to approximately 100 Hz, but also across the entire frequency range in the case of low-damped systems, and a second aspect of the embodiments described herein solves the same problem with higher attenuations, for example in a higher frequency range and / or with small distances between movable elements and adjacent structures, for example implemented for a high packing density, for example at frequencies from approximately 100 Hz.While in the lower frequency range an equalization of the fluid volume displaced by means of a movable element on both sides of the movable element brings advantageous effects, it was recognized according to the invention that at medium frequencies above 100 Hz an equalization of the fluidic damping is particularly advantageous.

[0064] Fig. 1 shows a schematic top view of at least parts of a MEMS 10 according to an embodiment of the first aspect. For better understanding, but without any limiting effect, an exemplary Cartesian coordinate system with the x, y, and z directions is provided. An x / y plane can be understood as a substrate plane that extends parallel to a wafer plane or the like in the MEMS 10. The z direction can be understood as a height direction or thickness direction and is arranged, for example, perpendicular to the x / y plane.

[0065] The MEMS 10 comprises a substrate 12 with a cavity 14 arranged therein. The substrate 12 may, for example, comprise a semiconductor material such as silicon or the like, although the embodiments discussed herein are not limited to semiconductor materials or to silicon specifically. Thus, another electrically conductive or electrically non-conductive material could also be used as the substrate material 12, for example, plastic, glass, metal, or the like.

[0066] The MEMS 10 comprises a fluid displacement device 16. This device has an in-plane movable element 18 suspended in the cavity 14 and arranged parallel to the substrate plane x / y in the cavity 14. A suspension 22 of the movable element 18 can be provided on one side of the movable element 18, but is preferably provided on both sides. Suspension elements, such as solid-state springs or the like, can be used; however, the movable element 18 can also be formed integrally with a material of the substrate 12 and suspended thereby.

[0067] An actuator device 24 of the fluid displacement device 16 is configured to move main sides 18A and 18B of the movable element 18 alternately along opposite directions 26i and 262, for example, along the positive and negative y-direction. By moving the movable element 18 along the direction 26i, the fluid contained in the cavity 14 can be displaced from a first fluid volume 28i. Starting from a rest position, center position, or reference position, a movement of the movable element 18 along the opposite direction 262 can lead to the displacement of the fluid from a second fluid volume 28i. The main sides 18A and 18B of the movable element 18 can be deflected alternately along the directions 26i to displace the fluid volume 28i and along the opposite direction 262 to displace the fluid volume 282.This is done in-plane, i.e. along a direction parallel to the x / y plane.

[0068] At least one, but possibly both, main sides 18A and 18B are curved in an unactuated state of the movable element 18. In Fig. 1, the main side 18A is shown curved by way of example, which means that it has a changing course of the surface direction along the x / y plane. Possibly, but not necessarily, the main side 18B follows the course of the main side 18A in whole or in part, as described in connection with exemplary embodiments. A curved main side in the unactuated state can also be understood such that the movable element 18 has a curved neutral fiber parallel to the substrate plane. A corresponding curvature enables a minimization or reduction of the lid pull-in compared to an upper or lower MEMS layer and / or the avoidance of a plate mode in the movable element.

[0069] The fluid volumes 28i and 282 have essentially the same volume size. The fluid flowing through openings of the cavity 14 can therefore make an equal or comparable contribution to a pressure, such as a sound pressure level (SPL), for both movement contributions along the direction 26i on the one hand and 262 on the other, which is particularly advantageous with regard to avoiding or reducing THD. Minor deviations are permissible. Preferably, the first fluid volume and the second fluid volume are equal within a tolerance range of a maximum of 20%, a maximum of 15%, or a maximum of 10%, particularly preferably a maximum of 5%, in order to approximate the ideal state of 0% deviation.

[0070] The MEMS 10 can optionally include a control device 32 configured to apply a control signal 34 to the actuator device 24 and thus control a movement of the movable element 18. However, the control device 32 can also be a device independent of the MEMS 10 or external to it, which is to be coupled to the MEMS 10 at a later time.

[0071] The actuator device 24 can use different actuator principles to deflect the movable element 18. Thermal or magnetic actuators can possibly be used. Alternatively or additionally, it is also provided within the scope of exemplary embodiments to use piezoelectric and / or electrostatic / electrodynamic actuator principles. It is possible but not necessary for the actuator device to be arranged in part on or in the movable element 18. It is also not necessary for the actuator device to be arranged in a MEMS plane of the movable element. According to exemplary embodiments, it is also provided that a force-generating actuator device or the mechanism providing the movement is arranged in a different MEMS plane than the movable element and transmits the force and movement to the plane of the movable element 18, for example by means of coupling elements.

[0072] Fig. 2a shows a schematic side sectional view of a MEMS 20 formed in accordance with the MEMS 10 and illustrating one embodiment of the suspension of the movable element 18. The movable element 18 is, for example, directly or immediately connected to the substrate 12. A deformation of the movable element 18 and / or a movement of the movable element 18 can occur along the positive and negative y-direction. The actuator device can comprise a stator electrode arrangement 36 with at least one or two stator electrodes 38i and 38s. Stator electrodes 38i and / or 38s can comprise an electrically conductive material, such as a metal material or an electrically conductive semiconductor material. For example, the stator electrodes 38i and / or 38s can be formed from doped polysilicon.The movable element 18, which can also be referred to as a fin, can also be electrically conductive and / or electrically conductive layers or regions configured for use as an electrode can be arranged thereon or therein to enable actuation by electrostatic and / or electrodynamic forces. Optionally, a continuous or perforated or at least partially arranged electrical insulator 42i and / or 422 can be arranged between the stator electrode 38i and the movable element 18 and / or between the movable element 18 and the stator electrode 38i. Although Fig.2a shows an arrangement of the electrical insulator 42i on the stator electrode 38i and an arrangement of the electrical insulator 422 on the stator electrode 382, ​​at least one of the electrical insulators 42i and 422 can also be arranged on the movable element 18 or elsewhere in order to avoid an electrical short circuit between the electrical drive part connected to the movable element 18 and the stator electrode 38i or 382.

[0073] When implementing the electrostatic / electrodynamic drive, one of the two stator electrodes 38i and 382 is optional. However, using both stator electrodes (e.g., as bottom and cover electrodes) enables symmetrical force introduction along the z-direction, which can reduce or prevent torsion of the movable element 18 about the x-axis. Alternatively, a corresponding torsion can also be deliberately adjusted, if desired during operation. A respective gap 44i and 442 can be arranged between the movable element 18 along the positive and negative z-direction to enable the movement of the movable element 18. A dimension of the respective gap 44i and 442 can be used to adjust the fluidic resistance, preferably avoiding an acoustic short circuit or excessive fluid flow around or through the respective gap 44i and 442.With appropriate dimensioning, gaps 44i and 442 can also provide electrical insulation between the movable element 18 and at least one of the stator electrodes 38i and 382. Not shown is additional substrate material that can be located along the positive and / or negative z-direction to define the cavity, particularly in the event that one of the stator electrodes 38i and 382 is not required.

[0074] Fig. 2b shows a schematic side sectional view of part of the view from Fig. 2a in a rotated representation.

[0075] Fig. 3a shows a schematic top view of a portion of a MEMS 30i, which can be used, for example, as MEMS 10 and / or 20. The main sides 18A and 18B run essentially parallel to each other and are both curved. The movable element 18 essentially follows a sinusoidal shape, although alternatively, a cosinusoidal shape with pre-deflection is also readily possible. In this case, the curvatures of both main sides 18A and 18B parallel to the substrate plane x / y are inversion-symmetric or point-symmetric, approximately relative to a center point 46 of the movable element 18. Instead of the sinusoidal shape or cosinusoidal shape, other shapes can also be easily implemented, e.g., a higher number of curvature arcs, other symmetries, or even asymmetrical shapes.

[0076] The center point 46 can be located in a central region 48 in a reference state from which the displaced volumes 28i and 282 of Fig. 1 are substantially equal. From the geometry of the movable element 18 in Fig. 3a, it is clear that a curvature in the x / z plane is reduced or prevented by the additional stiffness of the movable element 18 obtained by means of the curved main sides 18A and 18B when the movable element 18 is deflected along the directions 26i and / or 262.

[0077] Fig. 3b shows a schematic top view of parts of a MEMS 302, which has an alternative configuration of the movable element 18 compared to the MEMS 30i and is also compatible with the MEMS 10 and / or 20. Compared to the movable element 18 of Fig. 3a, the movable element 18 of Fig. 3b has sections of straight areas, for example, instead of the sine minima and maxima. Such straight areas 52i and / or 52s can be obtained while maintaining point symmetry around the center point 46. Straight sections enable a good design of the coupling points to the substrate.

[0078] Just like the movable element 18 of the MEMS 20, the movable elements 18 of the MEMS 30i and 30s can also be suspended in-plane on both sides and symmetrically from the substrate 12, for example by means of a fixed clamping. Alternatively, springs can be arranged at the ends, which have a comparatively low compliance in the direction of the substrate. The symmetry can relate to the magnitude of the forces induced in-plane and / or perpendicular thereto out-of-plane, but can also relate to the position of the fastenings, although this is not necessary, as explained in connection with Fig. 3c and Fig. 3d. While in Fig. 3a and Fig.3b the movable element can be suspended such that, in the reference state from which the displaced fluid volumes are approximately equal, clamping regions or fastening regions 58i and 58s to which the movable element 18 is connected or coupled or fixed to the substrate can be arranged along an axis of the central region 48 arranged perpendicular to the y-direction and perpendicular to the z-direction, an advantageous deviation from this is described with reference to Fig. 3c and Fig. 3d.

[0079] In other words, Fig. 3a shows a micromechanical transducer with point-symmetric fin geometry, in which the arcs of the sine and Euler-Bernoulli curves (bending mode) assume a convex shape. Despite its shape, the deflection of the fin is similar to that of a straight fin for small arc amplitudes. Therefore, this approach is feasible both for air-displacing elements driven via electrodes in the cap and bottom wafers, as well as for air-displacing elements connected to a drive plane via a coupling point. Straight sections along the fins or different arc amplitudes along the fin, as described, for example, in connection with Figs. 5a and 5b, can also be used to optimize the maximum deflection and the bending curve. For a sinusoidal fin, a high number of arcs does not significantly improve the THD. Even a sinusoidal shape shows a significant effect.In connection with other shapes, however, the higher number of arcs can certainly have positive effects. Fig. 3c shows a schematic top view of the MEMS 30i from Fig. 3a for discussing an advantageous embodiment. Although the deflection directions 26i and 262 have been described as being arranged essentially parallel to the y-direction in space, the curvature of the at least one main side 18A and / or 18B in combination with the clamping of the movable element 18 at the clamping regions 58i and 582 can lead to the actual direction of movement 26'i and / or 262 being inclined by an angle Qi or θ2. The angles Qi and θ2 can be individual and different from one another, but can also have the same values. Even if the representation in Fig.3c is strongly exaggerated, the angles 01 and 02 can have a value of greater than 0° to approximately 10° corresponding to 80° to 90°, preferably less to an axis 51 parallel to the chip edge or the course of the substrate 12 and / or the y-direction.

[0080] Fig. 3e shows a schematic top view of a MEMS 304 according to an embodiment, which can be constructed similarly to the MEMS 30i from Fig. 3c, in which the directions 26i and 262, along which the movable element 18 is to be deflected, are arranged at least approximately parallel to the x-axis, wherein here too the actual directions of movement 26'i and 262 can be shifted by the angles αi and α2, respectively, to which the value ranges explained above can be applied. This movement, which is orthogonal to the MEMS 30i, can be generated, for example, by a correspondingly arranged force device in the same or a different MEMS plane, which generates a force parallel to the x-direction.For example, corresponding electrostatic, piezoelectric, or differently designed drives can be rotated by 90° compared to other MEMS described herein, so that a force coupling at the point or region 53 is deflected accordingly and the movable element 18 is deformed accordingly. This can be useful, for example, because in some embodiments of the MEMS 30i and / or 304, a movement of the region 53 due to the curved shape along the x-direction occurs with a significantly greater amplitude than a deflection along the y-direction. By appropriately aligning the actuator, for example an electrostatic comb drive, in the other MEMS plane, large proportions of mechanical transverse loads on the drive plane can be reduced. Fig. 3f shows a schematic plan view of a MEMS 30s that has the features described in connection with Fig.3d, the displacement of the fastening regions 58i and 582 on the substrate 12 has the advantages explained above. Displacements 57i of the fastening region 58i and / or 572 of the fastening region 582 relative to the position in Fig. 3e in the form of a corresponding displacement make it possible to further reduce angles a'i and a'2 relative to the angles ai and 02 in Fig. 3e.

[0081] This means that, depending on the direction of displacement, the angles Qi and O2 have a value between 0° and 10° or between 80° and 90°. A point 53 of maximum deflection of the movable element 18 can move along the directions 26' 1 and 26' 2 based on this, or at least provide corresponding forces based on the mechanical stiffness.

[0082] In an exemplary embodiment of a MEMS, such as that shown in Fig. 3f, the maximum is shifted by approximately 100-200 μm along the beam length or the x-direction for a deflection along the directions 26'i or 262 of 5-10 μm, for example for a sinusoidal shape with an amplitude of the sweep or arc of approximately 4 μm, which can be associated with angles κ and θ2 with a value of greater than 0° to approximately 10°, 5° or less. The approach described is particularly suitable for absolute tilt angles between approximately 80-90° or 0-10°, since otherwise the packing density of the MEMS could be impaired. The angle and the "direction of rotation" can depend heavily on the selected geometry, such as the amplitude and / or the number of sweeps or the like.

[0083] The deflection of the movable element 18 can therefore lead to a lateral offset of the coupling region 53 along the x-direction.

[0084] Thus, in some embodiments, it may be provided to generate a force for deflecting the movable element 18 in a MEMS plane different from the movable element 18, for example, by means of electrostatic drives and / or piezoelectric drives of the actuator device. These forces can be generated by coupling elements that are mechanically connected to the movable element 18, for example at point 53, which then serves as a coupling region. However, alternative configurations are also conceivable for coupling the coupling region to a mechanical device in a different MEMS plane.

[0085] For example, a force generated parallel to the y-direction can then be influenced by mechanical resistances through the tilted directions 26'i and 262, which can lead to inefficiencies and / or mechanical stresses.

[0086] These inefficiencies and / or mechanical stresses can be at least partially remedied by a design of a MEMS according to Fig. 3d. The MEMS 30s shown there has an axis 48' of the central region or the connection between the clamping regions or fastening regions 58i and 582, which is inclined in the opposite direction to the tilts in the directions 26'i and 26'2. The displacement of the fastening regions 26'i and 26'2 relative to one another along the deflection direction 26i and 262 makes it possible to at least partially compensate for a lateral offset of the coupling region 53 of the movable element 18 transverse to the deflection direction and thus to reduce the mechanical loads. The inclination of the axis 48' can approximately correspond to the angles αn and α2 or approximately to an average value of the angles.

[0087] Even if the advantageous embodiment of the inclination of the axis 48 is described by a displacement of the fastening regions 58i and 582 along the y-direction and / or parallel to the chip direction starting from the MEMS 30i, all other movable elements described herein are suitable for such an improvement, although the effect can be more pronounced the fewer arcs the movable element has.

[0088] In other words, Fig. 3d and Fig. 3f show a rotation of the structure, so that the maximum deflection preferably occurs essentially or even always parallel to the chip edge or to the force guide in the other MEMS level, since this is often guided parallel to the chip edge there for reasons of space. This configuration enables easier coupling to a drive level that works parallel to the chip, for example by arranging comb structures. A low transverse load on the coupling point is achieved. The improvement can be achieved both in configurations where the actuator device is designed to move the movable element perpendicular to a direction between suspension regions of the movable element, see Fig. 3d, and in configurations where the actuator device is designed to move the movable element parallel to a direction between suspension regions of the movable element, see Fig. 3f. Fig.4a shows a schematic top view of parts of a MEMS 40i according to an embodiment, which, like the MEMS 40s of FIGS. 4b and 40s of FIG. 4c, can be combined with the MEMS 10 and / or 20. Compared to the MEMS 30i and 30s, the movable element 18 of the MEMS 40i has at least two arcs 54 along each of the directions 26i and 262. While the MEMS 30i and 30p each show one arc, according to other embodiments, at least two, at least three, at least four, or even more, such as eight arcs as shown in FIG. 4a, can be shown. This makes it possible to achieve high compliance along the directions 26i and 262 and, compared to FIG. 3a or FIG. 3b, to utilize a comparatively small extension along the y-direction.

[0089] Fig. 4b shows a schematic top view of a MEMS 402 in which, unlike a possibly continuous transition between sections, as shown in Figs. 3a, 3b, or 4a, a zigzag pattern, such as a discontinuous transition at bends 56, can be implemented. Regions between bends 56i and 562 can be partially straight or curved. However, by means of bends 56i and 562, an overall curved profile of the main sides 18A and 18B is nevertheless obtained.

[0090] While in Fig. 4b main sides of the movable element 18 are formed in sections straight in the unactuated state, which nevertheless provides a curvature, main sides according to other embodiments, for example according to Fig. 4a or Fig. 4c, can be continuously curved in the unactuated state.

[0091] Fig. 4c shows a schematic top view of a MEMS 40s with a fin or movable element 18, which can have a point-symmetric beam eigenmode, which can be achieved by appropriate clamping at clamping regions 58i and 582. For example, the movable element 18 can be arranged straight or perpendicular to the substrate 12, while the MEMS 30i and 302 can enable an inclined force introduction.

[0092] Fig. 5a shows a schematic top view of a MEMS 50i according to an embodiment, the details of which can be readily combined with the explanations for the MEMS 10, 20, 30i, 302, 40i, 402 and / or 40s, which also applies to the MEMS 502. While the arcs of the MEMS 40i, 40s and 40s can have the same arc height or curvature amplitude along the directions 26i and 262, a curvature amplitude 62 of an individual arc 54 of the movable element 18 in the MEMS 50i can be individually adjusted or deviate from other arcs, whereby point symmetry around the center point 46 may, but is not necessarily, observed. For example, arcs 543 and 544 have a curvature amplitude 623 and 624 that is larger than arcs 54i and 546 at edges of the movable element 18. In the embodiment of Fig. 5a, a curvature amplitude 62 decreases continuously from the center point 46 toward the substrate 12.However, neither a monotonic nor a strictly monotonic decrease or increase of a corresponding curvature amplitude is mandatory for embodiments of the present invention.

[0093] For example, Fig. 5b shows a portion of a MEMS 50p, in which the movable element 18 can also have, for example, six arches 54i-54e, although any other number is also possible. Starting from the center point 46 of a point-symmetric configuration of the movable element 18, inner arches 543 and 544 can have a comparatively small or minimal curvature amplitude 623 and 624, respectively. Just as in the MEMS 50i, the arches can have different sizes. In the MEMS 502, small or minimal sizes or curvature amplitudes n are arranged centrally, and a curvature amplitude 62 can increase toward outer arches 542 and 54i on the one hand, and 545 and 546 on the other. It is possible but not necessary that a curvature amplitude 62i is larger or smaller than a curvature amplitude 622. The same applies to the curvature amplitude n 625 and 62ß.

[0094] Both embodiments of MEMS 50i and 503 provide a MEMS in which a path of the movable element 18 parallel to the substrate plane, starting from the center point 46 to one end of the substrate 12, has at least two arcs of curvature that have a mutually different curvature amplitude 62 along a deflection direction 26i, 262 of the movable element. For example, a path of the movable element 18 of the MEMS 50i and / or 502 parallel to the substrate plane, starting from a center point, such as the center point 46, to one end on the substrate 12, can have at least two arcs of curvature 54 that have a mutually different curvature amplitude 62 along a deflection direction y of the movable element. Embodiments of the present invention relate to micromechanical systems, MEMS, that can be used as sound-generating devices. The achievable sound pressure level is an important parameter.In the low-frequency range, the goal is to achieve a very high frequency response to enable active noise cancellation. For an electrostatic loudspeaker in a multi-layer system [4], a bulge in the center can occur at higher frequencies with straight air-displacing elements. This reduces the maximum achievable sound pressure level. Furthermore, the resulting harmonic distortion (THD) is a quality characteristic for the reproduction of broadband signals. This is caused by nonlinearities in the system. These nonlinearities can be either part of the driving force or part of the system stiffness or damping.

[0095] A linear driving force can be realized, for example, by a multi-layer system as proposed in [4, 5].

[0096] The nonlinearity resulting from the mechanical stiffness of the air-displacing plane is caused by its expansion stiffening. Reducing this stiffening can reduce harmonic distortion and reduce or prevent instabilities in the frequency response caused by an unfavorable interplay of force, damping, and expansion stiffening. Furthermore, expansion stiffening reduces the achievable maximum deflection of the elements in the air-displacing plane, the plane of the moving elements 18, and thus the maximum sound pressure level.

[0097] The MEMS components described herein are, for example, layer stacks consisting of at least one substrate layer in which the electrodes and the passive elements are arranged. Further layers include a base, which can also be referred to as a handling wafer, and a cover, which can also be referred to as a lid wafer. Both the lid wafer and the handling wafer can be connected to the substrate plane using material-to-material processes, preferably bonding, which can create acoustically sealed gaps in the component. The deformable components, the movable element 18, can be located in this gap in the device plane. In other words, the deformation occurs in-plane.

[0098] The layers can, for example, comprise electrically conductive materials, such as doped semiconductor materials and / or metal materials. The layer-by-layer arrangement of electrically conductive layers can enable a simple design, since electrodes (for deflectable elements) and passive elements can be formed by selectively dissolving them from the layer. If electrically non-conductive or insulating materials are to be arranged, this can be achieved by applying these materials layer by layer using a deposition process.

[0099] Embodiments of the present invention allow the stiffness of the air-displacing element to be adjusted and selected such that the highest possible sound pressure level can be generated while maintaining a low THD. For this purpose, fins with a non-straight neutral fiber or at least one curved surface are used. The inventive implementation of a curved main surface of the MEMS is synonymous herein with a non-straight, i.e., continuously or discontinuously curved, bent, or kinked neutral fiber.

[0100] This can reduce the strain stiffening of the moving element and thus the total electric field strength (THD). This is possible for fins / moving elements clamped on both sides using non-straight fins, e.g., fins with a curved neutral fiber. If the fin is inversion-symmetric in the x / y plane, the displaced volume can be the same during forward and backward movement. A symmetric or point-symmetric profile of the electrodes and / or a point-symmetric arrangement of dielectric layers in the gap can enable a high packing density of moving elements.

[0101] Through this implementation, the effects of strain stiffening can be reduced. In particular, instabilities occurring at a certain resonant frequency can be reduced, especially when the damping is insufficient for the applied force. Furthermore, the reduced stiffness can result in an increase in sound pressure level (SPL). Furthermore, a reduction in strain stiffening can also reduce the THD. By means of embodiments of the present invention, the even harmonics, i.e., second, fourth, sixth, and higher orders, in the THD can be reduced or prevented, and buckling can be prevented. Instead, a displaced volume that is equal in the forward and backward directions can be created. This can be achieved, for example, using the point symmetry described herein. The air gap or trench can remain constant along the fin, the movable element, to facilitate etching.A constant fin thickness can be specified. Furthermore, the fin itself can be used to reduce strain stiffening. This eliminates the need for additional space on the chip.

[0102] Second, increasing the fin's stiffness in the x / z plane can increase the displaced air volume. A straight fin without bending can deform at high frequencies of high damping in the system, which would result in less air or fluid being pumped out of the system. This means that part of the system's energy is invested in deforming the fin rather than in displacing the air (SPL). One possible solution here would be to make the fins thicker. This would increase the fin's mass and stiffness in the y-direction, which in turn would increase the distortions from strain stiffening, which is avoided by the embodiments described herein.

[0103] This will be further discussed in the following sections. Figures 6a and 6b show a schematic perspective view of a planar, undeformed fin 64 parallel to the x / z plane. Figure 6b shows a fin 64' that is deformed in the x / z plane and has a curvature. The curvature of Figure 6b leads to increased stability of the fin 46'.

[0104] Fig. 7a shows a multiply corrugated fin arranged in edge regions on a solid substrate 66. Fig. 7b shows a corresponding positioning of the fin 64. In response to a force 68, for example, a weight force or an actuating force, the plate without corrugations in Fig. 7b bends much more easily than the plate 64" with corrugations, which bends more difficultly. The folds in the fin 64" thus contribute to the fin deforming less for the same thickness.

[0105] This finding is transferred to embodiments according to the invention and the MEMS obtained thereby. Although the folds reduce the strain stiffening and thus the THD and increase the stiffness in the x / y plane and thus the SPL, embodiments enable the fins / movable elements with folds or arcs to reach the vertical pull-in earlier, i.e., at a lower drive voltage, than the unfolded fins. One reason for this is that the folds reduce the stiffness in the z-direction and increase the effective length of the fins. Both effects can reduce the vertical pull-in voltage. This effect can be addressed by using a larger gap between the top electrodes / bottom electrodes on the one hand and the fin on the other. Alternatively or additionally, a small number of arcs or folds with a low amplitude relative to the neutral fiber of the fin can be used.

[0106] Fig. 8a shows a schematic perspective view of a portion of a MEMS 80 according to one embodiment. The movable elements 18i-18s are shown only in part. Furthermore, a cover layer 68 of the MEMS is shown, which may, for example, include the stator electrode 38i or serve as a carrier therefor.

[0107] As an alternative to the representation of the five movable elements I 81-I 85, the MEMS 80 may also have a different number of at least one movable element, for example a number of one, two, three, four, six or more.

[0108] Fig. 8b shows a schematic plan view of a part of a MEMS 80', in which four movable elements 181-184 are arranged as an example.

[0109] Like other movable elements described herein, one or more of the movable elements I 81-I 84 may be substantially uncurved in an unactuated state along a direction perpendicular to the substrate plane, ie, along the z-direction.

[0110] The MEMS 80 also has, as an example for other MEMS described herein, a movable element in which a path of the movable element 18 parallel to the substrate plane from a first end to a second end, for example, between the two clamping points, has at least two sign changes of a derivative of a curvature path. This can refer to a fin with at least two arcs, to the zigzag pattern shown in Fig. 8b, to a sinusoidal path, and / or to a cosinusoidal path.

[0111] Fig. 9a shows a schematic top view of a part of a MEMS 90 according to an embodiment. According to the embodiment, the actuator device has a stator electrode arrangement. Stator electrodes 38i and 382 and optionally additional stator electrodes can be arranged on the substrate of the MEMS or form parts of the substrate. The actuator device can be designed to deflect the movable element 18 along the direction 26i based on an electrostatic force between the movable element 18 and the stator electrode 38i. Furthermore, the actuator device according to the embodiment is designed to deflect the movable element in the direction 262 based on an electrostatic force between the movable element 18 and the stator electrode 382.For this purpose, for example, the control device 32 of the MEMS 10 can be configured to apply corresponding control signals to the electrodes 38i, 382 and an electrode of the movable element 18. Fig. 9 can show an exemplary undeflected or force-free state or, alternatively, a center position of the movable element 18 centered by a pre-deflection, which can serve as a starting point for the same volume size of displaced volumes. In this state, the movable element 18 can be arranged centrally opposite the stator electrodes 38i and 382. This enables a homogeneous and symmetrical generation of forces.

[0112] According to one embodiment, the MEMS 90 is configured to arrange the movable element 18 opposite the first stator electrode 38i and the second stator electrode 382 in every operating position. This means that even when the movable element 18 is deflected, both stator electrodes 38i and 382 are opposite the movable element 18, at least in sections or in a partial area of ​​the movable element 18. This can be achieved, for example, by a corresponding arrangement of the stator electrodes 38i and 382 adjacent to the substrate 12, so that when the movable element 18 is highly deflected in a central area, the movable element 18 is opposite only one of the electrodes 38i and 382, ​​but at least in the edge areas at the clamping location, both stator electrodes 38i and 382 are always opposite each other in every intended operating position.This enables a homogeneous and continuous generation of electrostatic forces.

[0113] Fig. 9a shows a further embodiment that can be implemented independently, in which an electrode geometry of at least one of the stator electrodes 38i and 382 of the actuator device has a curvature that is adapted to a curvature of the movable element in the actuated or unactuated state. This enables further homogenization of the generated forces and can be particularly advantageous if an identically shaped movable element is arranged adjacent to the movable element 18 along the positive and / or negative y-direction, which is to be deflected together with the movable element 18. The electrode geometry is to be understood in connection with the curvature in such a way that parallel to the substrate plane and / or parallel to the x / y plane or in the plan view shown, an edge 76i, 76s, 76s and / or 764 of one or more stator electrodes 38i and 38s is adapted to a course of the main sides 18A and / or 18B.

[0114] The stator electrodes 38i and 38s may be spaced apart by a gap 78 to prevent electrical short circuiting.

[0115] The gap 78 may, but does not necessarily, be arranged opposite the neutral fiber of the movable element 18.

[0116] The embodiment of Fig. 9a can also be understood as a MEMS with curved drive electrodes, wherein the curved shape can be adapted to the shape and / or geometry of the movable element and can also be adapted to other shapes of the movable elements according to the embodiments described herein.

[0117] Fig. 9b shows a schematic perspective view of the MEMS 90 from Fig. 9a, with auxiliary lines HLi to HI-4 being drawn to illustrate the curvature in a plane parallel to the x / y direction.

[0118] Fig. 9c shows a schematic perspective front view of the MEMS 90 from Fig. 9a and Fig. 9b.

[0119] Fig. 10 shows a schematic perspective view of at least part of a MEMS 100 according to an exemplary embodiment. Optionally but not necessarily, the curved shape of the stator electrodes, i.e., the shape adapted to the shape of the movable element, is shown therein; this configuration is optional. The cavity 14 can be delimited entirely or partially along the positive and negative z-direction by a cover 681 and / or a base 682, for example, comprising substrate material and / or semiconductor material. At least one stator electrode arrangement, for example, the stator electrodes 38i,1 and 382,1 on the one hand and / or 38i,2 and 382,2, can be arranged thereon. However, generating electrostatic forces on the possibly conductive movable element 18 along both stator electrode arrangements 381.1 / 382.1 and 381.2 / 382.2 allows for controlling or preventing torsion of the movable element 18 about the x-axis.Also shown are fluid passages or openings 82i and 82s in the lid 681 and the base 682, respectively, to enable an exchange of fluid located in the cavity 14 with the environment. This means that fluid can flow into and out of the cavity 14 through the opening 821, and fluid can also flow into and out of the cavity 14 through the opening 822. Preferably, the openings 82i and 822 are arranged on opposite sides of the MEMS 100 along the z-direction, for example, to prevent an acoustic short circuit when operating as a loudspeaker. By means of the movable element 18, the cavity 14 can be divided into partial cavities 84i and 84p, wherein a fluidic or acoustic short circuit between the partial cavities 84i and 84p can be prevented by the comparatively small gaps between the movable element 18 and the cover 681 on the one hand and the base 682 on the other hand.

[0120] According to the illustrated embodiment, a MEMS is provided that has at least one opening 82i and / or 822 in the substrate, which opening is arranged along a direction parallel to a surface normal arranged perpendicular to the substrate plane x / y, i.e., along the z-direction, and which has an opening geometry that is adapted to a curvature of the movable element 18 in the actuated state or the unactuated state. In particular, when using a plurality of movable elements arranged adjacent to one another approximately along the y-direction, the cover 681 and / or base 682 can also have a plurality of openings that are adapted individually, in groups, or each individually to a shape of the movable elements.

[0121] Fig. 11 a shows a schematic top view of possible courses or geometries of movable elements 18, 18' and 18" of a MEMS 1 10i. While a course of the movable element 18 can be substantially consistent with the discussions on the MEMS 90 and / or 100, which can be readily combined with other MEMS described herein, a course 18' shows an exemplary triangular geometry, which is implemented, for example, in Fig. 4b. As another limiting case, a course 18" can be considered a rectangle compared to the triangle of the course 18', which, however, can also have a point symmetry around the respective common center point 46.

[0122] All configurations of the movable elements 18, 18', and 18" can be understood as curved fins or fins that deviate from a straight shape, for example, with a convex and / or concave fin shape. Fig. 11b shows a schematic plan view of at least part of a MEMS H O2, in which the movable element 18 can have an overhang when oriented along the x-direction.

[0123] A transition is understood to mean that the movable element 18 overlaps itself along the positive or negative y-direction, for example, starting from the center point 46. A point geometry is retained in the embodiment of Fig. 11b, as are other advantages described herein. This means that if the movable element 18 extends parallel to the substrate plane x / y from a first end to a second end, the movable element 18 can have an overhang.

[0124] Fig. 12a shows a schematic top view of parts of a MEMS 120 according to an embodiment, which comprises a plurality of adjacently arranged movable elements 181 and 182, possibly with additional movable elements. These are arranged with a main curvature oriented opposite to one another, such that the curvatures face each other.

[0125] Fig. 12b shows a schematic perspective view of the MEMS 120. The MEMS 120 has a plurality of adjacently arranged movable elements that have mutually facing main curvatures.

[0126] Fig. 13a shows a schematic plan view of at least parts of a MEMS 130, in which a plurality of at least two movable elements 181 and 182 arranged adjacent to one another are also provided, in which oppositely oriented main curvatures are also implemented, in such a way that the main curvatures face away from one another.

[0127] Fig. 13b shows a schematic perspective view of the MEMS 130.

[0128] Compared to MEMS 120, MEMS 130 exhibits reduced fluidic damping due to the opposing principal curvatures, while conversely, MEMS 120 exhibits comparatively high fluidic damping for the fluid moving in the cavity. In other words, Figures 12a and 12b show an increase in damping at the same clamping distance compared to the embodiments of Figures 13a and 13b, which show a reduction in damping at the same clamping distance.

[0129] Fig. 14 shows a schematic plan view of at least parts of a MEMS 140 in which a plurality of at least two, three, four, or more movable elements are arranged adjacent to one another. The movable elements I 81-I 84 can each have an equal distance d o from neighboring elements, although the distances can also be different from one another. However, equal distances enable a homogeneous generation of sound pressures.

[0130] Fig. 15 shows a schematic plan view of at least parts of a MEMS 150, in which a plurality of at least two movable elements 181-18e are arranged on the substrate 12. These have a mixture of the configuration of the MEMS 120 and 130. For example, the movable elements 181 and 182 are arranged with main curvatures facing away from each other according to the MEMS 130. A similar pair of movable elements is formed by the movable elements 183 and 184 and 185 and 18e. These pairs can form a configuration according to the MEMS 120 when adjacent to each other on the outer movable elements 182 and 183 and / or 184 and 185. This means that regions with high and low fluidic damping can be created. The distances between the movable elements can be I 81 / I 82 ; 183 / 184 ; 185 / 18e with a distance dp different from a distance between the pairs di, wherein preferably the distance di is greater than dp.The distances can, for example, have a value of at least 0.1 pm and at most 1 mm, at least 0.5 pm and at most 500 pm and preferably of at least 1 pm and at most 100 pm.

[0131] This means that the adjacent movable elements 181-18e can be arranged with an alternating distance.

[0132] Figs. 12a-15 show a further aspect of the embodiments described herein that can be implemented independently. The movable elements of the MEMS 120, 130, 140, and 150 can be mirror-symmetrical with respect to an axis of symmetry 86, which extends, for example, in the z / y plane or parallel thereto. The number of arcs can be one or greater than one. The axis of symmetry 86 is arranged, for example, perpendicular to the main extension direction x of the movable element. The MEMS 140 and 150 thus exhibit a periodic repetition of mirror-symmetric fins in a MEMS.

[0133] With a periodic repetition of exemplary mirror-symmetric fins, the fin spacing can either be the same, as shown in Fig. 14, or can be periodically varied, for example, when the structure is alternately mirrored, as shown in Fig. 15. In the latter case, the asymmetry in the damping can be compensated. If the spacing of the fins in their central region is large, the fin spacing is selected to be smaller. If the spacing in the fin center is small, the fin spacing is selected to be larger. Aside from mirror symmetry, other forms of design deviating from point symmetry can also be selected for the movable elements.In the event that a curvature of the main side deviates from a point symmetry, for example due to deliberate shaping of the movable elements and / or due to manufacturing tolerances, a control device of the MEMS, such as the control device 32, can be designed to set a pre-deflection, bias, of the movable element adapted to the curvature in order to equalize the first fluid volume and the second fluid volume to one another by means of the pre-deflection.

[0134] Alternatively, different control amplitudes can be applied, for example to compensate for different degrees of compliance or mechanical stiffness. The drive in the direction of the lower or higher displaced volume can be increased or decreased until the displaced volumes in the forward and backward directions are at least essentially the same. One implementation example is detuning the DC voltage, as can be used, for example, for mirror-symmetric fins. This makes the approach robust for production even in the event of deviations from the target symmetry. The cover electrodes follow this fin shape in exemplary embodiments, whereby the outlet support can also be adapted to the shape. A variation of the initial overlap between fin and cover electrode is possible, so that in places with little deflection the overlap can be large without overtravel occurring.Such a configuration is shown, for example, in Fig. 10.

[0135] Although mirror-symmetric fins can produce even harmonic distortions because their compliance differs for forward and backward deflection, for example, along the positive and negative y-direction, such shapes can be used in exemplary embodiments. The mirror-symmetric shape can lead to different deflection amplitudes in the two directions. In reduced-order models, this can result in an effective quadratic elastic force for force ranges where buckling does not yet occur. Pre-deflection using the selected drive voltage can be used to minimize the even harmonics. For the cover drive, the DC components are detuned symmetrically around the average applied DC voltage, for example, this can be achieved using the parameter δ, so that with reference to the voltages Ui and Us in Fig. 9a, the following can be obtained:

[0136] Ui = UDC (1 + Ö) + UAC

[0137] U2= UDC (1 -Ö)-UAC

[0138] This results in a DC component for the resulting force. For reduced-order models, the DC component in the force is equivalent to a scaled elastic force without an AC component. By choosing the DC component, the quadratic spring effect can be minimized.

[0139] The combination of a mirror-symmetric fin with pre-deflection results in a lower effective linear stiffness while maintaining the same cubic stiffness. This can further increase the maximum achievable deflection at the same signal amplitude.

[0140] Fig. 16 shows a schematic flow diagram of a method 1600 according to an embodiment. A step 1610 includes providing a substrate having a cavity in the substrate.

[0141] A step 1620 comprises arranging a fluid displacement device with an element suspended in the cavity and arranged parallel to a substrate plane in the cavity and movable in-plane, which element is designed to interact with a fluid in the cavity.

[0142] A step 1630 comprises providing an actuator device configured to deflect main sides of the movable element alternately along a first direction to displace a first fluid volume and along an opposite second direction to displace a second fluid volume in-plane. The method is carried out such that at least one of the main sides of the movable element is curved in an unactuated state, and such that the displaced first fluid volume and the displaced second fluid volume have substantially the same volume size.

[0143] The fluidic damping between the fins can be optimized for a mirror-symmetric shape. The contribution of this damping is smaller for larger distances. Regions in which the fin deflects further contribute more to the damping. For the fin clamped on both sides, this is the central region for its first bending mode. A mirror-symmetric shape, such as a cosine, increases the fin spacing in the center and thus reduces the damping. An arrangement that reduces the gap in the center increases the fluidic damping, as described in connection with Figs. 12a, 12b, 13a, and 13b. Embodiments according to the second aspect address an adaptation or adjustment of the fluidic damping to optimize MEMS.

[0144] The previously described embodiments relate to the design of displaced fluid volumes such that they have essentially the same volume size. Such designs demonstrate particularly high effectiveness in the low frequency range, for example at frequencies up to 100 Hz, 200 Hz, or even 600 Hz. Although the advantages achieved thereby are unrestricted in higher frequency ranges, further effects occur at higher frequencies, for example in the adjacent mid-frequency range above 100 Hz, 200 Hz, or 600 Hz or beyond, which can influence the sound quality. It should be noted that transitions between the lower frequency range and the mid-frequency range can be fluid, depending on the application. At mid-frequencies above 100 Hz, 200 Hz, or 600 Hz, additionally or alternatively, damping effects can become more noticeable.According to a further aspect of the present invention, the same or identical advantages for the design of the displaced volumes at higher frequencies are also obtained by designing the fluidic damping acting in a respective partial cavity on a first or second main side of the movable element in a substantially identical manner.

[0145] Fig. 17 shows an exemplary schematic graph of a sound pressure level curve that can be generated in exemplary embodiments over the frequency to illustrate the low frequency range Afu and the mid-frequency range Afw, which border each other at approximately 600 Hz. The specified frequency is to be understood merely as an example and does not limit the exemplary embodiments discussed herein. For example, the resonance frequency f res present in a range between 3 kHz and 5 kHz and the first peak in the THD is due to the dynamics at f res / 3, i.e., starting at approximately 1 kHz. These values ​​may vary for different systems. Depending on the degree of attenuation, the peak may sooner or later become a significant contributor to the THD. With very low attenuation, the attenuation may also play a subordinate role across all frequencies.

[0146] Embodiments of the second aspect described below can be combined without restrictions with the embodiments of the first aspect.

[0147] Fig. 18 shows a schematic top view of a MEMS 180 according to an embodiment of the second aspect. The MEMS 180 comprises the substrate 12, in which the movable element 18 is arranged, which is shown deflected relative to a reference position or reference deflection 92 of the movable element 18. This deflection shown can also be considered an unactuated state, which can be obtained due to a pre-deflection or a bias relative to an undeflected state, as described, for example, for asymmetric fins.

[0148] At this point, it should be noted that embodiments according to the second aspect do not exclude the possibility that one or both main sides of the movable element are curved in an undeflected state, but that this feature is not necessarily important when adjusting the damping and the statements can also apply to non-curved or straight movable elements, at least in the undeflected state, which is why the reference state 92, for example, can also be assumed by the movable element 18 at certain times.

[0149] In accordance with the first aspect, the MEMS 180 comprises a fluid displacement device comprising the movable element 18. Furthermore, an actuator device (not shown) is provided, which is configured to deflect main sides 18A and 18B of the movable element alternately along the directions 26i and 26b in-plane to displace a respective fluid volume. Although in the MEMS 180 both main sides 18A and 18B are curved in the unactuated state, it is sufficient that only one of the main sides 18A or 18B is curved, as described in connection with Fig. 1. The illustrated state of the movable element can be implemented as a cosine fin with DC bias, so that a respective individual or matching distance d o is then implemented once between blue and the respective wall of the partial cavities 84i and 84p, respectively.The displaced first fluid volume is displaced from a first partial cavity 84i on the main side 18A of the movable element, and the second fluid volume is displaced from a second partial cavity 842 on a main side 18B of the movable element 18 opposite the main side 18A. The partial cavities 84i and 842 have or provide essentially the same Fuldian damping for the deflection of the movable element. The essentially equal damping can be equal within a tolerance range of at most 20%, at most 10%, or at most 5%, preferably less, for example, taking into account the mathematical relationships discussed herein. While this can be accomplished in a trivial case of a straight movable element (see reference state 92) by congruent partial cavities 84i and 842, the solutions for such a task are more complex for movable elements with at least one curved side.

[0150] It is possible but not necessary that, according to embodiments, the partial cavities

[0151] 841 and 842 are geometrically congruent, meaning that according to embodiments, the partial cavities 84i and 842 are incongruent. Adjustment of the fluidic damping can be achieved by designing the partial cavities 84i and 842, for example, in a passive manner, but actively deflectable elements can also be provided. According to one embodiment, an active element is provided in the MEMS, which is designed to adjust the fluidic damping of at least one of the two partial cavities 84i and 842, for example, via the distance between the movable element and that of the partial cavity 84i and

[0152] 842opposite cavity wall.

[0153] Both a passive attitude and an active attitude can be based on the recognition that: where w is the curvature profile; t, a dimensionless quantity of the relative position x along the total length l of the movable element x / l, T is the dimensionless time related to the period at the mechanical resonance frequency; a is a real scaling parameter, and i()0 describes the deflection profile of the beam for the respective load situation. For an electrostatically driven beam, this can be the first bending mode of the beam. For fluidic damping, when considered as squeeze-film damping, the following applies: where F SQ the squeeze film damping; <5[ describes the dimensionless distance between the main side of the movable element and the passive and / or active wall of the partial cavity adjacent to this main side

[0154] For consideration as squeeze-film damping, the distance between the movable element and an adjacent cavity wall can be at least 0.1 pm and at most 1 mm, at least 0.5 pm and at most 500 pm, and preferably at least 1 pm and at most 100 pm. The distance depends heavily on the geometry (e.g., the depth of the movable element or the aspect ratio) as well as the fluidic boundary conditions (size of the outlet / inlet openings in the substrate or the so-called clearance, i.e., leaks between the movable element and the substrate).

[0155] From this context, the following term can be highlighted: where J describes the integral over the effective spatial differences in damping along the fin, which significantly influences the squeeze film damping. <5[^] describes the dimensionless distance between the main side of the movable element and the passive and / or active wall of the partial cavity that borders this main side; a is the time-dependent scaling parameter of the spatial beam bending and >o [fl describes the spatial bending profile of the actuator during actuation. This can, but does not have to, be the first bending mode. For example, a different bending shape can be assumed as the basic bending if several coupling points are arranged along the fin. The following can apply for a distance do(x) between the reference state 92 and a cavity wall: d = d0[x] — w [x, t] where d is the distance to the boundary structure in the deflected state and do = the distance in the undeflected state at the operating point.This may include a pre-deflection by a bias; x denotes the position along the fin; w denotes the deflection of the fin by a time-dependent actuation at time t.

[0156] The squeeze film damping considered according to embodiments can effectively, or at least largely, depend on the aforementioned term J. The relationship shown takes into account the nonlinearities of the squeeze film damping in a lumped parameter model (LPM). In the embodiment of the present invention, the fact that the shape of the stator provides a degree of freedom can be used to influence the term. The numerator is weighted by the square of the beam's bending profile t0[f|]. The term also shows that a higher damping effect can be achieved where the gap between the movable element and the cavity edge becomes smaller.

[0157] Fig. 19 shows a schematic top view of a MEMS 190 according to an embodiment. The MEMS 190 is embodied, for example, as an ultrasonic transducer, although this is not necessary in connection with embodiments described herein. The movable element 18 is designed, for example, to be connected to a reference potential GND. Arranged opposite a main side 18A is a stator electrode 38 which can be connected to an electrical potential U2 in order to provide a potential difference between the movable element 18 and the stator electrode 38. Unlike described, for example, in connection with the MEMS 20, 90 or 100, the stator electrode 38 can be arranged in the same MEMS plane as the movable element 18, ie in-plane opposite one another.A volume displaced from the partial cavity 84i can escape from the partial cavity 84i along the positive and / or negative z-direction, for example, through an opening not shown. Alternatively or additionally, the stator electrode 38 can also have an opening to allow fluid flow. Similarly, the partial cavity 842 can be connected to an external environment, wherein, for example, only the volume displaced from the partial cavity 84i contributes to the operation of the MEMS 190.

[0158] The partial cavities 84i and 842 are geometrically incongruent, whereby a contour or shape and / or size of the partial cavity 842 can be adapted to the operation and properties of the partial cavity 84i by means of a boundary structure 94. The boundary structure 94 can be an electrically active or passive element, for example, made of substrate material or additionally inserted. Optionally, the boundary structure 94 can be connectable to an electrical potential, such as the reference potential GND, which makes it possible to reduce or prevent the generation of electrostatic forces between the boundary structure 94 and the movable element 18, given the same electrical potential between the boundary structure 94 on the one hand and the movable element 18 on the other.The shape and geometry of the boundary structure 94 thus allows for efficient adjustment of the fluidic damping, for example in accordance with the relationship discussed above. It is possible, but not necessary, to provide a boundary structure 94. At least some of the sidewall structures 12A and / or 12B of the partial cavities 84i and / or 842 can also be adapted.

[0159] However, according to embodiments, a size and / or a course of limiting structures 94 can be adapted to a non-linear stiffness of the movable element 18 and / or a clamping of the movable element.

[0160] A shape or profile of the movable elements 18 according to the second aspect is cosine-shaped merely by way of example but not by way of limitation. All of the shapes of the movable elements described in connection with the first aspect are also possible.

[0161] Fig. 20 shows a schematic top view of at least parts of a MEMS 200, in which the boundary structure 94, unlike in the MEMS 190, has a planar side 94A facing the movable element 18. This nevertheless enables a precise design of the partial cavity 842. In other words, stators or static elements 94 and / or 38 can be continuous or structured. The potential U2 at the stator electrode 38 can be used to generate a signal from the MEMS 190 and / or 200, and an applied direct current (DC) voltage can be used to generate a pre-deflection (bias). The exemplary embodiments relate to a compensation of the fluidic nonlinearity by means of a geometry in the partial cavity 842. It can be taken into account that at positions along the x-direction where a gap change is comparatively large, a higher contribution to damping is made.Compensation for fluidic nonlinearity can alternatively or additionally be achieved via a distance between the movable element 18 and the boundary structure 94. This means that the two cavities do not necessarily have to be geometrically identical to compensate for fluidic nonlinearities in the small signal range. In both MEMS 190 and MEMS 200, passive cavity adjustment is achieved via a distance and geometry of boundary structures.

[0162] At this point, attention is drawn to the difference between the partial cavity itself and the displaced fluid volume. Even with unequal partial cavities, the fluid volume displaced during movements along opposite directions can be designed identically, which allows embodiments according to the first aspect to be implemented without restrictions, taking into account the findings of the second aspect, in order to obtain an advantageous design across a wide frequency range.

[0163] In other words, Figs. 19 and 20 show a passive cavity adjustment via a distance and a geometry of a confinement structure.

[0164] Fig. 21 shows a schematic top view of at least parts of a MEMS 210 according to an embodiment of the second aspect, which may include an active element 96 configured to adjust the fluidic damping of at least the partial cavity 842. An active or passive element may be arranged as the boundary structure.

[0165] This can be achieved, for example, by applying an electrical potential U1 between the active element 96 and a stator electrode 382 arranged on a base layer or cover layer, which can generate an electrostatic force for deflecting the active element 96. Other actuator mechanisms are also readily possible, for example, magnetic actuation, mechanical actuation by means of an actuator, piezoelectric actuation, or the like. By deflecting the active element 96, a local distance between the active element 96 and the movable element 18 can be adjusted without affecting the partial cavity 84i in order to adjust the fluidic damping of the partial cavity 84s. The stators 38i and 38s can be structured individually or jointly, or formed continuously.While the potential U2 between the movable element 18 and the stator electrode 38i can be used to generate a signal, both in the sense of a loudspeaker and in the sense of a microphone, the movable element 18 and the active element 96 can possibly, but not necessarily, be subjected to the same potential, such as the reference potential GND, in order to avoid an electrostatic force between the two elements. By applying an electrical potential between the stator electrode 38s and the active element 96, a deflection of the active element 96 can occur, which allows adjustment of the cavity 842 acting on the movable element 18 in order to adjust the fluidic damping. Preferably, the deflection is caused by a static voltage that can be calibrated at a specific time, such as at the beginning of an operating interval, via the frequency response of the system.

[0166] It is also possible for the active element 96 to be set into some kind of oscillation in order to adapt the cavity 842 for large-signal behavior. This can be particularly important at low frequencies with regard to resonance, including large deflections of the movable element 18. According to exemplary embodiments, however, the size of the partial cavity 842 is set quasi-statically compared to a dynamic operation of the movable element set for sound generation. While, for example, an acoustic sound or even an ultrasound can be generated with the movable element 18, exemplary frequencies for adapting the partial cavity 842 are arranged, for example, at frequencies that are lower than the target range. In a loudspeaker configuration, these can be frequencies below 1 kHz, for example, while in an application in the ultrasound range, frequencies below 10 kHz are easily possible.

[0167] The size of the partial cavity can be adjusted once or multiple times. A control device and / or a regulation device can be used for the adjustment. A one-time active adjustment can be achieved, for example, by pre-deflecting the movable element into an at least temporarily unchangeable reference state, for example to compensate for manufacturing tolerances. This can represent a preferred embodiment. For active control, for example with a closed control loop, a closed feedback loop, a sensor such as a microphone can be used to record the sound pressure signal in the time domain. A spectrum can be derived from this, for example using a fast Fourier transform (FFT). The spectrum can then be analyzed to determine whether the presence or generation of at least one harmonic is detected.Alternatively or additionally, the impedance of the converter can be recorded and monitored, which enables the determination of the harmonics in the capacitance measurement.

[0168] With the MEMS 210, it is also optionally possible for the partial cavity 84s to be acoustically short-circuited with another partial cavity 843 between the stator electrode 38s and the active element 96 used as a boundary structure. This changes the boundary condition for the design of the fluidic damping and also the fluidic damping itself. However, the fluidic damping in the partial cavity 842 can still be adjusted via the geometry of the cavity. By deflecting the actively adjustable element 96, the fluidic damping between it and the movable element 18 can be adjusted. This can include adjusting the distance between the element 96 and the movable oscillating element 18. The distance between the active element 96 and the movable element 18 can be adjusted by statically or dynamically deflecting the active element 96, which can, for example, be pre-deflected once.This changes or adjusts the fluidic damping for the movable element. The geometry to be adjusted can thus be viewed as the fluid volume adjacent to the movable element. Active adjustability by means of a movable element 96, in contrast to a possibly non-deformable element 94, enables adaptation to different operating points, which can change, for example, at different temperatures or fluid densities, so that the correct operating point can always be found. The invention recognized that distortions can also depend on the operating point. For example, the sensitivity of a microphone depends on the operating point, particularly in the example of an ultrasound application. The sensitivity to even harmonic distortions also depends on the operating point.Measurements regarding the movement of the movable element are easily possible by observing the deflection behavior or by capacitance or inductance measurements to establish a control loop. Alternatively or additionally, embodiments also provide for the use of a sound pressure signal in the loudspeaker / ultrasonic transducer or a voltage signal in the microphone as the measured variable.

[0169] In other words, Fig. 21 shows an active cavity adjustment by means of a further fin or active element or movable element, which can also be designed as a further movable element 18, for example.

[0170] Fig. 22 again shows the representation of the MEMS 150 from Fig. 15 for discussing an advantageous embodiment according to the second aspect. The plurality of movable elements 181-18e are arranged adjacent to one another in the cavity, wherein adjacent movable elements I 81 / I 82; I 82 / I 83; I 83 / I 84; ... are arranged alternately with a first distance di and a second distance d2, and in each case a main side of a movable element has the distance di from a first adjacent movable element, for example the movable element 183 along the negative y-direction to the movable element 182, and an opposite main side has the first distance di from another movable element, for example to the movable element 184. In this case, for example, the active element 96 of the MEMS 210 can be formed as the movable element 18, which can also be easily implemented in the MEMS 210.

[0171] A larger distance di is preferably arranged on that side of the movable element along which a preferred direction of deflection of the movable element is arranged, which can result from the geometry and / or mechanical rigidity of the movable element 18 in the case of an implementation of the movable element deviating from a point-symmetric design, for example in the case of a cosine configuration.

[0172] The embodiments described herein for the second aspect are a further example of the fact that the geometry of the cavity is not limited to a specific shape. According to the invention, it was recognized that smaller gaps can increase damping compared to larger gaps. Examples of passive damping adjustment via the geometry of the partial cavities are shown in Fig. 19 and Fig. 20. Embodiments according to the second aspect solve the problem of nonlinearities from the fluidic damping, which generate harmonic distortions. This force depends on the velocity and therefore cannot be easily compensated for via elastic nonlinearities or electrostatic nonlinearities.This problem is solved according to the invention in that a combination of two cavities is provided, one on each side of the fin / movable element, which effectively have the same damping effect in order to minimize distortion. For this purpose, embodiments provide for the effective damping force to be adjusted in both deflection directions of the movable element, whereby the geometry of the respective partial cavity does not have to be the same for this purpose. Active elements can enable adjustment after completion. Embodiments according to the second aspect enable linear transmission behavior even at higher frequencies, which can be particularly relevant for loudspeakers and / or microphones. The geometry of the cavities can, but does not have to, be the same, which can result in additional design freedom. Active elements or active fins enable subsequent adjustment of the damping.In contrast, passive elements, such as the limiting structure 94, enable more sophisticated models for accurate prediction, but require less space and may not require a dedicated voltage signal.

[0173] Fig. 23 shows a schematic flow diagram of a method 2300 for producing a MEMS according to the second aspect. A step 2310 comprises providing a substrate with a cavity in the substrate. A step 2320 comprises arranging a fluid displacement device with an element suspended in the cavity and arranged parallel to a substrate plane in the cavity and movable in-plane, which element is designed to interact with a fluid in the cavity. A step 2330 comprises providing an actuator device that is designed to deflect main sides of the movable element alternately along a first direction for displacing a first fluid volume and along an opposite second direction for displacing a second fluid volume in-plane, such that at least a first main side of the movable element is curved in an unactuated state.Steps 2310, 2320, and 2330 may be performed in accordance with method 1600. Deviating from method 1600, method 2300 is performed such that the displaced first fluid volume is displaced from a first sub-cavity of the MEMS at a first main side of the movable element and the displaced second fluid volume is displaced from a second sub-cavity of the MEMS at an opposite second main side of the movable element, and the first sub-cavity and the second sub-cavity provide substantially equal fluidic damping for the deflection of the movable element.

[0174] In other words, several optimization parameters exist for the components described herein. At maximum deflection of the movable element, it can be ensured that no overtravel occurs, meaning that the movable element is always positioned at least partially opposite both stator electrodes. The fin benefits from an overlap with both cover electrodes; otherwise, significant harmonic distortions could occur. Furthermore, harmonic distortions must be minimized for the same or higher SPL or swept area of ​​the substrate. Furthermore, the surface area to the cover should be small to avoid cover pull-in. Deformation in the x / z plane (curvature of the fin) can reduce efficiency and must therefore also be minimized.

[0175] Possible component-specific parameters for this optimization include the non-straight fin shape, the corresponding shape of the electrodes in the lid / bottom wafer, as well as the depth, width, and position of optional V-grooves on the fins. Examples of implementations of non-straight fin shapes are provided.

[0176] Other arched shapes, such as beam eigenmodes, show similar results to the sinusoidal shape. Overhang-like structures, such as those shown in Fig. 11b, are also possible. For manufacturing purposes, maintaining point symmetry is easier with a smaller number of arches. However, for coupling elements between the air-displacing fin plane and a second drive plane, many arches or a combination with straight sections can be used to achieve an effectively mirror-symmetrical bending curve of the fin and thus minimize shear forces at the coupling point. Furthermore, the use of a higher number of arches can be used to further increase the torsional stiffness of the fin and thus avoid curvature along the fin depth, the z-direction. The structuring can therefore also be used as a degree of freedom for designing the frequency and shape of the fin eigenmode.

[0177] In connection with the arcs 54 of the movable elements described herein, an amplitude of less than 100 pm for the fin arcs can already be sufficient to achieve a significant reduction in strain stiffening, which is why, for example, a sinusoidal fin does not require any significant additional space compared to the conventional straight fin structure. The driving force of the fin can be implemented as a so-called balanced drive in all possible combinations for the use of electrodes in the cap wafer and / or bottom wafer.

[0178] For use with a drive element, this can also be controlled in both forward and reverse directions. For fin shapes that result in a deviation between the displaced volume in the forward and reverse directions, this difference can be compensated for by asymmetrically controlling the drive elements.

[0179] Embodiments of the present invention rely on at least some or combinations of the following features:

[0180] • A non-straight neutral fiber of the movable element to reduce strain stiffening

[0181] • A point-symmetrical shape to avoid straight harmonics and buckling

[0182] • A symmetrical shape, e.g. with DC offset, to minimize even harmonics

[0183] • Increased stiffness in x / z plane due to the non-neutral fiber

[0184] • Curved cover electrodes

[0185] • Notches in the moving elements

[0186] • Straight sections in the moving elements

[0187] Although some aspects have been described in connection with a device, it is understood that these aspects also represent a description of the corresponding method, so that a block or component of a device can also be understood as a corresponding method step or as a feature of a method step. Similarly, aspects described in connection with or as a method step also represent a description of a corresponding block, detail, or feature of a corresponding device.

[0188] The above-described embodiments are merely illustrative of the principles of the present invention. It is understood that modifications and variations of the arrangements and details described herein will occur to others skilled in the art. Therefore, it is intended that the invention be limited only by the scope of the following claims and not by the specific details presented in the description and explanation of the embodiments herein.

[0189] Sources

[0190] [1] Alcheikh, Nouha, Hassen M. Ouakad, and Mohammad I. Younis. "Dynamic analysis of straight stepped microbeams." International Journal of Non-Linear Mechanics 128

[0191] (2021 ): 103639.

[0192] [2] Hajjaj, Amal Z., and Nizar Jaber. "Controlling Resonator Nonlinearities and Modes through Geometry Optimization." Micromachines 12.11 (2021): 1381.

[0193] [3] Alcheikh, Nouha, Hassen M. Ouakad, and Mohammad I. Younis. "Dynamics of V- shaped electrothermal MEMS-based resonators." Journal of Microelectromechanical Systems 29.5 (2020): 1372-1381 . [4] WO 2022 / 117197 A1

[0194] [5] EP 4 147 461 A1

Claims

Patent claims 1. A MEMS comprising: a substrate (12) having a cavity (14) in the substrate (12); and a fluid displacement device (16) having an in-plane movable element (18) suspended in the cavity (14) and arranged parallel to a substrate plane in the cavity (14), said element being configured to interact with a fluid in the cavity (14); and having an actuator device (24) configured to alternately deflect main sides of the movable element (18) in-plane along a first direction (26i) to displace a first fluid volume (28i) and along an opposite second direction (26s) to displace a second fluid volume (28s); wherein at least one of the main sides (18A-B) of the movable element (18) is curved in an unactuated state; wherein the displaced first fluid volume (28i) and the displaced second fluid volume (28s) have substantially the same volume size.

2. MEMS according to claim 1, wherein the movable element (18) is suspended in-plane on both sides and symmetrically on the substrate (12).

3. MEMS according to claim 1 or 2, wherein the movable element (18) is suspended in-plane on both sides of a respective first and second fastening region on the substrate, wherein the first fastening region and the second fastening region are shifted relative to one another along a deflection direction of the movable element (18) in order to at least partially compensate for a lateral offset of a coupling region of the movable element transverse to the deflection direction.

4. MEMS according to claim 3, wherein the lateral displacement of the coupling region during actuation corresponds to the displacement of the bending eigenmode of the movable element.

5. MEMS according to claim 3 or 4, wherein the lateral offset is based at least in part on the curvature of the main side.

6. MEMS according to one of claims 3 to 5, wherein the coupling region is coupled to a mechanical device in another MEMS level.

7. MEMS according to one of the preceding claims, wherein the actuator device is configured to move the movable element perpendicular to a direction between suspension regions (58) of the movable element (18); or wherein the actuator device is configured to move the movable element parallel to a direction between suspension regions (58) of the movable element (18).

8. MEMS according to one of the preceding claims, wherein a curvature of the main side (18A) parallel to the substrate plane is inversion-symmetric.

9. MEMS according to one of the preceding claims, wherein a curvature of the main side (18A) substantially follows a course of a sine function or cosine function.

10. MEMS according to claim 9, wherein a course of the movable element (18) parallel to the substrate plane and from a first end to a second end has at least two changes of sign of a derivative of a curvature course of the curvature.

11. MEMS according to one of the preceding claims, in which a curvature of the main side (18A) deviates from a point symmetry and a control device (32) of the MEMS is designed to set a pre-deflection of the movable element (18) adapted to the curvature in order to equalize the first fluid volume (28i) and the second fluid volume (28s) to one another by means of the pre-deflection.

12. MEMS according to claim 11, wherein the curvature of the main side (18A) is substantially mirror-symmetrical to a plane perpendicular to a main extension direction of the movable element (18).

13. MEMS according to one of claims 1 to 12, wherein the main side (18A) of the movable element (18) is continuously curved in the unactuated state.

14. MEMS according to one of claims 1 to 12, wherein the main side (18A) of the movable element (18) is formed in sections straight in the unactuated state.

15. MEMS according to one of the preceding claims, wherein a course of the movable element (18) parallel to the substrate plane, starting from a center point (46) to one end, has at least two curvature arcs which have a different curvature amplitude (62) along a deflection direction of the movable element (18).

16. MEMS according to one of the preceding claims, with a plurality of adjacently arranged movable elements (18), with mutually facing or mutually remote main curvature.

17. MEMS according to one of the preceding claims, comprising a plurality of adjacently arranged movable elements (18) which have an identical or oppositely oriented main curvature.

18. MEMS according to one of the preceding claims, wherein a course of the movable element (18) parallel to the substrate plane from a first end to a second end has an overhang 19. MEMS according to one of the preceding claims, wherein the actuator device (24) has a stator electrode arrangement (36) with at least a first stator electrode (38i) and a second stator electrode (38s), which are arranged on the substrate (12) or are part of the substrate (12); wherein the actuator device (24) is designed to deflect the movable element (18) in the first direction (26i) based on an electrostatic force between the movable element (18) and the first stator electrode (38i); and to deflect the movable element (18) in the second direction (26s) based on an electrostatic force between the movable element (18) and the second stator electrode (38s).

20. MEMS according to claim 19, wherein in the unactuated state the movable element (18) is arranged centrally opposite the first stator electrode (38i) and the second stator electrode (38s).

21. MEMS according to claim 19 or 20, wherein the movable element (18) is arranged opposite to the first stator electrode (38i) and the second stator electrode (38s) in each operating position.

22. MEMS according to one of the preceding claims, wherein an electrode geometry of a stator electrode (38) of the actuator device (24) has a curvature which is adapted to a curvature of the movable element (18).

23. MEMS according to one of the preceding claims, wherein the actuator device (24) has a first stator electrode arrangement and a second stator electrode arrangement opposite perpendicular to the substrate plane, wherein the movable element (18) is arranged between the first stator electrode arrangement and the second stator electrode arrangement.

24. MEMS according to one of the preceding claims, wherein the movable element (18) is substantially uncurved in the unactuated state along a direction perpendicular to the substrate plane.

25. MEMS according to one of the preceding claims, with at least one opening (82) in the substrate (12) which is arranged along a direction parallel to a surface normal arranged perpendicular to the substrate plane and which has an opening geometry which is adapted to a curvature of the movable element (18).

26. MEMS according to one of the preceding claims, wherein the displaced first fluid volume (28i) is displaced from a first sub-cavity of the MEMS at a first main side of the movable element, and the displaced second fluid volume (28s) is displaced from a second sub-cavity of the MEMS at an opposite second main side of the movable element; and the first sub-cavity and the second sub-cavity provide substantially equal fluidic damping for the deflection of the movable element.

27. A MEMS comprising: a substrate (12) having a cavity (14) in the substrate (12); and a fluid displacement device (16) having an element (18) suspended in the cavity (14) and arranged parallel to a substrate plane in the cavity (14) and movable in-plane, said element being designed to interact with a fluid in the cavity (14); and having an actuator device (24) designed to alternately deflect main sides of the movable element (18) in-plane along a first direction (26i) to displace a first fluid volume (28i) and along an opposite second direction (26s) to displace a second fluid volume (28s); wherein the displaced first fluid volume (28i) is displaced from a first partial cavity of the MEMS on a first main side of the movable element and the displaced second fluid volume (28s) is displaced from a second partial cavity of the MEMS on an opposite second main side of the movable element;and the first sub-cavity and the second sub-cavity provide substantially equal fluidic damping for the deflection of the movable element; 28. MEMS according to claim 27, wherein at least one of the main sides (18A-B) of the movable element (18) is curved in an unactuated state; 29. MEMS according to claim 27 or 28, wherein the first partial cavity and the second partial cavity are geometrically incongruent.

30. MEMS according to claim 27 or 28, comprising an active element configured to adjust the fluidic damping of at least the second partial cavity.

31. MEMS according to claim 30, wherein the active element is configured to adjust a size of the second partial cavity by deforming the active element.

32. MEMS according to claim 31, wherein the active element is designed to adjust the size of the second partial cavity quasi-statically with respect to a dynamic operation of the movable element adjusted for sound generation.

33. The MEMS of any one of claims 27 to 32, wherein a plurality of movable elements are arranged adjacent to one another in the cavity, wherein adjacent movable elements of the plurality of movable elements are arranged alternately at a first distance and a second distance, and wherein a main side of a movable element has the first distance from a first adjacent movable element; and an opposite other main side of the movable element has the second distance from a second adjacent movable element.

34. MEMS according to claim 33, wherein a larger distance from the group comprising the first distance and the second distance is arranged on a side of the movable element along which a preferred direction of deflection of the movable element is arranged.

35. MEMS according to one of claims 27 to 34, wherein the fluidic damping is based on a squeeze film damping based on the rule is set, wherein; and the fluidic damping in the first partial cavity and the second partial cavity is substantially the same.

36. MEMS according to one of claims 27 to 35, wherein a size and a course of limiting structures are adapted to a nonlinear stiffness of the movable element and / or a clamping of the movable element.

37. MEMS according to one of claims 27 to 36, wherein the displaced first fluid volume (28i) and the displaced second fluid volume (28s) have substantially the same volume size.

38. A method (1600) for manufacturing a MEMS comprising the following steps: Providing (1610) a substrate having a cavity in the substrate; and Arranging (1620) a fluid displacement device with an element suspended in the cavity and arranged parallel to a substrate plane in the cavity and movable in-plane, which element is designed to interact with a fluid in the cavity; Providing (1630) an actuator device configured to deflect main sides of the movable element alternately along a first direction to displace a first fluid volume and along an opposite second direction to displace a second fluid volume in-plane; such that at least one of the main sides of the movable element is curved in an unactuated state; such that the displaced first fluid volume and the displaced second fluid volume have substantially the same volume size.

39. A method for manufacturing a MEMS comprising the following steps: Providing a substrate having a cavity in the substrate; and Arranging a fluid displacement device with an element suspended in the cavity and arranged parallel to a substrate plane in the cavity and movable in-plane, which element is designed to interact with a fluid in the cavity; Providing an actuator device configured to deflect main sides of the movable element alternately along a first direction to displace a first fluid volume and along an opposite second direction to displace a second fluid volume in-plane; such that at least one of the main sides of the movable element is curved in an unactuated state; such that the displaced first fluid volume (28i) is displaced from a first partial cavity of the MEMS on a first main side of the movable element and the displaced second fluid volume (28s) is displaced from a second partial cavity of the MEMS on an opposite second main side of the movable element; and the first partial cavity and the second partial cavity provide substantially equal fluidic damping for the deflection of the movable element.

Images