EP4002407A1 - Microelectromechanical switching element, device and manufacturing method - Google Patents

Abstract

A microelectromechanical switching element (1) is specified, comprising a bending element (10) with at least one first switching contact (11) arranged on the bending element, the bending element (10) being deflected from a central position (p0) in two opposite directions (r1 ,r2) can be deflected,- and a first cover substrate (100) and a second cover substrate (200),- the two cover substrates (100,200) being designed in such a way that the bending element (10) is in a cavity (300) between the two cover substrates (100,200) is arranged, - and wherein each of the two cover substrates (100, 200) in the area of the bending element (10) has a control electrode (110, 210) with which a deflection of the bending element (10) can be influenced Device with such a switching element (1) and a method for producing such a switching element (1).

Description

The present invention relates to a microelectromechanical switching element, which comprises a flexible element with at least one first switching contact arranged on the flexible element. Furthermore, the invention relates to a device with such a switching element and a method for producing such a switching element.

Microelectromechanical switching elements are known in principle from the prior art and are also referred to as MEMS switches in the technical field. These are mechanical solid-state switching elements that are structured in the micrometer to nanometer range and include electrostatically actuated bending elements so that they can be switched by changing an electrical voltage. A plurality of such individual MEMS switches is often arranged to form an array, in particular in order to achieve a sufficiently high current-carrying capacity and/or dielectric strength. Such MEMS switches and switching devices based on them are, for example, in DE102017215236A1 and the WO2018028947A1 described.

A disadvantage of the MEMS switches according to the prior art is that the flexible elements typically require relatively high forces in order to be brought from their basic position into a deflected position. Or to put it another way, the bending elements are typically designed in such a way that in the deflected position they exert a relatively high mechanical restoring force, by means of which they can independently return to their basic position after the switching voltage has been removed. On the one hand, this high restoring force has the advantage that the switching elements very reliably return to the basic position after the switching voltage has been removed and do not stick in the deflected state.

However, this also has the disadvantage that the movement of the bending element into its deflected position requires a switching voltage that is comparatively high in absolute terms. The switching voltage is the voltage that is applied between a portion of the flexure and an opposing control electrode to cause deflection of the flexure by electrostatic force. With conventional MEMS switches, these switching voltages are often around 10 V. In addition, with conventional switching elements, the switching duration is often undesirably long, which can also be attributed to the rigid design of the bending elements described: At least with a fixed switching voltage, a stiffer design of the bending element leads to this to a higher switching time.

The object of the invention is therefore to provide a switching element which overcomes the disadvantages mentioned. In particular, a switching element is to be made available which can be switched with a comparatively low switching voltage and/or which can be switched comparatively quickly. A further object is to specify a device with such a switching element and a method for producing such a switching element.

These objects are achieved by the switching element described in claim 1, the device described in claim 11 and the manufacturing method described in claim 14.

The microelectromechanical switching element according to the invention has a bending element with at least one first switching contact arranged on the bending element, wherein the bending element can be deflected from a central position into two opposite deflection directions. The switching element also has a first and a second cover substrate, the two cover substrates being designed in such a way that the bending element is arranged in a cavity between the two cover substrates. In doing so, each of the two cover substrates in the area of the bending element has a control electrode with which a deflection of the bending element can be influenced.

A microelectromechanical switching element is to be understood here as meaning a switching element which is produced using microsystems technology. The term microsystems technology is generally understood to mean the technology that is able to produce microscopically small mechanically active components, such as switches or gears that can move. In a broader sense, the term microsystems technology should also include nanosystems technology, which enables corresponding structures in the submicrometer to nanometer range. In general, technologies that are generally known from semiconductor production are used here. Such MEMS switches can be manufactured on glass and/or semiconductor substrates (so-called wafers), for example made of silicon or gallium arsenide. In this case, the length of a MEMS switching element is less than 1 mm, preferably less than 100 μm. Here, the largest structural element of a single MEMS switching element is typically the bending element. This bending element is expediently shaped in an elongate manner in order to enable a defined, resilient deflection in the manner of a straight leaf spring. The bending element is therefore often also referred to as a bending beam or as a switch tongue in the specialist circles. In principle, however, other proportions are also possible.

The bending element is located in an inner cavity of the superordinate switching element component, which is composed of several flat substrates and is therefore present overall as a flat component. In the case of a horizontal orientation, this component is covered on the outside "above" and "below" (that is to say towards its two main surfaces) by two cover substrates in such a way that the inner cavity is delimited by these two cover substrates. In conventional MEMS switching elements, the region of the flexure is in contrast typically limited only on one side by a cover substrate which, for example, as in the DE102017215236A1 can be designed as a glass wafer. In the known MEMS switches, this cover substrate carries the control electrode, which is placed opposite the bending beam and to which the switching voltage is applied. In the prior art, this control electrode is sometimes also referred to as the gate electrode. The bending element can be deflected by electrostatic interaction between the control electrode and the bending element. For example, it can be moved in the direction of the cover substrate due to electrostatic attraction, so that this deflection can result in the formation of an electrical contact between a contact element of the bending beam and a contact element of the cover substrate.

A significant advantage of the switching element according to the invention is that the cavity is delimited by two cover substrates and that each of these two cover substrates carries an associated control electrode in the area of the bending beam. In particular, a deflection of the bending element can then be influenced or brought about with each of these two control electrodes by temporarily applying a switching voltage. Since these two control electrodes (seen when the component is oriented horizontally) lie above and below the bending element, this bending element can thus be actively moved both from above and from below. In particular, this enables controlled deflections upwards and downwards from a central position. In principle, an upward or downward deflection can also be effected with just one control electrode, for example by means of an overhead control electrode causing an attractive or repulsive electrostatic force depending on the desired direction of movement. The latter variant is technically much more difficult to implement. In any case, actuation with two separate opposing control electrodes allows for much more controlled and precise guidance of the movement. In particular, by means of electrostatic attraction by means of the upper control electrode An upward deflection can be effected from a central position, and a downward deflection can be effected by electrostatic attraction by means of the lower control electrode. In this way, it is possible to switch between a central position, an upwardly deflected position and a downwardly deflected position of the bending element in a particularly controlled manner. For example, the central position can be a basic position of the bending element, into which the bending element automatically returns without the action of electrostatic forces. In principle, however, the bending element could also be prestressed, so that a position deflected “up” (ie to the first cover substrate) or a position deflected “down” (ie to the second cover substrate) can also form the mechanical basic position.

In general and independently of the precise configuration, the bending element is arranged between the two control electrodes in the switching element according to the invention, so that it can be actuated electrostatically by the interaction of both control electrodes. This results in several advantages for the control or the movement of the switch: On the one hand, the bending element can be designed to be significantly softer, in other words it can have a lower mechanical restoring force than if it were only deflected with a control electrode. This is because, for example after an electrostatic attraction of the bending beam in the direction of the first control electrode, the subsequent return in the direction of the second control electrode can be supported by a correspondingly attractive switching voltage on the second control electrode. In other words, the resetting from a first (electrostatically) deflected position does not have to take place purely mechanically, but can be supported electrostatically in a simple manner by the second electrode. On the other hand, a softer bending element with a smaller restoring force also allows the use of lower absolute values Switching voltages compared to the prior art (on both control electrodes).

Despite a softer configuration of the bending element, sticking of the bending element in a deflected state can be reliably prevented in the switching element according to the invention, since the second control electrode can support detachment from the deflected state by developing an electrostatic restoring force. Further advantages result for the dynamics of the switching behavior, since the two control electrodes and the softer bending element generally enable a faster and, if required, also a dynamically controlled switching process. Not only binary switching signals (switching voltage on or off) can be applied to the control electrodes, but more complex voltage profiles can be applied. The movement of the bending element can then be controlled particularly precisely, in particular through the interaction of the two opposing control electrodes. For example, the bending element can be moved in the direction of this first control electrode by switching on the switching voltage on the first control electrode (when the so-called "pull-in voltage" is exceeded). With a suitable voltage profile on the opposite second control electrode, braking can also be effected shortly before the bending element mechanically impacts, which reduces the mechanical stress on the bending element (and its contact elements) and can thus significantly extend the service life of the switching element. Expressed more generally, suitable voltage profiles at the two control electrodes can be used to achieve predetermined switching characteristics and thus the corresponding movement characteristics of the flexible element much more precisely than with just one control electrode. It is also possible, in particular, not only to return to the basic position when resetting from a deflected position of the bending element and an associated electrical disconnection of the corresponding contact elements, but also to temporarily move the bending element beyond the neutral basic position to "overbend". In this way, an arc between the contact elements that have just been separated can be extinguished much more reliably than if only the basic position were reached, since the distance between the contact elements is increased and the electrical field is thus further reduced.

The device according to the invention has a switching element according to the invention or an array of several such switching elements according to the invention as partial element(s). Regardless of the specific application of the device, its advantages are analogous to the above-described advantages of the switching element according to the invention, in particular with regard to a reduced switching voltage, shorter switching time and/or a more precise setting of a desired movement profile.

The method according to the invention serves to produce a switching element according to the invention. In this case, the bending element is formed by subtractive manufacturing from at least one flat layer and is then freed. This layer forming the bending element or the layer system containing this layer is connected to the first cover substrate and the second cover substrate by a total of two wafer bonding steps. The "subtractive production" of the bending element is to be understood as meaning that material is removed within the planar layer in an area around the bending element, so that the bending element remains within its layer like an island or at least like a peninsula. In this way, the bending element is separated from the "mainland" (ie the surrounding areas of the same layer), so to speak, to such an extent that it can, in principle, be deflected independently of these areas. At most, a type of web remains in the foot area of the bending element, by which it can be connected to the other areas of the same layer.

Under the subsequent "release" of the bending element should be understood that at least on a major Part of the longitudinal extension of the bending element, the mechanically load-bearing flat layers that may be adjacent here on both main surfaces (if still present) are removed so that the bending element can be deflected upwards and downwards in the manner of a straight leaf spring perpendicular to the layer surface. Typically, the bending element is already exposed to one of its main surfaces in this step, so that a mechanically supporting adjacent layer only has to be removed on a rear side in order to enable the required mobility.

Overall, in any case, the layer or the layer system which contains the flexible element is connected to the first cover substrate and the second cover substrate by a total of two wafer bonding steps. In this case, an additional carrier substrate can optionally be used, which, for example, initially carries the layer system of the flexible element before it is connected to the first cover substrate on its free side. After subsequent removal of this carrier substrate, the opposite side of the layer system (which is then exposed) can also be connected to the second cover substrate in a similar way via a wafer bonding step. In this way, an overall structure is achieved in which the layer system that contains the bending element is sandwiched between the two cover substrates. A recess is expediently created in the two cover substrates before the respective wafer bonding step in the area in which the bending element is arranged after the bonding step. These two recesses form the hollow space in which the bending element is arranged so that it can be deflected in the finished state.

Advantageous refinements and developments of the invention emerge from the claims dependent on claims 1, 11 and 14 and from the following description. The described configurations of the switching element, the device and the production method can generally be advantageously combined with one another. So can in particular the first deflection direction of the bending element can correspond to a deflection towards the first cover substrate, and the second deflection direction can correspond to a deflection towards the second cover substrate. In other words, the bending element can be deflected “up” and “down” when the overall component that is shaped flat is horizontal. Such mobility out of the layer plane of the bending element is particularly easy to implement with a bending element shaped like a leaf spring.

According to a particularly preferred embodiment, the bending element can be deflected in the direction of the first cover substrate by applying voltage to the control electrode of the first cover substrate, and it can be deflected in the direction of this second cover substrate by applying voltage to the control electrode of the second cover substrate. This can be achieved in particular by temporarily applying a switching voltage to each of the two control electrodes for deflecting the bending element in the relevant direction, which causes an electrostatic attraction force between the bending element and the associated cover substrate. Alternatively, however, the respective control electrode can in principle also be subjected to a switching voltage, which results in a repelling electrostatic force between the bending element and the associated cover substrate. In connection with the present invention, it is particularly important that a movement of the bending element in both directions of deflection can be actuated separately due to the presence of two control electrodes. By simultaneously applying voltage to both control electrodes, a particularly precise movement control according to a predefined characteristic can be achieved in general.

In a generally advantageous manner, the bending element can essentially be formed from a semiconductor material. This is particularly favorable because it is advantageous for maintaining defined switching properties if the mechanical properties of the bending element can firstly be precisely reproduced from component to component in a defined manufacturing process and secondly if they remain stable for a given component over a longer period of operation. This can easily be achieved with known semiconductor layer systems, since the processing of such semiconductors can be easily controlled and has reached a very advanced level of development. In addition, it is comparatively inexpensive. Silicon, for example, is therefore particularly suitable as a semiconductor material. Monocrystalline silicon in particular can be processed very reproducibly in terms of its mechanical and electrical properties and is stable over the long term.

Alternatively, however, the bending element can also be formed essentially from a metal and/or one or more dielectrics. "Essentially" is intended here to mean in general that the materials mentioned form the main component, with other additional materials, e.g. in the form of coatings of the bending element, not being excluded. In particular, the bending element can carry one or more contact elements, which can be applied to the bending element in the form of structured metallization. If the bending element is formed essentially from a dielectric material, it can expediently carry a control electrode on its upper and lower side, which is arranged opposite the respective control electrode of the cover substrate in the finished switching element. On the other hand, if the bending element is configured from a semiconducting or metallic material, the bending element itself can form the counter-electrode for these two control electrodes of the cover substrates.

Generally advantageously, at least one of the cover substrates and in particular even both cover substrates can be embodied as an insulating cover substrate. In particular, they can essentially be made of glass. Again, this should be understood to mean that the glass or other insulating Material forms the main component of the respective substrate and additional elements, in particular in the form of local coatings, should not be excluded. In particular, the named control electrodes and additional contact elements can be applied as further elements on the surface of the cover substrates. In particular, control electrodes and contact elements can be arranged on the side of the cover substrates facing the bending element. However, metallizations in the form of line elements and/or contact elements and/or other components, for example, can also be present on the outside of the cover substrates. Electrical feedthroughs can be provided between the two main surfaces of the respective cover substrate, in particular in the form of so-called vias, which extend through the substrate perpendicularly to the substrate surface in order to electrically connect elements on the top and bottom. Alternatively, however, conductive substrates can also be coated with an electrically insulating layer on the side facing the bending element, so that an electrically insulating cover substrate is formed purely functionally. Such an electrically insulating layer can be a silicon dioxide layer or else a polymer layer, for example.

According to a generally advantageous embodiment, the cavity for the bending element can be formed by recesses in the two cover substrates. In particular, these recesses can be arranged opposite one another in the two cover substrates and can correspond to one another in terms of their outline, so that a common cavity is formed as a result of their interaction. The bending element can be correspondingly deflected upwards and downwards in this common cavity. The metallizations of the two control electrodes are advantageously arranged in the area of the recesses of the respective cover substrates. In this way, the respective control electrode can interact electrostatically with the bending element arranged within the cavity.

In a generally advantageous manner, the switching element can have a silicon-on-insulator layer system or part of such a layer system in the area between the two cover substrates. A silicon-on-insulator (abbreviated: SOI) layer system includes, in particular, a silicon-insulator-silicon layer sequence. Advantageously, one of the two silicon layers and particularly preferably both silicon layers can be monocrystalline silicon. The insulator layer can preferably essentially be formed by a silicon dioxide layer (SiO ₂ ). In particular, it can be implemented as a so-called buried oxide layer or BOX layer for short. The SOI technology, which is known per se, enables a particularly precise definition of the layer thicknesses and the other material properties of the individual layers. In particular, the bending element or at least a mechanically supporting part of the bending element can be essentially realized by one of these silicon layers. In this case, the bending beam can be produced subtractively, for example by etching away surrounding areas of the silicon. In particular, the layer of the cantilever is the thinner of the two silicon layers of a typical SOI substrate. The mechanical connection of the bending element to the other areas of the component can be mediated in particular via an insulator layer of the SOI substrate. In other words, the bending element can be coupled in its foot area via the insulator layer to a mechanically supporting part of the switching element. These and other advantageous features and variants of the SOI layer system and its processing are detailed in the DE102017215236A1 described, which should therefore be included in its entirety in the disclosure of the present application.

In the case of the finished component according to the invention, however, the SOI layer system described no longer has to be completely preserved. In particular, the originally thicker of the two silicon layers can be completely removed after the component has been completed. The insulator layer can also be completely or at least partially removed. What is essential in this embodiment is that at least the bending beam has been produced from a partial layer of the SOI layer system.

In general, it is expedient if the first cover substrate carries a first counter-contact, which can be electrically contacted with a first switching contact of the bending element in a first deflection position of the bending element. When the flexible element is deflected towards the first cover substrate and is in contact with it in its end region, an electrical contact should be established between the first switching contact of the flexible element and a first mating contact arranged opposite on the first cover substrate. Particularly preferably, the first cover substrate even carries a pair of first mating contacts, both of which can be contacted with the first switching contact of the bending element. In this way, the pair of first mating contacts is electrically connected to one another in the first deflection position. These two first mating contacts can be designed as so-called load contacts of a first load circuit. The first load circuit can thus be closed by means of the first switching contact. With regard to the first load circuit, the switching element is then in an “ON” position.

According to an advantageous development of the invention, the second cover substrate can also carry a second mating contact, which can be contacted with a second switching contact of the bending element in a second deflection position of the bending element. The functioning of the second switching contact in interaction with the second counter-contact is analogous to that already described in connection with the first switching contact and the first counter-contact. Here too, on the second cover substrate, in particular a pair of second mating contacts can be present, which can be electrically connected to one another by means of the second switching contact. In this way, in particular, a second load circuit can be formed be, which is closed in the second deflection position of the bending element. In other words, the bending element can then be used to switch between an “ON” position for the first load circuit and an “ON” position for the second load circuit. In principle, it is also possible for the bending beam to assume a central basic position, in which the switching element is switched to “OFF” with respect to both load switching circuits.

The bending element can be designed in such a way that this central basic position is assumed when the two control electrodes are in a voltage-free state. Alternatively, the bending element can also be mechanically prestressed in the first or second deflection direction by the type of processing, so that the "ON" position of the first load circuit or the "ON" position of the second load circuit is the basic position of the switching element forms. Such a bias can be achieved, for example, as in the not yet published European patent application with the application file number 20182568.4 is described. This application should therefore also be included in the disclosure of the present application.

The configuration with a second switching contact and a second mating contact is particularly advantageous since it can be used to form a changeover switch. Such a changeover switch is also referred to as a changeover relay and enables switching between two load circuits. An input line can thus be optionally connected to a first or a second output line. This can be implemented, for example, by electrically connecting an individual contact of the upper contact pair via a via to an individual contact of the lower contact pair, so that the two individual contacts are connected together to the input line. The realization of a changeover switch is particularly advantageous for applications in telecommunications, for example, or for all types of logic circuits in which switching between two or more exits is desirable. A combination of several such switching elements, in particular connected in series, can also be used to switch between more than two outputs.

In general and independently of the precise configuration of the individual switching element, the superordinate device can have an array of a plurality of switching elements according to the invention. Such an array can be a parallel connection and/or a series connection of several such switching elements. A parallel connection of several switching elements can be used in particular to increase the current carrying capacity of the entire device compared to a single switching element. A series connection of a plurality of switching elements can be used in particular to increase the electric strength compared to a single switching element. In this way, the use of an array can help the device meet the specifications of a circuit breaker of an electrical power distribution line, particularly in a low-voltage or medium-voltage network. The number of individual switching elements in the array can be aligned with the respective specifications. For example, it can be a few 10 to a few 1000 switching elements and even be several hundred thousand for higher power ranges.

According to a further advantageous embodiment of the device, in addition to the at least one switching element, it can comprise one or more semiconductor elements electrically connected in series or parallel thereto. This can be, for example, transistors or other semiconductor switching elements, as in the WO2018028947A1 described. In principle, the additional semiconductor elements can be manufactured on the same substrate as the bending element, that is to say they can be monolithically integrated, or they can in principle also be manufactured on a different substrate and only subsequently connected to the switching element.

The higher-level device can be designed for use in very different applications. For example, it can be designed as a switching device or contactor, as a converter or as an inverter, as a logic circuit and/or as a logic gate. The switching device or contactor can in particular be a device for a low-voltage or medium-voltage network. In general, the device can also be a programmable logic controller, in particular a controller for an industrial plant. Switching elements according to the invention can be used, for example, in an input stage, an output stage and/or in a safety relay of such a system controller.

According to an advantageous development of the manufacturing method, the bending element can be obtained by subtractive manufacturing from a silicon-on-insulator layer system, as has already been described above for the corresponding embodiment of the switching element. In particular, the bending element can be formed by exposing part of a silicon layer of the SOI layer system. The production method can include a large number of further optional production steps, which are known in principle from MEMS and semiconductor processing in particular. For example, additional metallization steps can be provided, for example in the form of a coating by vapor deposition, sputtering or galvanic deposition. The metallic layers for the electrodes and/or contact elements can include, for example, gold, chromium or silver or other metals that are common in semiconductor production. The (complete or partial) removal of individual layers can take place, for example, by etching and/or mechanical/chemical polishing and/or by a lift-off process. Above all, etching processes such as chemical etching with hydrofluoric acid, reactive ion etching (RIE for "reactive ion etching" or DRIE for "deep reactive ion etching") can be used for the defined removal of (partial) layers. A precise structuring can usual lithographic structuring methods can be achieved.

In principle, the respective substrates can be handled during production by gripping and positioning individual "functional substrates" of the component, e.g. the SOI and cover substrates. Alternatively, however, such a functional substrate can also be held during processing by an additional carrier substrate and, so to speak, carried piggyback through the process by this. Such an additional carrier substrate can be removed in a simple manner, for example by a so-called tilt-release step, without the functional substrates being damaged in the process.

Figur 1

eine schematische Querschnittsdarstellung eines Schaltelements nach dem Stand der Technik zeigt,

Figur 2

eine perspektivische Ausschnittsdarstellung eines weiteren bekannten Schaltelements zeigt,

Figur 3

eine Querschnittsdarstellung eines erfindungsgemäßen Schaltelements zeigt,

Figuren 4 bis 9

mehrere Prozessstadien bei der Herstellung des Schaltelements der Figur 3 zeigen,

Figuren 10 bis 12

schematische Aufsichten auf Teilbereiche eines erfindungsgemäßen Schaltelements zeigen und

Figur 13

eine schematische Querschnittsdarstellung eines Schaltelements nach einem weiteren Beispiel der Erfindung zeigt.

The invention is described below using some preferred exemplary embodiments with reference to the attached drawings, in which:

figure 1

shows a schematic cross-sectional representation of a switching element according to the prior art,

figure 2

shows a perspective detail view of another known switching element,

figure 3

shows a cross-sectional representation of a switching element according to the invention,

Figures 4 to 9

several process stages in the manufacture of the switching element figure 3 demonstrate,

Figures 10 to 12

show schematic views of parts of a switching element according to the invention and

figure 13

shows a schematic cross-sectional representation of a switching element according to a further example of the invention.

Elements that are the same or have the same effect are provided with the same reference symbols in the figures.

figure 1 shows a microelectromechanical switching element 1, which is known from the prior art. This switching element is structured in a similar way to that in DE102017215236A1 described, but it is shown here in a somewhat simplified manner for the sake of clarity. The switching element has a silicon-on-insulator layer system 50, which here includes a first silicon layer 51, then a first oxide layer 52 and then a second silicon layer 53. The layer 53 can also be followed by a second oxide layer as part of the SOI layer system 50 or at least have been present here during production. A flexible element 10 is defined from this SOI layer system 50 and in particular from its second silicon layer 53 by subtractive production and is then released. The production of the bending element includes the removal of the surrounding areas of the silicon layer 53 and the local removal of the parts of the layers 51 and 52 adjoining the bending element 10. In this way, a bending element 10 that can be deflected in the direction of thickness d was produced, similar to that in FIG DE102017215236A1 described. The other manufacturing steps can be analogous to that DE102017215236A1 be performed. The foot area 10a of the bending element, in which this is mechanically connected to the remaining parts of the SOI layer system, is designed and shown here in a somewhat simplified manner. As an alternative to this simpler realization, the foot area 10a can also be analogous to DE102017215236A1 or the not yet published European patent application 20182568.4 be designed.

Towards the top, the component is covered by a cover substrate 100, which can be made of glass, for example. This cover substrate 100 can have been connected to the remaining layers of the SOI layer system 50 by a wafer bonding step, for example. In the area of the bending beam, the cover substrate has a recess so that together with the removed parts of the layers 51 and 52 a cavity is formed in which the bending element 10 can be deflected. In order to bring about this deflection, a control electrode 110 is provided on the cover substrate 100 in the region above the bending element 10 . By applying voltage of the control electrode 110, a deflection of the bending element can be electrostatically actuated. If the bending element 10 is deflected upwards in direction d, its end region can be brought into contact with the cover substrate to such an extent that a switching contact 11 applied to the bending element 10 and a counter-contact 120 applied in the recess of the cover substrate 100 are electrically connected to step. In this way, the switching element can be switched to "ON", and an associated load circuit can be closed with the switching element 1.

The switching element 1 shown can be present as part of a higher-level device, which in particular can have an array of switching elements that are similar to one another. Such a MEMS array can be constructed monolithically from the same substrates. In this case the in figure 1 The area shown is to be understood as a section from a larger component, with the lateral layers extending even further to the right and left (and of course also perpendicular to the plane of the paper) and in these spatial directions can also include other similarly constructed switching elements.

figure 2 shows a perspective detail view of such a conventional switching element, which is particularly similar to the switching element in terms of its layer sequence figure 1 can be constructed. However, the spatial orientation is the opposite of that figure 1 selected so that the direction d points downwards here and the cover substrate 100 lies below the bending element 10 . In the foot area 10a of the bending element, this is connected to the other layers of the SOI layer system 50 and connected to the covering substrate 100 via the planar wafer connection. In the central region, the bending element is relatively wide and electrostatically interacts with the control electrode 110 on the cover substrate, in particular when a voltage is applied to it via the control circuit outlined. In the end area 10c, the bending element is also relatively wide and carries a switching contact that is not visible here. Opposite In this example, there is a pair of adjacent mating contacts 120 on the cover substrate (that is, an input contact 121 and an output contact 122). When the bending element is deflected, these two contacts 121 and 122 are electrically connected to the switching contact of the bending element and also to one another through this. The "waisting" shown - i.e. the reduced width of the bending element 10 both between the base area 10a and the middle area 10b and between the middle area 10b and the end area 10c - serves to enable this twisting to be corrected even if there is any torsion of the bending element, see above that as extensive a contact as possible or an extensive interaction of the mutually opposite regions of the bending element 10 and the covering substrate 100 can come about.

figure 3 shows a schematic cross-sectional representation of a switching element 1 according to a first example of the invention. The basic functionality is analogous to the functionality of the conventional switching element 1 of Figures 1 and 2 , unless otherwise described below. In particular, this functionality is expanded as described below.

So also has the switching element 1 of the invention figure 3 a bending element 10 which can be deflected perpendicularly to the substrate plane. Here, too, the bending element is produced subtractively from a silicon-on-insulator layer system, for example from a silicon layer 53 of such a layer system. Similar to in figure 1 the component is covered at the top by a first cover substrate 100, which has a recess in the region of the bending element and carries a first control electrode 110 and a first mating contact 120 (or a pair of such mating contacts) there. The bending element 1 has a first switching contact 11, which can be brought into connection with the at least one counter-contact when it is deflected upwards. The deflection can also be brought about here by applying a voltage to the control electrode 110 .

In contrast to the switching element of figure 1 However, here the first silicon layer and the first oxide layer of the SOI layer system have been removed from the underside of the component. Instead, the component is covered here by a second cover substrate 200, which is designed analogously to the first cover substrate 100 and can therefore also be essentially made of glass. This second cover substrate was connected to the remaining layers of the SOI layer system (here the silicon layer 53) by a planar connection in a wafer bonding step. This flat connection also results in an encapsulation of the component towards the bottom. In the section shown, the planar connection is only interrupted in the immediate vicinity of the bending element 10, since the second cover substrate 200 has a recess here, which together with the corresponding recess in the first cover substrate forms a superordinate cavity 300 in which the bending element 10 can be moved .

It is essential in connection with the present invention that the second cover substrate 200 also has a control electrode in the area of the bending element 10 . This second control electrode is labeled 210 here. Thus, by applying voltage to the two control electrodes 110 and 210 (either simultaneously or alternately), a movement of the bending element in two opposite deflection directions r1 and r2 can be brought about. This movement can be controlled much more precisely according to a desired characteristic than is possible with only one control electrode in the conventional switching element.

In addition to the elements and functionalities described so far, the switching element of the figure 3 have further features, which, however, are to be regarded as optional in connection with the present invention and these only in particular advantageously further develop: The bending element can have a second switching contact 21 on its underside (ie opposite to the first switching contact). In addition, the second cover substrate 200 can have one or more second mating contacts 220 in the corresponding area, which can be brought into an electrically conductive connection with the second switching contact 21 when the bending element is deflected downwards. In this embodiment variant, the switching element forms a changeover switch which can switch over between a first load circuit and a second load circuit. The first load circuit is closed when the first switching contact 11 is connected to the at least one first mating contact 120, and the second load circuit is closed when the second switching contact 21 is connected to the at least one second mating contact 220.

The choice of materials for the individual layers and elements can, for example, be analogous to that already cited DE102017215236A1 be designed. The layer thicknesses and other properties can also be configured analogously to the embodiments there. Likewise, the configuration of the foot area of the bending element can deviate from the simplified form shown here analogously, as in FIG DE102017215236A1 or the as yet unpublished European application 20182568.4 be realised.

Reference is also made to the two patent applications cited in the previous paragraph for the production method for the switching element according to the invention. The manufacturing method according to the invention initially takes place analogously to the method steps described there and only deviates from the manufacturing method there after the connection of the first cover substrate 100 to the SOI layer system 50 . Figures 4 to 9 show a sequence of selected process stages for an exemplary manufacturing method according to the present invention. So shows figure 4 an intermediate stage in the manufacture of the switching element figure 3 , which largely in analogous process control as in the Figures 1 to 13 the DE102017215236A1 was produced. An essential difference to the method there is merely that the two underlying layers 51 and 52 of the SOI layer system 50 are still retained over the entire area here and not as in the method there figure 13 were opened in the area of the bending beam. This from the component of the local figure 13 deviating state can be achieved in that the SOI layer system in an earlier process stage, which of the local figure 5 corresponds, with the up to the stage of the local figure 12 processed cover substrate (corresponds to the cover wafer 200 there) is connected over a large area in a first wafer bonding step. In this process stage, the bending element 10 has already been subtractively manufactured within the layer 53, but has not yet been released from its areal connection to the adjacent layers 51 and 52.

Further, rather insignificant differences between the process stage shown here figure 4 and the component of the local figure 13 exist in the slightly more complex foot area of the bending element there and in further optional additional elements such as the layer 240 there.

Based on the stage of the process figure 4 the further process steps are now described according to an example of the production method according to the invention. Examples of this are in the Figures 5 to 9 some other selected stages are shown. So is in figure 5 the lower silicon layer 51 has been at least largely removed. This removal can be implemented, for example, by chemical-mechanical polishing or by an etching process or a combination of these methods. In this case, however, the thin first oxide layer 52 largely remains as a residual layer in the component. The possible remaining of a thin residual layer of the silicon layer 51 should not be ruled out either. In the between figure 5 and figure 6 In the next step, this residual layer is then opened locally in the end area of the bending element, so that a recess 52a is formed in the first oxide layer 52. This local opening can be made, for example, by lithographic structuring and an etching process. A metallic layer can then be deposited in the region of this opening 52a, as a result of which the second switching contact 21 is formed here. This corresponds to the stage of figure 7 . Subsequently, the residual layer 52 can be removed (for example, again by an etching step), whereby the in figure 8 stage shown is reached. Finally, in a subsequent step, the component that has been thinned down in this way is connected areally to the second cover substrate 200 in a second wafer bonding step. Previously, this second cover substrate was provided with at least one recess 205 analogously to the first cover substrate 100 and provided there with the control electrode 210 and the optional second counter-contact (or a corresponding pair of counter-contacts), analogously to the processing of the first cover substrate 100 and thus to the "glass wafer". the DE102017215236A1 . However, the structuring or generation of the second switching contact is in accordance with the Figures 6 and 7 to be regarded as optional within the scope of the present invention. What is important, however, is that the component is covered on both sides by two cover substrates 100, 200 with their corresponding recesses 105 and 205 and the fact that in each of these two cover substrates 100 and 200 an associated control electrode 110 or 210 is formed.

In the Figures 10 to 12 Schematic views are shown for an exemplary realization of a contact topology, with which in particular the functionality of a changeover switch can be achieved. The figures each show a view along the thickness direction d, with the same detail being shown in each case, but different partial planes (partially superimposed). So shows figure 10 shows the bending element 10 in its end region together with the first switching contact 11 arranged thereon this figure shows the first mating contacts arranged in the region of the first cover substrate, ie overlying it. In this example, there is a pair of first mating contacts, namely a first input contact 121 and a first output contact 122. These two contacts 121 and 122 become electrically conductive with the first switching contact 11 and thus also with one another when the bending beam is correspondingly deflected towards the first cover substrate tied together.

figure 11 shows the same bending element 10 together with the second switching contact 21, which is arranged on the second cover substrate 200 facing surface of the bending element. In addition, this figure shows the second counter-contacts arranged in the region of the second cover substrate, that is to say underneath. Here too there is a pair of second mating contacts, namely a second input contact 221 and a second output contact 222. The figures 10 and 11 also show a via 400, through which the two input contacts 121 and 221 are connected across the component levels to form a higher-level input. Depending on the deflection position of the bending element, the input can thus alternatively be connected either to a first output or to a second output. In contrast, in the central position of the bending element, the input is not connected to either of the two outputs. A changeover switch is thus implemented, which can be switched to either the first output, the second output or to "OFF". figure 12 finally shows an overlay of the figures 10 and 11 elements already shown, through which this switching from a common input (connected contacts 121 and 221) to two alternative outputs (122 or 222) is illustrated.

figure 13 finally shows a schematic cross-sectional representation of a switching element 1 according to a further example of the invention. The area around the flexure 10 with its surrounding control electrodes and contacts is analogous to, for example, FIG figure 3 designed. The example of figure 13 differs This is essentially due to the fact that the two control electrodes 110 and 120 are connected here to contact points located on the surface of these cover substrates by two vias 400 passed through the two cover substrates 100 and 200 . The second mating contact 220 is also connected here by such a via 400 to a contact point in the region of the covering substrate 100 lying opposite. In a similar way, the first mating contact or optionally present further contacts (not shown here) (e.g. 121,122,221,222 as in the example of figure 12 ) must be connected to an external contact point by a corresponding via. In principle, it is also sufficient for contacting the individual contacts and electrodes if such vias are only routed through one of the two cover substrates to the outer surface.

Reference List

1

MEMS switching element

10

bending element

10a

footer

10b

middle area

10c

free end

11

first switch contact

21

second switch contact

50

SOI layer system

51

first silicon layer

52

first oxide layer

52a

Opening of the first oxide layer

53

second silicon layer

100

first cover substrate

105

Recess of the first cover substrate

110

first control electrode

120

first mating contacts (first pair of load contacts)

121

first input contact

122

first output contact

200

second cover substrate

205

Recess of the second cover substrate

210

second control electrode

220

second mating contacts (second pair of load contacts)

221

second input contact

222

second output contact

300

superior cavity

400

Via

i.e

thickness direction

p0

central position

r1

first deflection direction

r2

second deflection direction

Claims (15)

Microelectromechanical switching element (1), comprising - a bending element (10) with at least one arranged on the bending element first switching contact (11),
wherein the bending element (10) can be deflected from a central position (p0) in two opposite deflection directions (r1, r2), - and a first cover substrate (100) and a second cover substrate (200), - wherein the two cover substrates (100, 200) are designed such that the bending element (10) is arranged in a cavity (300) between the two cover substrates (100, 200), - And wherein each of the two cover substrates (100, 200) in the area of the bending element (10) has a control electrode (110,210) with which a deflection of the bending element (10) can be influenced. Switching element (1) according to Claim 1, in which the first deflection direction (r1) of the bending element (10) corresponds to a bend in the direction of the first cover substrate (100) and the second deflection direction (r2) corresponds to a bend in the direction of the second cover substrate (200). . Switching element (1) according to one of Claims 1 or 2, in which the bending element (10) - Can be deflected in the direction of the first cover substrate (r1) by applying a voltage to the control electrode (110) of the first cover substrate (100). - And by applying a voltage to the control electrode (210) of the second cover substrate (200) in the direction of the second cover substrate (r2) can be deflected. Switching element (1) according to one of the preceding claims, in which the bending element (10) is essentially formed from a semiconductor material, in particular silicon, and/or a metal and/or one or more dielectrics. Switching element (1) according to one of the preceding claims, in which the cover substrates (100, 200) are functionally designed as electrically insulating cover substrates and are in particular formed essentially from glass. Switching element (1) according to one of the preceding claims, in which the cavity (300) for the bending element (10) is formed by recesses (105, 205) in both cover substrates (100, 200). Switching element (1) according to one of the preceding claims, which has a silicon-on-insulator layer system (50) or a part (53) of such a layer system in the region between the two cover substrates (100, 200). Switching element (1) according to one of the preceding claims, in which the first cover substrate (100) carries a first mating contact (120) which can be contacted with a first switching contact (11) of the bending element (10) in a first deflection position of the bending element (10). . Switching element (1) according to Claim 8, in which the second cover substrate (200) also carries a second mating contact (220) which can be contacted with a second switching contact (21) of the bending element (10) in a second deflection position of the bending element (10). Switching element (1) according to Claim 9, which is designed as a changeover switch. Device with a microelectromechanical switching element (1) or an array of several microelectromechanical switching elements (1) according to one of the preceding claims. Device according to claim 11, which in addition to the at least one microelectromechanical switching element (10) or has a plurality of semiconductor elements connected electrically in series or in parallel. Device according to one of Claims 11 or 12, which is designed as a switching device, as a converter or inverter, as a logic circuit and/or as a logic gate. Method for producing a switching element (1) according to one of the preceding claims, in which - The bending element (10) is formed by subtractive manufacturing from at least one flat layer (53) and then freed - and this layer (53) forming the bending element (10) or a layer system containing this layer (53) is connected to the first cover substrate (100) and the second cover substrate (200) by a total of two wafer bonding steps. Method according to Claim 14, in which the bending element (10) is obtained by subtractive manufacture from a silicon-on-insulator layer system (50).

Images