MEMS devices and methods of using and manufacturing MEMS devices

Abstract

A micro-electro-mechanical system (MEMS) device is disclosed that comprise a substrate, a suspended membrane suspended over the substrate between a first support and a second support, and an actuation mechanism for moving the suspended membrane relative to the substrate. The actuation mechanism has a first actuation electrode and a second actuation electrode associated with the substrate, and at least one floating electrode supported by the suspended membrane, wherein each floating electrode overlies at least a portion of both the first and second actuation electrodes and is electrically unconnected to the first and second actuation electrodes. Application of a potential difference between the actuation electrodes creates a first charge on a surface of the first actuation electrode and a second charge on a surface of the second actuation electrode, the first and second charges inducing respective first and second opposite charges on overlapping regions of the floating electrode, which causes the floating electrode and the suspended membrane to electrostatically move the towards the substrate.

Description

MEMS devices and methods of using and manufacturing MEMS devicesTECHNICAL FIELD

[0001] The present disclosure relates to a micro-electromechanical systems (MEMS) device, a method of using a MEMS device and a method of manufacturing the MEMS device.BACKGROUND

[0002] Micro-electromechanical systems (MEMS) devices are microscopic devices which incorporate electrical parts and moving mechanical parts. MEMS devices generally range in size from the micron scale to the millimetre scale; in general the dimensions are below 25 sq mm in area. The components of a MEMS device typically have an area in the order of 100 square microns up to 25 square millimeters . MEMS devices have many applications including, but not limited to, switches, actuators and sensors. One application is for use in optical devices, such as optical switches and spectral filters.

[0003] Any discussion of documents, acts, materials, devices, articles or the like which has been included in the present specification is not to be taken as an admission that any or all of these matters form part of the prior art base or were common general knowledge in the field relevant to the present disclosure as it existed before the priority date of each of the appended claims.

[0004] SUMMARY

[0005] In accordance with a first aspect of the present disclosure, there is provided a micro-electro-mechanical system (MEMS) device comprising a substrate, a suspended membrane suspended over the substrate between a first support and a second support; and an actuation mechanism for moving the suspended membrane relative to the substrate, the actuation mechanism comprising a first actuation electrode and a secondactuation electrode associated with the substrate, and at least one floating electrode supported by the suspended membrane, wherein each floating electrode overlies at least a portion of both the first and second actuation electrodes and is electrically unconnected to the first and second actuation electrodes, wherein application of a potential difference between the first and second actuation electrodes creates a first charge on a surface of the first actuation electrode and a second charge on a surface of the second actuation electrode, the first and second charges inducing respective first and second induced opposite charges on the respective overlapping regions of the floating electrode, which causes the floating electrode and the suspended membrane to electrostatically move the towards the substrate.

[0006] In an embodiment, the suspended membrane is electrically insulating.

[0007] In an embodiment, the suspended membrane has an inner region which does not have any floating electrodes, and the at least one floating electrode is positioned adjacent the inner region

[0008] In an embodiment, a dielectric layer is positioned between the first actuation electrode and the suspended membrane.

[0009] In an embodiment, the MEMS device comprises at least one first actuation electrode that surrounds at least one second actuation electrode, the at least one second actuation electrode surrounding an inner region of the substrate.

[0010] In an embodiment, the suspended membrane comprises an inner region and at least 2 floating electrodes disposed on opposite sides of the inner region.

[0011] In an embodiment, the MEMS device comprises a plurality of floating electrodes on each side of the inner region.

[0012] In an embodiment, the plurality of floating electrodes are electrically connected together.

[0013] In an embodiment, each floating electrode is electrically isolated from the other floating electrodes.

[0014] In an embodiment, each floating electrode has an outer portion overlapping the first actuation electrode and an inner portion overlapping the second actuation electrode, and wherein the inner portions of the floating electrodes are staggered so that the inner portion of at least one floating electrode is positioned inwardly of the inner portion of at least one other floating electrode.

[0015] In an embodiment, the first actuation electrode and the second actuation electrode are provided in a common layer.

[0016] In an embodiment, the MEMS device comprises a plurality of floating electrodes provided in a common layer.

[0017] In an embodiment, the device is configured so that application of a potential difference between the first and second actuation electrodes causes a central region of the suspended membrane to move towards the substrate and remain substantially parallel to the substrate during the movement.

[0018] In an embodiment, the device is configured such that a separation gap between the suspended membrane and the substrate is dependent on a magnitude of the potential difference applied between the first and second actuation electrodes.

[0019] In an embodiment, the separation gap has a maximum distance, wherein the maximum distance is 12,000 nm or less.

[0020] In accordance with a second aspect of the present disclosure, there is provided an optical device comprising the MEMS device of any one of the preceding claims and an optical component attached to one or both of the substrate and the membrane.

[0021] In an embodiment, a first mirror or reflector is mounted to the membrane and a second mirror or reflector is mounted to the substrate.

[0022] In accordance with a second aspect of the present disclosure, there is provided an optical spectrometer for measuring the optical spectrum of light, the spectrometer comprising an optical filter including the MEMS device of any of the preceding claims, a first mirror or reflector mounted to substrate and a second mirror or reflector mounted to the membrane.

[0023] In accordance with a third aspect of the present disclosure, there is provided a method of operating the MEMS device of the first aspect, comprising applying a potential difference between the first actuator and the second actuator in order to cause the membrane to move relative to the substrate.

[0024] In accordance with a fourth aspect of the present disclosure, there is provided a method of manufacturing the MEMS device according to the first aspect, the method comprising forming a first actuation electrode and a second actuation electrode over a substrate forming a pair of terminals through which a voltage can be applied between the first and second actuation electrodes, forming a support structure for supporting a suspended membrane, the support structure including a first support and a second support, and forming a suspended membrane supported by the first support and second support over the substrate and forming a floating electrode supported by the suspended membrane, wherein the floating electrode overlies at least a part of the first actuation electrode and the second actuation electrode.

[0025] In an embodiment, the method comprises forming a sacrificial layer over the substrate and the first and second actuation electrodes, removing at least part of the sacrificial layer to form a gap between the membrane and the substrate, such that the membrane is suspended over the gap between a first fixed support and the second fixed support.

[0026] Further aspects and features of the present disclosure are provided in the following description and the appended claims.

[0027] BRIEF DESCRIPTION OF DRAWINGS

[0028] Examples of the present disclosure will now be described, by way of nonlimiting examples, with reference to the accompanying drawings, in which:

[0029] Figures 1 A to 1C show an example of a conventional MEMS cantilever-type device;

[0030] Figures 2A-2C show a MEMS device according to a first example of the present disclosure;

[0031] Figure 3 is a graph showing a relationship between displacement and actuation voltage for the MEMS devices of Figures 2A-2C, where a first capacitance Ci is equal to a second capacitance C2;

[0032] Figures 4A-4C show a MEMS device having a dielectric layer according to a second example of the present disclosure;

[0033] Figure 5 is a graph showing a relationship between displacement and actuation voltage for the MEMS device of Figure 4 A for several different area ratios a = —;A

[0034] Figures 6 is a schematic diagram of a MEMS device having a cantilever arrangement according to a third example of the present disclosure;

[0035] Figures 7A-7C show a MEMS device having a cantilever arrangement according to the third example of the present disclosure;

[0036] Figure 8 shows an example of an optical spectrometer device including a MEMS device according to a fourth example of the present disclosure;

[0037] Figures 9A-9C show an example MEMS device according to a fifth example of the present disclosure;

[0038] Figure 9D shows an example of deflection of part of a membrane and a floating electrode of the MEMS device of Figure 9A to 9C;

[0039] Figure 9E is a cross sectional view of the MEMS device shown in Figure 9A, the view showing an example of a tilting instability phenomenon;

[0040] Figure 10A shows a plan view of a first actuation electrode and a second actuation electrode of the MEMS device shown in Figure 9A according to an example;

[0041] Figures 10B to 10D show plan views of alternative designs of first and second actuation electrodes;

[0042] Figure 11 is a graph showing a relationship between applied actuation voltage and suspended membrane displacement for the MEMS device of Figures 9A to 9C;

[0043] Figures 12A to 12D show an example MEMS device that has split floating electrodes according to a sixth example of the present disclosure;

[0044] Figure 13 shows equivalent circuits to the MEMS device of Figures 12A to 12D;

[0045] Figure 14 is a graph showing a relationship between applied actuation voltage and suspended membrane displacement for the MEMS device of Figures 12A to 12D;

[0046] Figures 15A to 15D show an example MEMS device having split floating electrodes that are staggered or of different area (and hence capacitance) according to a seventh example of the present disclosure;

[0047] Figure 16 is a graph showing a relationship between applied actuation voltage and suspended membrane displacement for the MEMS device of Figures 15A to 15D;

[0048] Figure 17 shows an example method of manufacturing a MEMS device according to the present disclosure; and

[0049] Figure 18 shows a pictorial example of a method of manufacturing a MEMS device according to the present disclosure.

[0050] DESCRIPTION OF EMBODIMENTS

[0051] Throughout this specification the word "comprise", or variations such as "comprises" or "comprising", will be understood to imply the inclusion of a stated element, integer or step, or group of elements, integers or steps, but not the exclusion of any other element, integer or step, or group of elements, integers or steps. The terms "includes" means includes but not limited to, and the term "including" means including but not limited to. The term "based on" means based at least in part on. The term "number" means any natural number equal to or greater than one. The terms "a" and "an" are intended to denote at least one of a particular element.

[0052] The term micro-electromechanical systems (MEMS) device refers to a microscopic device which incorporates both electrical parts and moving mechanical parts. In some examples, the MEMS devices described herein may have a size of between 10 micrometres to 5 millimetres.

[0053] The term “suspended” refers to a part which is mechanically suspended over another part such that there is a separation gap between the two parts. The separation gap will be at least partially empty (in that it is not filled by solid material), such as an air gap or vacuum, which allows movement of the suspended part towards the part over which it is suspended. The term “actuation electrode” refers to an electrode to which an actuation voltage is applied to actuate movement of part of the MEMS device. The term “floating electrode” refers to an electrode that is electrically floating in the sense that it is not connected by an electrical conductor to the actuation electrode(s) or any other part driven by an external voltage.

[0054] Figure 1 A shows an example of a conventional micro-electromechanical systems (MEMS) device 1. The MEMS device 1 comprises a first electrode 10 and a second electrode 20. The first electrode 10 is fixed and is disposed over a substrate 30.The second electrode 20 is suspended over the first electrode 10 such that there is a separation gap between the first and second electrodes. The second electrode 20 may be referred to as a suspended electrode. The second electrode 20 may be supported by a suspension structure. The second electrode 20 is movable relative to the first electrode 10, and the movement may be facilitated by providing flexibility in the suspension structure and / or flexibility in the suspended electrode itself. The dotted lines 22 show an initial position of the second electrode 20 in which there is a separation distance d between the first electrode 10 and the second electrode 20. The full lines 24 show a subsequent position of the second electrode 20 after the MEMS device 1 has been actuated and the second electrode 20 has moved towards the first electrode 10 by a distance Ax, so that the separation distance between the two electrodes is d - Ax = x.

[0055] In the initial position, the second (suspended) electrode 20 extends substantially parallel to the first (fixed) electrode 10 so that the electrodes 10, 20 form a parallel plate capacitor. The first and second electrodes 10, 20 are electrically connected to a voltage source 40, and the MEMS device 1 is activated by applying a voltage between the first electrode 10 and the second electrode 20. The actuation voltage causes a first charge to accumulate on the first electrode 10 and an opposite second charge to accumulate on the second electrode 20. The actuation voltage may be direct current (DC) or alternating current (AC). Where the actuation voltage is an AC voltage, the frequency should be greater than, and preferably significantly greater than, the mechanical resonant frequency of the MEMS structure.

[0056] The accumulation of opposite charges on the first and second electrodes 10, 20 generates an electrostatic attraction between the electrodes, which causes the second electrode 20 to move towards the first electrode 10 as shown by the arrow Ax. In this way, the MEMS device 1 acts as an electrostatic actuator which varies the separation between two parallel plates.

[0057] Figure IB shows a conventional cantilever structure for realizing the MEMS device 1 of Figure 1A. The first electrode 10, also referred to as the fixed electrode is supported by a substrate 30. The second electrode 20, also referred to as the suspendedelectrode, is suspended over the first electrode 10 by a support structure 50, such as a dielectric cantilever as shown in Figure IB. The dielectric cantilever may comprise a first portion 51 extending along a surface of the substrate 30, a second (sidewall) portion 52 extending vertically upwards from the substrate and a third portion 53 extending out over a gap d between the second electrode 20 and the first electrode 10. The second electrode 20 may be deposited on the support structure 50 such that it extends up the vertical sidewall 52 as shown in Figure IB. For example, the second electrode 20 may be deposited by a physical vapor deposition (PVD) process.

[0058] The type of MEMS device shown in Figures 1 A and IB can be difficult to manufacture, because metal deposition on the vertical side-wall is difficult and unpredictable. In many cases a discontinuity or crack 60 will form in the second electrode 20, or electrical connections leading to the second electrode, as shown in Figure 1C. If there is a discontinuity 60 in the second electrode 20, then the MEMS device 1 will not work in a predictable way as the resistance of the second electrode 20 becomes highly variable. Even if there is no discontinuity, there are still likely to be variations in the thickness of the vertical deposition which causes the resistance to be variable. Therefore, the impedance characteristic of this type of MEMS device, both in DC and AC forms, is typically variable on any given fabrication-run. As a result, due to the significant variations and unpredictability of the impedance characteristics, it is difficult to develop a reliable consistent performance drive circuit for such a MEMS device.

[0059] The above issue is not limited to cantilever structures, but is a general problem when implementing MEMS devices of the type shown in Figure 1 A. As the first electrode 10 and second electrode 20 are on different layers and at different heights relative to each other, creating an electrical connection to a suspended electrode so as to apply a voltage to the second electrode generally involves metal deposition on a vertical side wall, which as explained above can result in variable track resistance.

[0060] The present application proposes a MEMS device structure in which a suspended electrode is electrically floating and mechanically suspended over a pair ofactuation electrodes. By applying a voltage between the actuation electrodes, corresponding opposing charges are induced in the suspended electrode and electrostatic attractive forces are generated between the actuation electrodes and the suspended electrode which causes the suspended electrode to move towards the actuation electrodes.

[0061] In this way, electrostatically actuated movement between upper and lower electrodes can be achieved without the need to create an electrode portion on a vertical sidewall of the device that provides a voltage to the upper electrode. This simplifies construction of the device and may provide a MEMS device with more consistent and predictable impedance characteristics.

[0062] Figures 2A and 2B show an example of a MEMS device 100 according to this structure. The device 100 comprises first and second actuation electrodes 110, 111 supported by a substrate 130, and a suspended electrode 120 suspended over the substrate 130. The suspended electrode 120 overlies at least a portion of the first actuation electrode 110 and at least a portion of the second actuation electrode 111. The suspended electrode 120 is electrically floating and is therefore electrically unconnected to ground or any voltage source. The first actuation electrode 110 and the second actuation electrode 111 are electrically connected during use to a voltage source 140, as shown in Figure 2B. A potential difference or actuation voltage may be applied between the first actuation electrode 110 and the second actuation electrode 111 in order to actuate the device 100.

[0063] Figure 2A shows an initial position of the floating electrode 120 in which no actuation voltage is applied between the first and second actuation electrodes 110, 111. Figure 2B shows the first portion of the floating electrode 120 in broken lines 121 and a second position of the floating electrode 120 in full lines 123 after a potential difference is applied between the first actuation electrode 110 and the second actuation electrode 111. As shown in Figure 2B, the applied potential difference creates a first charge on a surface of the first actuation electrode 110, a second charge on a surface of the second actuation electrode 111 and opposing induced charges on first and secondregions 122, 124 of the suspended electrode 120 which overlap the first and second actuation electrodes 110, 111 respectively. This arrangement creates an attractive force between each charged region of the suspended electrode 120 and the corresponding first and second actuation electrodes 110, 111. The first region 122 is defined as the area of the floating electrode 120 which overlaps (in this case directly overlies) the first actuation electrode 110 and the second region 124 is defined as the area of the floating electrode 120 which overlaps the second actuation electrode 111.

[0064] In Fig. 2C, the actuation electrodes 110, 111 have the same depth as the suspended electrode 120. Therefore in Figs. 2A-2C, the area of each actuation electrode 110, 11 lis equal to the area of the corresponding overlapping region 122, 124 of the floating electrode above. That is the area of the first overlapping region 122 is equal to the area of the first actuation electrode 110, and the area of the second overlapping region 124 is equal to the area of the second actuation electrode 111. However, in other implementations this may not be the case. It will however be understood that where the area of the actuation electrode 110, 111 differs from the area of the overlapping region 122, 124, the capacitance and resulting force on the floating electrode 120 depends upon the size of the overlapping area, rather than on the area of the actuation electrode. For simplicity, in the present specification we refer to the area of the actuation electrode 110, 111 as a shorthand for the area of the overlapping region of the floating electrode 120. Therefore, unless context demands otherwise, references to an area Ai of the first actuation electrode 110, an area A2 of the second actuation electrode 111 and the like, should be understood as referring to the area of the respective overlapping region 122, 124 of the floating electrode.

[0065] In the example of Figure 2B, the first actuation electrode 110 has a positive charge which induces a negative charge on the first surface region 122 of the suspended electrode 120, and the second actuation electrode 111 has a negative charge which induces a positive charge on the second surface region 124 of the suspended electrode 120. However, in other examples, the polarities could be reversed. As the surfaces of the actuation electrodes induce opposite charges on the suspended electrode, this creates an attractive force and causes the suspended electrode 120 to electrostaticallymove towards the actuation electrodes 110, 111 and the substrate 130. The independence of drive polarity means that the actuation electrodes could also be driven with an alternating-current (AC) voltage to achieve electrostatic attraction between the suspended electrode and the actuation electrodes. As the suspended electrode 120 is electrically floating and electrically unconnected to the actuation electrodes below, manufacture of the MEMS device is simplified and reliability may be improved compared to the design of Figure 1 A.

[0066] As shown in Fig 2B, each actuation electrode 110, 111 and the respective charged suspended electrode region 122, 124 above it form a respective capacitor. Thus the first actuation electrode 110 and first region 122 form a first capacitor Ci and the second actuation electrode 111 and the second region 124 form a second capacitor C2. Ci and C2 are connected in series across the applied voltage source 140. Each capacitance Ci, C2 will be directly proportional to the area (Al, A2) of the associated actuation electrode 110, 111 and inversely proportional to the distance (di, d2) between the corresponding actuation electrode 110, 111 and the charged region 122, 124 of the suspended electrode 120. When the suspended electrode 120 is flat and parallel to the substrate 130, and the two actuation electrodes 110, 111 are disposed in the same plane, the two distances di and d2 are the same. If the areas Al, A2 are also the same, the two capacitances are also the same (Ci = C2), and the voltage across Ci is the same as the voltage across C2 (VI = V2). With Al = A2, the force on the two regions 122, 124 of the suspended electrode will be the same (Fl = F2). In this case, the applied voltage for a given displacement will be double that required for a parallel-plate structure of the type shown in Figure 1 A.

[0067] Figure 3 shows a relationship between applied voltage and displacement of the suspended electrode 20, 120 for the arrangements shown in Figures 2A to 2C and 1 A to 1C. As shown, snap-down (discussed in more detail below) occurs for both devices at the same value of normalized displacement. However, with the present arrangement shown in Figures 2A to 2C, if the floating electrode 120 becomes tilted, the two distances di and d2 will no longer be the same. For example if di were less than d2, Ciwould become greater than C2, and the voltage division in the series capacitancecircuit would result in V2 being greater than VI, which would therefore result in the force F2 being greater than Fl. This force difference would serve to reduce d2 more than di, thereby compensating for the tilt in the suspended electrode. In many MEMS devices, suspended electrode and membrane tilt is a key issue limiting performance. As such, the architecture shown in Figures 2 A to 2C will improve performance by compensating for tilt in the suspended structure.

[0068] Figure 2C shows one example implementation of the design of the MEMS device 100 of Figures 2A and 2B. The MEMS device 100 of Figure 2C has a suspended membrane 150 suspended over the substrate 130 between a first support 160 and a second support 170. The first support 160 and second support 170 may be separate pieces or different parts of the same piece. For instance, in some examples the first support 160 and second support 170 may be different parts of a wall which surrounds a central part over which the electrode 120 is suspended. The electrically floating electrode 120 is supported by the suspended membrane 150 and overlies at least a portion of both the first and second actuation electrodes 110, 111. The floating electrode 120 may for example be attached to the top of the suspended membrane 150 (as shown in Figure 1C), attached to the bottom of the suspended membrane or embedded in the suspended membrane.

[0069] The suspended membrane 150 may be formed of a dielectric or electrically insulating material. The actuation electrodes 110, 111 and suspended electrode 120 may be formed of a conductive material, such as a metal or semiconductor. The first and second actuation electrodes 110, 111 may be deposited directly on the substrate 130, embedded in or deposited on a layer overlying the substrate 130, or sandwiched between the substrate 130 and one or more overlying layer(s). The first and second actuation electrodes 110, 111 may be in the same single layer of the MEMS device, as shown in Figures 2A-2C, or may be provided in different layers. A lateral gap g must however exist between the first and second actuation electrodes 110, 111 as shown in Figure 2C.

[0070] The first and second supports 160, 170 are support structures which suspend the suspended membrane 150 over the substrate 130 between a first anchor point 160A and a second anchor point 170A. The first and second supports 160, 170 may be fixed in position relative to the substrate 130. In the example of Figure 2C, the first and second supports 160, 170 are support posts extending between the substrate 130 and the suspended membrane 150. In other examples, the supports may take the form of other structures and may support the suspended membrane 150 from above, or from the sides, instead of from below. The supports 160, 170 need not extend vertically and may include any part external of the suspended membrane that is attached to and supports the suspended membrane 150 over the substrate 130.

[0071] In some examples, there may be more than two supports. For instance, there may be a plurality of supports disposed circumferentially of the suspended membrane 150. In one example, the suspended membrane 150 may have a square, rectangular or other shape, and there may be a support attached to each corner of the suspended membrane 150. In some examples, the first and second supports may be different parts of the same support structure. For example, the support structure may include a square, rectangular, circular or circumferential wall surrounding a central region over which the suspended membrane 150 extends.

[0072] While the example in Figures 2A-2C includes one floating electrode 120, in other examples there may be more than one floating electrode 120. In such cases, each floating electrode 120 is supported by the suspended membrane 150, is positioned to overlie at least a portion of both a first and second actuation electrode 110, 111, and is electrically unconnected to the first and second actuation electrodes 110, 111. In some examples, each floating electrode 120 is confined to a single layer of the MEMS device, which simplifies fabrication.

[0073] The first and second actuation electrodes 110, 111 and the one or more floating electrodes 120 together form an actuation mechanism for moving the suspended membrane 150 relative to the substrate 130. The MEMS device 100 may have terminals for applying a potential difference between the first and secondactuation electrodes 110, 111. For example, the terminals may comprise contact pads disposed near an edge of the device 100 which are electrically connected to the first and second actuation electrodes 110, 111 and a voltage source 140.

[0074] For the MEMS device 100 of Figure 1 A, the actuation electrode 10 and the suspended electrode 20 form a single parallel plate capacitor. For the MEMS device of Figures 2A-2C, the two actuation electrodes 110, 111 and the suspended electrode 120 may be considered as two capacitors Ci, C2 connected in series, as shown in Figure 2B. As the two capacitors are in series, the applied actuation voltage Vsis divided between the two capacitors Ci, C2in inverse proportion to their respective capacitance. For example, if the two capacitors have the same capacitance (for example because the first and second actuation electrodes 110, 111 have same surface area), then when 100 V is applied across the two actuation electrodes 110, 111, 50 V is applied across each of the capacitors Ci, C2. Consequently, the voltage required for actuation of the floating electrode MEMS actuator of Figures 2A-C will be higher than for an equivalent area parallel plate MEMS actuator having the design of Figure 1 A.

[0075] The MEMS device 100 may be configured such that the separation gap d between the suspended membrane 150 and the substrate 130 is adjustable in a controlled manner through a plurality of increments Ax by varying the potential difference between the first and second actuation electrodes 110, 111. In some examples, by adjusting the potential difference the suspended membrane is controllable to move in increments of less than lOnm or less than lOOOnm. For instance, the suspended membrane 150 may be movable gradually over an actuation range as the voltage Vsis increased, rather than switching between binary on and off positions in a switch like manner. In this way, the MEMS device 100 may be used in optical and other applications which require the separation distance d to be varied in a gradual manner. For instance, an optical component may be mounted to one or both of the substrate 130 and the membrane 150. In this way, the MEMS device 100 may vary the position and / or orientation of the optical component, which is useful for a variety of applications, including but not limited to optical filters.

[0076] The actuation range of electrostatic MEMS devices is limited by a phenomenon referred to as snap-down (or pull-in), which causes the suspended electrode to rapidly snap down and contact the substrate or actuation electrode after a snap-down (or pull-in) separation distance limit is reached. Once the separation distance between the suspended and actuation electrode reduces to less than the snapdown (or pull-in) limit, the attractive electrostatic force between the suspended and actuation electrodes rapidly overwhelms the mechanical spring-restoring force of the suspension structure causing the suspended electrode to rapidly snap down.

[0077] This is because the electrostatic attraction force is proportional to l / (d-Ax)2while the mechanical spring-restoring force is proportional to k Ax (where d is the initial separation between the suspended and actuation electrode, Ax is the displacement of the suspended electrode away from an initial position towards the actuation electrode, and k is the spring constant of the suspension structure). Accordingly, the electrostatic attraction increases in proportion to the square of the deflection Ax, while the countering mechanical force increases only linearly, and consequently after the snap-down separation distance limit is reached the MEMS device becomes unstable and the suspended electrode ‘snaps down’ towards the actuation electrode.

[0078] It is desirable to accurately predict the snap-down separation distance limit, because the behavior of the device becomes unstable beyond this. The behavior of the device may be modelled by the following equation.(Equation 1) where Vsis the applied actuation voltage, d is the initial separation distance between the actuation electrode(s) and the suspended electrode, Ax the displacement of the suspended electrode towards the actuation electrode, k is the spring constant, Ai is the area of the first actuation electrode, A2 is the area of the second actuation electrode; andso is the dielectric constant of the empty space between the suspended and actuation electrodes.This equation assumes that in a stable region before snap down, the electrostatic attraction balances the spring restoring force, that the lateral gap g between the first and second actuation electrodes is negligibly small and the area A of the suspended electrode is approximately equal to the sum of the areas of the first and second actuation electrodes.

[0079] For scenarios where the actuation electrodes have different areas, factors n and m are included, as follows:(Equation 2)

[0080] Taking into account the factors n and m, Equation 1 can be re-written as:(Equation 3)

[0081] Therefore, the relationship between applied voltage Vsand deflection (displacement) Ax of the suspended electrode is given by:(Equation 4)

[0082] When the actuation electrodes have equal areas (m=l, n=2), this can be rewritten as:(Equation 5) k

[0083] Using y = — , this can be simplified to:(Equation 6)

[0084] Figure 3 A is a graph 200 showing a relationship between normalized displacement (Ax / d) and actuation voltage Vsfor the device 100 of Figures 2A-2C according to the above model when Ci = C2 (202), compared to the conventional device of Figure 1 A (204). It can be seen that for both devices 1, 100, the displacement varies non-linearly in the stable region before the snap-down voltage 206 is reached. Once the snap-down voltage is reached, the device becomes unstable and the suspended electrode snaps down onto the actuation electrode. The snap down voltage 206 corresponds to the maximum voltage shown on the graph 200 and the snap down limit 208 (maximum displacement before snap down) is the normalised displacement corresponding to the snap down voltage. Beyond this voltage, the device 1, 100 becomes unstable and the suspended electrode will snap down reducing the separation distance to zero.

[0085] It can be shown that the stable range of motion for both the floating electrode device (Figs 2A-2C) and the conventional design device (Fig 1 A) is up to one third of the vertical separation distance d. Thus, the snap down limit, or maximum displacement Ax before snap down is as follows:(Equation 7)For a floating electrode actuator device having first and second actuator electrodes of equal area (m=l, AI=A2 =A / 2), the snap down voltage is as follows:(Equation 8)

[0086] It will also be noted that the voltage required to achieve a given displacement for the floating electrode actuator device of Figure 2A is approximately double that for the conventional actuator device of Figure 1 A. This is due to the actuation voltage Vsbeing split equally between the two capacitors Ci, C2in the floating electrode design when Ci = C2 (see Figure 2B).

[0087] For different values of m (wherein m =1 indicates equal areas of first and second actuation electrodes AI=A2, m=2 indicates that the second actuation electrode has double the area of the first actuation electrode A2= 2*Ai, and m=3 indicates that the second actuation electrode has triple the area of the first actuation electrode A2= 3*Ai, and so on), there will be different forces induced on different areas of the suspended structure depending on the area ratio, and hence capacitance ratio. This phenomenon can be used to avoid tilting of a membrane during use.

[0088] It would be desirable to reduce the actuation voltage Vsneeded to displace the floating electrode in order to reduce the MEMS device power requirements. Figures 4A-4C show another example of a MEMS device 300 according to the present disclosure. The device 300 is similar to the device 100 of Figures 2A-2C, except that a solid dielectric layer 180 is positioned between the first actuation electrode 110 and the suspended membrane 150 and / or floating electrode 120. This forms a structure with a constant capacitance Ci in series with a variable capacitance C2. Including a dielectric layer 180 increases the capacitance Ci between the first actuation electrode 110 and the floating electrode 120, which has the effect of reducing the total actuation voltage Vsneeded to move the suspended portion of the floating electrode in comparison to that required for actuation when Ci = C2.

[0089] The dielectric layer 180 may take the form of a support post for supporting the suspended membrane 150. In the example of Figures 4A-4C, the support post has a dielectric constant £xand occupies a space between the first actuation electrode 110 and the suspended electrode 120, while a space between the second actuation electrode 111 and the suspended electrode 120 is empty (e.g. an air gap) and has a dielectric constant £0. The part of the suspended structure over the dielectric post 180 is not movable, and therefore both the separation gap d1and the capacitance between first actuation electrode 110 and the floating electrode 120 remains constant. In contrast, the separation gap d2between the suspended portion of the floating electrode 120 and the second actuation electrode I l l is variable.

[0090] As shown by the equivalent circuit in Figure 4B, the above arrangement may be considered equivalent to two capacitors Ci , C2 in series, with the applied actuation voltage Vs divided between them in inverse proportion to their capacitance. If the first and second actuation electrodes 110, 111 have the same area, the initial voltage induced across each of the series-connected capacitors then depends only on the values of the dielectric constants. Therefore, in the limiting case of»0, it can be deduced that(where V5is the applied actuation voltage, Vi is the initial voltage between the first actuation electrode 110 and the suspended electrode 120, and V2 is the initial voltage between the second actuation electrode 111 and the suspended electrode 120). Consequently, a structure with a dielectric layer 180 between one (but not both) of the first and second actuation electrodes 110, 111 and the suspended electrode 120 reduces the total applied actuation voltage (Vs) needed to deflect the suspended electrode 120 compared to the architecture of Figure 2 that contains no fixed capacitance.

[0091] In the standard actuation architecture of Figure 1 A, before snap-down, any incremental increase in the actuation voltage Vsresults in a change in Ax, which in turn results in an increase in the electric field in the gap, and an increase in the force on thesuspended electrode. However, the change in Ax also results in an increase in the mechanical restoring force, which balances the electrostatic force to hold the suspended electrode position stable. Snap-down is reached when the displacement Ax reaches about 1 / 3 of the original gap d. At snap-down, any incremental increase in the actuation voltage Vsresults in an increase in Ax that in turn increases the electrostatic force on the suspended electrode by a greater amount than the increase in the mechanical restoring force. This results in a run-away collapse of the suspended electrode Ax towards the fixed electrode. In contrast, in the architecture in Figure 4A, an incremental increase in Vsresults in an increase in Ax which in turn redistributes Vsin a greater proportion across the fixed capacitor Ci, thereby allowing the mechanical restoring force to hold the device in balance beyond the point of Ax=d / 3, and delaying the runaway collapse of the suspended electrode towards the fixed electrode. That is, the fixed dielectric post 180 serves to also increase the stable range of the device, i.e. the maximum possible displacement before snap down.

[0092] A MEMs device design having a dielectric 180 between the first actuation electrode and the overlapping region of the floating electrode, for instance as shown in Figs 4A-4C, can be referred to as a dielectric post design. It has been found that for a dielectric post design MEMS device, as the ratio of the second actuation electrode area to the first actuation electrode area — = n increases and the ratio of the dielectricconstants — = m increases, the capacitance ratio between the first and second £0 capacitors Ci and C2 (see Figs 4 A and 4B) increases. The voltage needed to reach a given deflection Ax also decreases as this capacitance ratio increases, but reaches a lower bound in the limit of Ci » C2. The behaviour of the MEMS device of Figures 4A-4C may be modelled by the following equation: (Equation 9)Where Vsis the actuation voltage, d the separation distance, Ax the deflection, y = — — l, n is the ratio A2 / A1 and m is the ratio — . Note that n and m denote different constants £0in Equation 9 compared to Equations 2-4. In Equation 9 and the following discussion, n = — and refers to the area ratio of the second to first actuation electrodes, while m = — and refers to the dielectric ratio of the first and second capacitors formed by the £0 first and second actuation electrodes and the suspended electrode.

[0093] Figure 5 is a graph 500 showing a relationship between normalised displacement and actuation voltage for dielectric layer MEMS devices of the type shown in Figures 4A-4C, but with different values of a . Note in the graph of Fig. 5, a is used instead of n to represent — and T] is used instead of m to represent — . A higher £0 value of m thus denotes a MEMS device having a dielectric layer with a higher dielectric constant. It can be seen that as a increases, the actuation voltage before snap down increases and the displacement limit increases. Thus by providing a dielectric layer between one of the actuation electrodes and the suspended electrode and / or by varying the ratio of the electrode overlapping areas in Fig 4C it is possible to increase the displacement limit.

[0094] Figure 6 is a schematic view of a MEMS device which has a cantilever arrangement. A floating suspended electrode 620 is suspended over first and second actuation electrodes 610, 611 which are supported by a substrate 630. However, unlike the MEMS devices of Figures 2 to 5, only one end 621 of the suspended electrode is fixed; the other end 622 is free to move. Figure 7A is a perspective view, Figure 7B a front view and Figure 7C a top down plan view of one example structure for implementing the MEMS device of Figure 6.

[0095] The behaviour of the cantilever arrangement can be modelled using a bisection model in which the cantilever suspended electrode 620 has a length L that is split into three parts, each with length L / 3 and its own respective capacitance Ci, C2 and C3, as shown in Figure 6. In this model, the first third of the suspended electrode to a pivot point 625 is fixed, and the rest of the electrode 620 tilts relative to the first third when it deflects towards the second actuation electrode 611 and the substrate 630. The capacitances in this model are given by:L(Equation 10)

[0096] where W is the width of the electrodes, L is the length of the suspended electrode, and L»Ax. LI and L2 are the lengths of the first actuation electrode 610 and second actuation electrode 611, respectively. The separation distance between the two actuation electrodes is denoted by g.

[0097] It can be shown that the maximum actuation voltage Vsbefore snap down according to this bi-section model is:(Equation 11)Predictions from the bi-section model were compared with an Ansys simulation for a cantilever arrangement with fixed post dielectric and found to be good in agreement as shown in Table 1 below.Cantilever Vmax (V) Vmax (V) Difference Difference (%) length (jim) from bisection from Ansys between model bisection model and Ansys simulation (V)150 8.81 8.73 0.08 0.92200 4.95 4.87 0.08 1.62250 3.17 3.11 0.06 1.89300 2.20 2.15 0.05 2.27350 1.62 1.57 0.05 2.53Table 1

[0098] It is possible to implement a MEMS device that has a cantilever arrangement with an electrode suspended over a pair of actuation electrodes, and with one end fixed and the other end free, and an example such device is shown in Figures 7A to 7C. Figures 7A to 7C show a structure that includes anchors 720, for example formed of silicon nitride, and a cantilevered suspended electrode 722 disposed over first and second actuation electrodes 724, 726. The electrodes may be formed of gold. The above model is also helpful to more accurately predict the behaviour of an arrangement in which a floating electrode is supported by a membrane suspended between a first support and a second support, but in which the floating electrode does not extend all the way to the second support.

[0099] One example of such an arrangement is a MEMS device 800 shown in Figure 8. In Figure 8 a suspended membrane 850 extends between a first support 860 and a second support 870. The suspended membrane 850 has an inner region 851 which does not have any floating electrodes. First and second floating electrodes 820, 822 are positioned on the suspended membrane 850 next to the inner region 851 such that the floating electrodes 820, 822 extend along outer regions 852, 854 of the suspended membrane 850. This arrangement may be modelled as a pair of cantilevers on sides of the inner region 851 of the membrane.

[0100] The inner region 851 is located inwardly of first and second supports 860, 870 and may be centrally disposed relative to the suspended membrane 850. This arrangement allows an electrical, optical or other component to be attached to the inner region without interfering with the floating electrodes 820, 822. In some examples, the inner region may have an area of 1 square cm or less.

[0101] In the example shown in Figure 8, a first optical component 801, in this example in the form of a first mirror, is mounted to the suspended membrane 850 and a second optical component 802, in this example in the form of a second mirror, is mounted to the substrate 830. In other examples, other types of optical component or electrical components could be mounted to the inner region of the suspended membrane 850 and / or the substrate 830. While the first optical component 801 isshown mounted above the suspended membrane 850, in other examples it could be mounted below the membrane 850 or embedded in the membrane 850, or have the membrane 850 embedded within it, or it could simply consist of the membrane 850.

[0102] The mirror arrangement of Figure 8 forms an optical filter (specifically a Fabry Perot filter), which may be used in an optical spectrometer. For example, it may be used in a short wave infrared spectrometer. However, it will be understood that the MEMS device is not limited to this application and may be used for other types of micro-optical and micro-electronic devices in which it is necessary to vary the separation between two parts in a controlled way.

[0103] An optical detector 803 may be provided under the second optical component 802. In the example of Figure 8, the optical detector 803 forms part of the MEMS structure and is disposed directly beneath the second optical component 802. However, in other examples there may be one or more transparent intermediate layers between the optical component 802 and the detector 803. In some examples, the MEMS device could be fabricated separately, then packaged with the detector 803 and slightly separated physically from the detector 803. For instance, the MEMS device could be fabricated on another transparent substrate, then flipped over so that it is face-to-face with the detector, or the detector 803 could be attached below the substrate 830.

[0104] The MEMS device 800 may be configured so that application of a potential difference between first and second actuation electrodes 810, 812 and 814, 816 causes the inner region 851 of the suspended membrane 850 to move towards the substrate 830, whilst remaining substantially flat / substantially parallel to the substrate 830 during the movement. Even if the movement of the suspended membrane 850 causes parts of the suspended membrane 850 between supports 860, 870 and the central region 851 to become tilted or angled towards the substrate 830, the central region 851 may remain substantially parallel to the substrate 830. In this way, the upper and lower mirrors 801, 802 in the example of Figure 8 remain parallel to each other. However, this structure may be used for many other applications where it is desired to vary the separation between two parallel members.

[0105] Keeping the inner region 851 of the suspended membrane 850 parallel to the substrate 830 during movement may be achieved by having a first floating electrode 820 extending over a first pair of actuation electrodes 810, 812 and a second floating electrode 822 extending over a second pair of actuation electrodes 814, 816, wherein the first and second floating electrodes 820, 822 extend over outer regions 852, 854 of the suspended membrane 850 and do not extend into the inner region 851 of the suspended membrane 850. A dielectric layer 860, 870 is positioned between each first actuation electrode 810, 814 and the corresponding region of the floating electrode 820, 822 above. The dielectric layer 860, 870 may serve as a dielectric support post.

[0106] While Figure 8 shows the first and second floating electrodes 820, 822 mounted to the bottom of the suspended membrane 850, in other examples the first and second floating electrodes 820, 822 may be mounted to the top of or embedded in the suspended membrane 850. Further, while first actuation electrodes 810, 814 are shown as two separate electrodes, they may in some cases form part of the same electrode, for instance they may be part of a ring or square shaped electrode surrounding the inner region 851. Likewise, the second actuation electrodes 820, 822 may be part of the same electrode, for instance they may be part of a ring or square shaped electrode which lies inside the first actuation electrode and surrounds an inner region of the substrate 830.

[0107] Figures 9 A to 9C show one possible implementation of a MEMS device 900 according to the present disclosure.

[0108] Figure 9A is a plan view from above, Figure 9B is cut-away perspective view showing half of the MEMS device, and Figure 9C is a cross-sectional view.

[0109] The MEMS device 900 comprises a substrate 930 and a suspended membrane 950 suspended over the substate 930 between a first support 960 and a second support 970. The substrate 930 may be formed of an insulating material such as, but not limited to, glass, sapphire or silicon. The suspended membrane 950 may be formed of an insulating material such as, but not limited to, silicon nitride or silicon. The supports 960, 970 are formed of a dielectric material, such as but not limited to polyimide PI-2611, silicon dioxide or silicon nitride. In one example, the dielectric material has a relative permittivity of 2.9; in other examples it may have a relative permittivity between 1 and 10, or much higher, for example up to 250 in the case of TiO2.

[0110] First actuation electrodes 910 and second actuation electrode 912 are supported by the substrate 930. The actuation electrodes may be formed of a conductive material such as, but not limited to, gold or aluminium.

[0111] Figures 10A to 10D show top down plan views of example structures of the first actuation electrode 910 and the second actuation electrode 912. The first actuation electrode 910 may be an outer actuation electrode and the second actuation electrode 912 may be an inner actuation electrode. The first actuation electrode 910 may comprise one or more electrodes which surround the second actuation electrode, and the second actuation electrode 912 may comprise one or more electrodes which surround an inner region of the substrate.

[0112] For example, in Figure 10A the first actuation electrode 910 surrounds the second actuation electrode 912 which surrounds an inner region of the substrate and both have a rectangular ring shape. Figure 10B is similar to Figure 10A, except that the second actuation electrode 912 is formed of a plurality of second actuation electrodes 912a, 912b, 912c and 912d which are electrically connected together.

[0113] Figure 10C is similar to Figure 10B, except that the first actuation electrode 910 comprises a plurality of first actuation electrodes 910a, 910b, 910c and 910d that are not electrically connected together and a plurality of second actuation electrodes 912a, 912b, 912c and 912d that are not electrically connected together. Figure 10D is similar to Figure 10 A, except that the first and second actuation electrodes 910, 912 have a circular ring shape. However, it will be understood that the first and second actuation electrodes 910, 912 may have any suitable shape. The first and / or second actuation electrode 910, 912 may comprise a plurality of unconnected independently actuated electrodes. However, in most cases each of the first and second actuation electrodes 910, 912 will comprise either a single electrode or a plurality of commonlydriven connected electrodes (as for example shown in Figure 10B) as this will reduce the number of electrical connections to an external voltage source or common ground.

[0114] In some examples, the first actuation electrode 910 may generally surround the second actuation electrode 912 and the second actuation electrode 912 may generally surround an inner region of the substrate 930 which is aligned with an inner region of the suspended membrane. This helps to provide a more stable structure which helps to keep the inner region of the suspended membrane 950 parallel to the substrate 930 when the membrane 950 moves towards the substrate 930.

[0115] The first and second actuation electrodes 910, 912 may be deposited directly on the substrate 930 (as shown in Figure 9C), embedded in or deposited on a layer overlying the substrate 930, or sandwiched between the substrate 930 and one or more overlying layer(s). The first and second actuation electrodes 910, 912 may be in a same single layer of the MEMS device, as shown in Figure 9C, or may be provided in different layers.

[0116] The suspended membrane 950 supports floating electrodes that in this example include an outer floating electrode 920 and inner floating electrodes 920a, 920b, 920c, 920d connected together by relatively thin strips 921a, 921b, 921c, 92 Id. Each floating electrode overlies at least a portion of a pair of first and second actuation electrodes 910, 912 and is electrically unconnected to the first and second actuation electrodes 910, 912. The dielectric layer (post 160, 170) fills some or all of the gap between the first actuation electrode 910 and the suspended membrane or floating electrode as shown in Figure 9C. The support posts 960, 970 may be separate from each other or part of the same support structure (for example in the form of a ring shaped wall).

[0117] In the example shown in Figures 9A to 9D, the suspended membrane 950 has a plurality of sides and a corresponding inner floating electrode 920a, 920b, 920c, 920d associated with each side of the membrane. While Figures 9A and 9B show a square membrane with four opposing sides, other shapes with a different number of sides, or a curved or circular shape could be used. Having an inner floating electrode on each sidehelps to stabilize the membrane 950 and maintain an inner portion 951 of the membrane parallel to the substrate 930 as it moves towards the substrate 930 when the MEMS device 900 is actuated. While the floating electrodes are shown on top of the suspended membrane 950 in Figure 9C, in other examples they could be embedded in or mounted to the bottom of the suspended electrode.

[0118] The membrane 950 has an inner region 951 and the inner floating electrodes 920a, 920b, 920c, 920d are positioned around the inner region 951 which does not have any floating electrodes. The inner region 951of the suspended membrane 950 corresponds to an inner region 931 of the substrate around which the second actuation electrode 912 is arranged.

[0119] The plurality of floating electrodes may be electrically connected together. For example, in Figures 9A to 9C each inner floating electrode 920a, 920b, 920c, 920d is connected to a common conductive structure 920, in this case a square ring, by a respective strip 921a, 921b, 921c, 92 Id. The common conductive structure overlies the first actuation electrode 910, while the inner part 920a, 920b, 920c, 920d of each floating electrode extends over the second actuation electrode 912.

[0120] A voltage source 940 may be used to apply a potential difference between the first and second actuation electrodes 910, 912 of the MEMS device. The potential difference may be applied via a pair of terminals (not shown). This creates a first charge on a surface of the first actuation electrode 910 which induces an opposite charge on the region of the floating electrode overlapping the first actuation electrode (the common conductive structure 920). The applied voltage also creates a second charge (opposite to the first charge) on a surface of the second actuation electrode 912 and induces an opposite charge on the region of the floating electrode overlapping the second actuation electrode (each inner floating electrode 920a, 920b, 920c, 920d). This creates an attraction between the floating electrodes and the actuation electrodes 912 and causes the floating electrodes and suspended membrane 950 to move towards the substrate 930 and the actuation electrodes 912.

[0121] As shown in Figure 9D, once the MEMS device is activated, the outer floating electrode 920 will remain in place as movement is prevented by the support post 160, 170, and the initial separation distance d with the first actuation electrode 910 and constant capacitance is maintained. However, the inner floating electrodes 920a, 920b, 920c, 920d will tilt downwards towards the second actuation electrode 912. The separation distance decreases towards the inner floating electrodes, such that at an outer end 980 of the inner part 920a, the separation distance is di, and at an innermost end 982 of the inner part 920a, the separation distance is d2, wherein d2 is smaller than di.

[0122] Referring to Figure 9 A, the potential on each inner floating electrode 920a, 920b, 920c, 920d and the common conductive structure 920 is constant everywhere, as it is an equipotential surface. Referring to Figure 9D, as d2 is less than di and the electrostatic field is proportional to the potential difference (voltage) and inversely proportional to the distance between the floating and actuation electrode, the electric field (and the electrostatic attraction force) is higher at d2 than at di. This can lead to snap-down as the force at point d2 is greater than at di, which causes the innermost end 982 to move closer to the actuation electrode, thereby reducing d2 and further increasing the electrostatic force at the innermost end 982.

[0123] For Figures 9A to 9C, the capacitance ratio between the capacitors Ci , C2 formed by the floating electrode and the first and second actuation electrodes depends upon the area ratio of the second actuation electrode to the first actuation electrode and the dielectric material of the posts 960, 970. By selecting appropriate values, a desired capacitance ratio and thus increased deflection before snap-down can be achieved

[0124] As shown in Figures 9A and 9B, one or more stress release notches 990 may be provided in a part of the suspended membrane 950 between the support posts 960, 970 and an area overlying the second actuation electrode 912. Each notch 990 may overlap part of the second actuation electrode and / or may be positioned at least partly between respective inner floating electrodes 920a, 920b, 920c, 920d. While the notches 990 are shown as having a square shape in Figures 9A and 9B, they may have a different shape, depending on the layout of the supporting membrane 950 and thedielectric posts 960, 970. The notches 990 may for example be positioned at comers of the membrane 950. The notches 990 help to relieve strain on the membrane 950 caused by forces pulling in different directions. In the absence of the notches, when the suspended membrane 950 moves, it is at greater risk of developing buckling at the four corners.

[0125] A protype was constructed in which the floating electrodes had a size of 600 m x 1100 gm and a size of the inner central region was 800 gm x 800 gm. The inner central region was used as an optical area. The separation gap d between the bottom surface of the membrane and the bottom electrode was approximately 2 jim. The area ratio of second to first actuation electrode n was 10. The floating electrodes were embedded in the membrane (in this example made from silicon nitride) with 200 nm thickness of the membrane on top of the floating electrodes and a 200 nm thickness of the membrane below the floating electrodes. Embedding the floating electrode in this fashion helped to maintain flatness of the inner central region of the membrane.

[0126] Figure 11 is a graph 1100 showing a relationship between applied actuation voltage and displacement of the central region of a suspended membrane for the prototype MEMS device described above, and where the suspended membrane is formed of silicon nitride and the tensile stress of the membrane is 200 Mpa. As can be seen in Figure 11, for an input voltage (V= 15.4V AC), a maximum displacement of 0.77 microns was achieved. The snap down or pull-in limit for this design was therefore around 38.5% of the initial gap between the activation and floating electrodes, which is more than the conventional 1 / 3 gap of Figure 1.

[0127] For conventional MEMS devices, the rapid movement and impact of snap down can seriously damage the device. There is also a risk of upper and lower parts of the device sticking together due to the Van der Waals attractive force. However, for the devices of Figures 2, 4, 6 to 10 and 12 this risk was found to be significantly reduced to the extent that it is possible to re-set and then continue using the MEMS device even after snap down. It is believed that this is because the suspended floating electrodedesign avoids a "hard" snap-down, which makes it easier to recover from snap-down, and may reduce or avoid the need for anti-stiction bumps.

[0128] For a conventional parallel-plate actuator device 1 of the type shown in Figure 1 A, the voltage across the electrodes is fixed so the electric field between the electrodes continues to increase as the electrodes move towards each other, resulting in increasing force and acceleration, and a "hard" snap-down. However, with the suspended floating electrode design of the present arrangement, there are two capacitors in series and this can be used to reduce the force exerted during snap down. With an embodiment of the present arrangement, a first and constant capacitor Ci is formed by the first actuation electrode 910 and a region of the floating electrode 920 overlapping the first actuation electrode, and a second variable capacitor C2 is formed by the second actuation electrode 912 and an inner floating electrode 920a overlapping the first actuation electrode 910 (see e.g. Figures 2B and 4B). As the suspended and actuation electrodes move closer together, the capacitance of the second capacitor C2 increases due to the reduced separation distance, but the capacitance of the first capacitor Ci remains the same (Figure 4B). Since the actuation voltage splits between the series capacitors in inverse proportion to their capacitance, as the capacitance of the second capacitor C2 increases, the voltage across the second capacitor decreases as a greater portion of the applied voltage moves to the first capacitor Ci. As the voltage across the second capacitor C2 decreases, this leads to a decreasing electric field and a force reduction on the part of the suspended electrode that is disposed over the second actuation electrode. Eventually, when the suspended electrode and second actuation electrode are very close and almost touching, the capacitance is so high that the induced voltage is almost zero, resulting in a "soft" snap-down. This effect makes the MEMS devices as disclosed herein robust and less vulnerable to damage due to snap down.

[0129] For a protype according to Figures 9 A to 9C and Figures 10A to 10D, it was found that after displacement of the floating electrodes beyond one third of the separation gap d, the structure became prone to tilting instabilities due to small fabrication imperfections. A consequence of this is that one side of the membrane tilts down slightly more, which can lead to increasing attraction and snap down on that side.Once a side snaps down, the rest of membrane also snaps down. Figure 9E is a cross section from left to right of Figure 9 A which illustrates the tilting instability phenomenon.

[0130] Referring to Figure 9E it can be seen that the suspended membrane 950 is tilted downwards to the right hand side. Therefore the second floating electrode 920b is closer to second actuation electrode 912 on the right side (separation distance d2) than the first floating electrode 920a is to the second actuation electrode 912 on the left side (separation distance di). As mentioned above, this can occur due to minor imperfections in the membrane 950, electrodes, substrate 930, or parts of the suspension structure. As the electrostatic force increases with decreasing separation distance, this can lead to run-away attraction and snap-down on one side. In other words, any minor tilt in the membrane is increasingly exaggerated as membrane actuation increases. It also causes tilting of the inner region 951 of the membrane, which is desired to be parallel to the substrate in many applications. Even if the snapdown is a ‘soft’ snap-down as discussed above, tilting in this way is still undesirable as it interrupts operation of the device.

[0131] One way of reducing the tilting effect is to provide inner floating electrodes 920a, 920b, 920c, 920d that are electrically isolated from each other. Figures 12A to 12D show an example of this approach. Figure 12A is a plan view from above, Figure 12B is cut-away perspective view showing half of the MEMS device, Figure 12C is a cross-sectional view showing capacitances of the device, and Figure 12D is a plan view of the actuation electrodes.

[0132] The MEMS device 1200 of Figures 12A to 12D is similar to Figures 9A to 9C and like reference numerals denote like parts. Whereas in Figures 9A to 9C, the inner floating electrodes 920a, 920b, 920c, 920d are electrically connected together by a common conductive structure 920, in the device of Figures 12A to 12D, each floating electrode is electrically isolated from the other floating electrodes. In this architecture, each floating electrode will independently re-distribute the input voltage across the two associated capacitances Ci, C2 , with lower voltages appearing across the variablecapacitances C2 as gap d decreases. This allows the structure to self-correct in the case of tilting, thus avoiding or reducing the risk of run-away attraction and snap down on one side.

[0133] Each of the floating electrodes has an outer floating electrode 923ai, 923a2, 923bi, 923b2, 923ci, 923C2, 923di, 923d2 overlying the first actuation electrode 910 and an inner floating electrode 925ai, 925a2, 925bi, 925b2, 925ci, 925C2, 925di, 925d2 overlying the second actuation electrode 912. The inner and outer floating electrodes may be joined by a thin electrically conductive strip. In the example shown in Figure 12A, the suspended membrane 950 has four sides and each side of the suspended membrane has two floating electrodes electrically isolated from each other and from the other floating electrodes. The parallel pairs may be referred to as split electrodes. The arrangement of Figure 12A thus comprises 8 split floating electrodes, with 2 floating electrodes on each side.

[0134] Figure 12D shows a plan view of a structure of the first actuation electrode 910 and the second actuation electrode 912. Each split floating electrode overlies a respective part of the first actuation electrode 910 and the second actuation electrode 912, and consequently each split floating electrode forms a respective pair of capacitors with the actuation electrodes beneath. As the floating electrodes are electrically isolated from each other, this allows each pair of capacitors to independently adjust the voltage distribution, which provides a degree of self-levelling of the device during actuation. For each floating electrode, for an applied actuation voltage across the pair of capacitors, the capacitance across the floating electrode and the first actuation electrode 910 is constant, while the capacitance between the floating electrode and the second actuation electrode 912 is variable. Therefore, when the movable part of a floating electrode on one side of the membrane is attracted more towards the second actuation electrode, the voltage across this region is less than for the floating electrode on the other side. Therefore, the movable part of the floating electrode on the other side of the membrane will have more electrostatic force and this has the effect of balancing the membrane. In this way the tilting instability is reduced.

[0135] For example, as shown in Figures 13 A and 13B, fixed outer floating electrodes 923ai, 923bi and movable inner floating electrodes 925ai, 925bi form first and second capacitors Ci, C2 with the parts (regions) of first and second acutation electrodes 910 and 912 which they overlie. Therefore, if the inner floating electrode 925aitilts down compared to the inner floating electrode 925bi, the capacitance C2 associated with the inner floating electrode 925aiincreases as the separation reduces. As the capacitance Ci is constant while C2 increases, and as the applied voltage Vsis split between Ci and C2 in inverse proportion to their capacitance, the voltage V2 across C2 decreases. Therefore the electrostatic attractive force between the inner floating electrode 925aiand the second actuation electrode 912 decreases. In contrast, the voltage V2 across C2 associated with the inner floating electrode 925bi does not decrease as the inner floating electrode 925bi has not tilted down. The force between the inner floating electrode 925bi and the second actuation electrode 912 is therefore greater than the force between the inner floating electrode 925aiand the second actuation electrode 912, which allows the inner floating electrode 925bi to ‘catch-up’ with the inner floating electrode 925aiand self correct the tilt. In this way, the inner region 951 of the membrane may be kept substantially parallel to the substrate and the titling instability reduced or eliminated. Splitting the floating electrodes on each side into two or more floating electrodes on each side of the membrane further reduces tilting instability.

[0136] As shown in Figures 12A to 12D, the inner and outer parts of each floating electrode may be connected by a thin strip and the inner part may be tapered towards the innermost end, in a similar manner to Figures 9A to 9C described above. Similarly, a plurality of notches 990 may be provided in the membrane to reduce the risk of buckling.

[0137] Note that while there are 8 independent pairs of capacitors in Figs 12A and 12B, Fig. 12C is a cross section which shows two of the capacitors pairs. Fig. 12C shows the device before actuation, at which time all of the eight pairs of capacitances are the same, and so the capacitor pair at the left of Fig 12C is shown as having the same capacitance Ci, C2 as the capacitor pair at the right of Fig 12C. While thecapacitances are initially the same, if tilting occurs then the capacitances and ratios of the voltage split change, as described above, causing the device to self-correct.

[0138] Figure 14 is a graph 1400 showing a relationship between applied actuation voltage and displacement of a central region of the suspended membrane for a prototype MEMS device 1200 shown in Figures 12A to 12C, where the suspended membrane is formed of silicon nitride and a tensile stress of 200 Mpa. It can be seen that the actuation range (i.e. displacement limit) of the centre of the optical area is extended from 38.5% to approximately 50% of the gap distance between the floating and actuation electrodes.

[0139] Figures 15A to 15D show a further example device 1500 in which inner floating electrode gates 925ai, 925a2, 925a3, 925a4 are radially spaced apart so that some are disposed closer to the centre than others. In this context, the term “floating electrode gate” or “floating gate portion” refers to an area of the floating electrode which overlaps a corresponding actuation electrode 910, 912 on the layer below. In this way, the split floating electrodes may be arranged in concentric rings around the actuation area 951. This distribution of electrode gates helps to further improve the flatness and orientation of the deflected area 951 during actuation. Inner and outer gate portions of each floating electrode may be linked by a thin connecting section. Figure 15A is a plan view from above, while Figure 15B is cut-away perspective view showing half of the MEMS device, Figure 15C is a cross-sectional view from the side and Figure 15D is a plan view of the actuation electrodes.

[0140] In the example of Figure 15A, each side of the structure has four split electrodes on each side. Each split electrode has an outer gate 923ai, 923a2, 923a3, 923a4 overlying / overlapping the first actuation electrode 910 and an inner gate 925ai, 925^, 925a3, 925a4overlying the second actuation electrode 912. The inner gates may be staggered so that they are disposed at different distances toward the centre. By staggering the gates, or having multiple rings of floating electrode gates, when an actuation voltage is applied a larger voltage is induced on the outer gates compared to the inner gates. Further, as the inner gates are staggered so that for example inner gates925aiand 925a4 are positioned inwardly of inner gates 925a2, 925a3, a larger voltage is induced at inner gates 925a2 and 925a3 compared to inner gates 925ai, 925a4. In this way, during actuation and displacement of the deflection region 951 there is a larger force on the outer electrode gates, which will tend to “flatten” the actuated area.

[0141] In summary, the device of Figs. 15A to 15D has split floating electrodes which are staggered and / or of different area (and, hence, capacitance) in order to result in different forces being applied to different regions of the suspended structure in order to flatten the suspended structure. This can be achieved by designing the capacitance ratios such that the force across the capacitance which is closer to the anchor edge so as to be greater than the force across the capacitance which is disposed further away from the anchor edge.

[0142] Figure 16 is a graph 1600 showing a relationship between applied actuation voltage and displacement of the central region of the suspended membrane for a prototype MEMS device as shown in Figures 15A to 15D. It can be seen that the actuation range (i.e. displacement limit) of the centre of the optical area is extended to approximately 60.7% of the gap distance between the floating and actuation electrodes. Thus it can be seen that the staggered electrode arrangement of Figure 15 A provides a larger actuation range than the non-staggered electrode arrangement of Figure 12 A.

[0143] Figure 17 shows an example method 1700 of manufacturing a MEMS device according to the present disclosure.

[0144] Step 1710 comprises forming a first actuation electrode and a second actuation electrode over a substrate, and forming a pair of terminals through which a voltage can be applied between the first and second actuation electrodes.

[0145] Step 1720 comprises forming a support structure for supporting a suspended membrane including a first support and a second support.

[0146] Step 1730 comprises forming a suspended membrane supported by the first support and second support over the substrate and a floating electrode supported by the suspended membrane, wherein the floating electrode overlies at least a part of the first actuation electrode and the second actuation electrode.

[0147] The method may further comprise, after step 1710 and before step 1720, forming a sacrificial layer over the substrate and the first and second actuation electrodes. The method may further comprise after step 1730, removing at least part of the sacrificial layer to form a gap between the membrane and the substrate and first and second actuation electrodes, such that the membrane is suspended over the gap between a first fixed support and the second fixed support.

[0148] Fig. 18 shows an example fabrication method in diagrammatic form. At step 1 shown in Figure 18A, actuation electrodes 1810, 1812 are formed on a substrate 1830. At step 2, a sacrificial layer 1840 is deposited over the actuation electrodes 1810, 1812, as shown in Figure 18B. At step 3, floating electrodes 1820 are deposited, as shown in Figure 18C. At step 4, a suspended membrane 1850 is deposited, as shown in Figure 18D. At step 5, part of the sacrificial layer 1840 is removed so that the suspended membrane 1850 and floating electrodes 1820 are suspended above the substrate 1830, as shown in Figure 18E. In this example, the remaining part of the sacrificial layer 1840 acts as the support structure. While the floating electrodes is shown beneath the suspended membrane 1850 in the example of Fig. 18, in another implementation it could be on top of or embedded in the suspended membranel850, in which case at least part of the suspended membrane 1850 may be formed before depositing the floating electrodes 1820.

[0149] The techniques, methods and structures described in this application may be applied to MEMS devices which adjust or vary the separation between first and second parts (e.g. the substrate and the suspended membrane). It may be particularly useful where it is desired to keep the first and second parts parallel to each other as they move towards or away from each other. It may for example be applied to MEMS devices for optical applications, such as an optical filter. The MEMS device may have a maximumseparation gap of 12,000 nm or less between the first and second parts and can be used to actuate small movements and adjustments of the gap.

[0150] At least some of the examples described herein make it possible to increase the extent of displacement of the membrane and the floating electrode before snap down. Further, due to increase in capacitance at shorter distances, the electrostatic force decreases when the suspended floating electrode and second actuation electrode are very close and when they are almost touching, the capacitance is so high that the induced voltage is almost zero, resulting in a "soft" snap-down. This may reduce or remove the need for anti-stiction bumps to prevent the electrodes sticking together. Further, in the embodiments with multiple electrically isolated or split floating electrodes, such as Figures 12A and 15 A, the positioning of multiple floating electrodes provides protection against tilting and leads to “flattening” of the displaced area. This allows the device to reduce early snap down resulting from membrane tilt. The flattening effect can be improved and the displacement range extended by staggering the split floating electrodes as shown in Figure 15 A. Furthermore, the arrangements with split floating electrodes may be less affected by inertial and vibration effects due to the “self-levelling” of the floating gate.

[0151] All of the features of the various example apparatus disclosed in this specification (including any accompanying claims, abstract and drawings), and / or all of the blocks of any method or process so disclosed, may be combined in any combination, except combinations where at least some of such features and / or blocks are mutually exclusive.

[0152] It will be appreciated by persons skilled in the art that numerous variations and / or modifications may be made to the above-described embodiments, without departing from the broad general scope of the present disclosure. The present embodiments are, therefore, to be considered in all respects as illustrative and not restrictive.

Claims

CLAIMS1. A micro-electro-mechanical system (MEMS) device comprising: a substrate; a suspended membrane suspended over the substrate between a first support and a second support; and an actuation mechanism for moving the suspended membrane relative to the substrate, the actuation mechanism comprising: i) a first actuation electrode and a second actuation electrode associated with the substrate; and ii) at least one floating electrode supported by the suspended membrane, wherein each floating electrode overlies at least a portion of both the first and second actuation electrodes and is electrically unconnected to the first and second actuation electrodes; wherein application of a potential difference between the first and second actuation electrodes creates a first charge on a surface of the first actuation electrode and a second charge on a surface of the second actuation electrode, the first and second charges inducing respective first and second induced opposite charges on the respective overlapping regions of the floating electrode, which causes the floating electrode and the suspended membrane to electrostatically move the towards the substrate.

2. The MEMS device of claim 1, wherein the suspended membrane is electrically insulating.

3. The MEMS device of claim 1 or 2, wherein the suspended membrane has an inner region which does not have any floating electrodes, and the at least one floating electrode is positioned adjacent the inner region.

4. The MEMS device of any one of claims 1 to 3, wherein a dielectric layer is positioned between the first actuation electrode and the suspended membrane.

5. The MEMS device of any one of the preceding claims, comprising at least one first actuation electrode that surrounds at least one second actuation electrode, the at least one second actuation electrode surrounding an inner region of the substrate.

6. The MEMS device of any one of the preceding claims, wherein the suspended membrane comprises an inner region and at least 2 floating electrodes disposed on opposite sides of the inner region.

7. The MEMS device of claim 6, comprising a plurality of floating electrodes on each side of the inner region.

8. The MEMS device of claim 7, wherein the plurality of floating electrodes are electrically connected together.

9. The MEMS device of claim 7, wherein each floating electrode is electrically isolated from the other floating electrodes.

10. The MEMS device of claim 9, wherein each floating electrode has an outer portion overlapping the first actuation electrode and an inner portion overlapping the second actuation electrode, and wherein the inner portions of the floating electrodes are staggered so that the inner portion of at least one floating electrode is positioned inwardly of the inner portion of at least one other floating electrode.

11. The MEMs device of any one of the preceding claims, wherein the first actuation electrode and the second actuation electrode are provided in a common layer.

12. The MEMS device of any one of the preceding claims, comprising a plurality of floating electrodes provided in a common layer.

13. The MEMS device of any one of the preceding claims, wherein the device is configured so that application of a potential difference between the first and second actuation electrodes causes a central region of the suspended membrane to move towards the substrate and remain substantially parallel to the substrate during the movement.

14. The MEMS device of any one of the preceding claims, wherein the device is configured such that a separation gap between the suspended membrane and the substrate is dependent on a magnitude of the potential difference applied between the first and second actuation electrodes.

15. The MEMS device of claim 14, wherein the separation gap has a maximum distance, wherein the maximum distance is 12,000 nm or less.

16. An optical device comprising the MEMS device of any one of the preceding claims and an optical component attached to one or both of the substrate and the membrane.

17. The optical device of claim 16, wherein a first mirror or reflector is mounted to the membrane and a second mirror or reflector is mounted to the substrate.

18. An optical spectrometer for measuring the optical spectrum of light, the spectrometer comprising an optical filter including the MEMS device of any of the preceding claims, a first mirror or reflector mounted to substrate and a second mirror or reflector mounted to the membrane.

19. A method of operating the MEMS device of any one of claims 1 to 15, comprising applying a potential difference between the first actuator and the second actuator in order to cause the membrane to move relative to the substrate.

20. A method of manufacturing the MEMS device according to any one of claims 1 to 15, the method comprising:forming a first actuation electrode and a second actuation electrode over a substrate; forming a pair of terminals through which a voltage can be applied between the first and second actuation electrodes; forming a support structure for supporting a suspended membrane, the support structure including a first support and a second support; and forming a suspended membrane supported by the first support and second support over the substrate and forming a floating electrode supported by the suspended membrane, wherein the floating electrode overlies at least a part of the first actuation electrode and the second actuation electrode.

21. The method of claim 20, further comprising: forming a sacrificial layer over the substrate and the first and second actuation electrodes; and removing at least part of the sacrificial layer to form a gap between the membrane and the substrate, such that the membrane is suspended over the gap between a first fixed support and the second fixed support.

Images