| JP3353556 | SEALED CONTACT DEVICE |
| WO/2000/044012 | MICROSWITCHING CONTACT |
| JP10233150 | MOVABLE CONTACT SUPPORT OF ELECTROMAGNETIC CONTACTOR |
SUY, Hilco (Betchworth House 57-65 Station Road, Redhill Surrey RH1 1DL, GB)
GOOSSENS, Martijn (Betchworth House 57-65 Station Road, Redhill Surrey RH1 1DL, GB)
STEENEKEN, Peter, G. (Betchworth House 57-65 Station Road, Redhill Surrey RH1 1DL, GB)
SUY, Hilco (Betchworth House 57-65 Station Road, Redhill Surrey RH1 1DL, GB)
GOOSSENS, Martijn (Betchworth House 57-65 Station Road, Redhill Surrey RH1 1DL, GB)
CLAIMS
1. A MEMS device comprising: first and second opposing electrodes (42,46), wherein the second electrode (46) is electrically movable to vary the electrode spacing between facing first sides of the first and second electrodes, wherein the second electrode (46) is at a spaced position from the first electrode (42) in a relaxed state in which no voltages are applied to the electrodes, wherein a first gas chamber (50) is provided between the electrodes, at a first pressure, and a second gas chamber (52) is provided on the second, opposite, side of the second electrode at a second pressure which is higher than the first pressure.
2. A device as claimed in claim 1 , wherein the first gas chamber (50) is at a pressure below SOOmbar, and the second gas chamber is at atmospheric pressure or higher than atmospheric pressure.
3. A device as claimed in claim 1 or 2, wherein the first gas chamber (50) is at a pressure below 100mbar.
4 A device as claimed in any preceding claim, wherein the second gas chamber (52) is at atmospheric pressure.
5. A device as claimed in any preceding claim, wherein the first gas chamber (50) has a height of between 50nm and 50μm and the second gas chamber (52) has a height of between 1 μm and 200μm.
6. A device as claimed in any preceding claim, wherein one or more stops (60) are provided in the second gas chamber (52), for limiting movement of the second electrode (46) away from the first electrode (42).
7. A device as claimed in any preceding claim, wherein the second electrode (46) is clamped between elastic members (62) around a portion of the second electrode (46) which defines the outer periphery of the first and second gas chambers.
8. A device as claimed in any preceding claim, wherein the second electrode (46) comprises a metal electrode layer (84), a piezoelectric layer (80) and a circuit for dissipating electrical energy generated by the piezoelectric layer.
9. A device as claimed in claim 8, further comprising a circuit for controlling the actuation of the piezoelectric layer to provide active damping.
10. A device as claimed in any preceding claim comprising a MEMs switch.
11. A device as claimed in any preceding claim comprising a capacitive switch, having a dielectric layer between the first and second electrodes.
12. A device as claimed in any one of claims 1 to 10 comprising a galvanic switch, in which the second electrode is movable to make and break contact with the first electrode.
13. A method of controlling the movement of an electrically movable electrode (46) of a MEMS device towards and away from a fixed electrode (42), wherein the movable electrode (46) is at a spaced position from the fixed electrode in a relaxed state in which no voltages are applied to the electrodes, the method comprising: providing a first gas chamber (50) provided between the electrodes with a first pressure; providing a second gas chamber (52) provided on an opposite side of the movable electrode facing away from the fixed electrode with a second pressure which is higher than the first pressure. |
DESCRIPTION
MEMS DEVICES
This invention relates to MEMs (micro electromechanical system) devcies.
MEMs technology is being used increasingly in electronic devices. Switches and tunable capacitors are examples of circuit components that can be made using MEMs technology.
MEMs switches can be used in a wide variety of applications, where high speed, typically low current, switching is required within a circuit. A MEMs switch has the advantage of a large capacitance switching range.
MEMs capacitors can also be used in a variety of circuits, such as tunable filters, tunable phase shifters and tunable antennas. One application of growing interest is in RF and microwave communications systems, for example for use in low cost reconfigurable/tunable antennas.
Figure 1 shows a MEMS capacitor with an electrically tunable dielectric and MEMS controlled dielectric spacing. The dielectric spacing is controlled in the manner of a switch, but the analogue electrical control of the dielectric properties can enable continuous tunability of the capacitor.
A tunable dielectric, ferroelectric or piezoelectric material can be used, such as Bai -x Sr x Tiθ3 or PZT as a dielectric layer 14. By combining a MEMS capacitor with a tunable dielectric, the advantages of the large capacitance switching range of RF MEMS switches are added to the advantages of the continuous tuning capability of tunable dielectrics. Moreover, use is made of the beneficial high dielectric constant of ferroelectrics, which can be 10-200 times higher than that of conventional dielectrics like Silicon Nitride. This dramatically reduces device size and increases continuous tuning range. The device comprises opposite capacitor plates 10 (e1 ) and 12 (e2).
The gap g is controlled by the MEMS switch represented by the spring k, based on the voltage applied to the plate 12. A dc voltage Vdc_switch is used
to provide this MEMS switching function, from a dc voltage source 18. An rf ac voltage source 16 represents the rf signal that is flowing through the MEMS device during operation. The tunable dielectric has a tunable dielectric value Ed, whereas the remaining dielectric spacing is air or vacuum, with dielectric value Eo. The tunable dielectric is controlled by the voltage Vdc_tune, so that the single voltage applied to the electrode 12 controls the MEMS switching and dielectric tuning. The capacitor C and resistor R are optional decoupling components.
Of course, if a non-tunable dielectric is used, the MEMs device can be used simply as a capacitive switch (low frequency or rf frequency). Alternatively, if no dielectric is provided, the device can be used as a galvanic switch.
Switches/tunable capacitors of the type shown in Figure 1 are usually operated in a gas atmosphere at a pressure close to 1 bar. A drawback of the presence of this gas is that it exerts a damping force on the moving electrode of the switch, which reduces its speed. Because the gas needs to be 'squeezed' away between the electrodes, this damping force is sometimes called the 'squeeze film damping' force, and is approximately given by:
F d = -bí ... [1 ] z
In equation [1], v is the speed of the electrode, b is a constant that depends on the geometry of the switch, and z is the separation between the plates. Clearly, the damping force increases tremendously when the plates approach each other and z becomes small. This strongly limits the switching time both for opening and for closing of the switch.
An obvious and well-known solution to this problem is to operate the switch in a vacuum or low pressure. At the limit, the gas damping force becomes close to zero and the switch closes and opens very quickly. However this solution does have another drawback: when the switch opens, the spring-mass system is strongly underdamped and the electrode will
oscillate for a long time before finally conning to a relaxed position. This increases the effective opening time (usually the switch can only be used after stabilizing, particularly in the case of capacitive MEMS switches).
Figure 2 shows the opening of a capacitive MEMs switch at atmospheric pressure (1000 mbar). The opening speed is low, especially when the electrode gap z is very small.
Figure 3 shows the opening of the same capacitive MEMs switch at a low pressure (1.4 mbar). Because the gas damping is very small, it takes many oscillations for the switch to relax to a stable position.
According to the invention, there is provided a MEMS device comprising: first and second opposing electrodes, wherein the second electrode is electrically movable to vary the electrode spacing between facing first sides of the first and second electrodes, wherein the second electrode is at a spaced position from the first electrode in a relaxed state in which no voltages are applied to the electrodes, wherein a first gas chamber is provided between the electrodes, at a first pressure, and a second gas chamber is provided on the second, opposite, side of the second electrode at a second pressure which is higher than the first pressure.
This arrangement enables movement of the second electrode towards the first electrode without needing to displace significant volumes of fluid, as the chamber between the electrodes is at a relatively low pressure. The movement of the second electrode towards the relaxed position is damped by the second gas chamber so as to prevent unwanted oscillations. The invention thus provides the ability for high speed switching, whilst also reducing settling times.
The first gas chamber is preferably at a pressure below δOOmbar, and the second gas chamber is at atmospheric pressure or higher than atmospheric pressure. For example, the first gas chamber can be at a pressure below 100mbar, or even below 50mbar, namely at a partial vacuum
pressure. The second gas chamber can for example be at atmospheric pressure. This simplifies the manufacture as the chamber does not need to be pressurised or depressuhsed as part of the manufacturing process.
The first gas chamber can have a height of between 50nm and 50μm and the second gas chamber can have a height of between 1 μm and 200μm.
The electrostatic control is across the first gas chamber, which has a relatively small height, whereas the second gas chamber is used for mechanical damping only, and therefore can have greater height.
One or more stops can be provided in the second gas chamber, for limiting movement of the second electrode away from the first electrode. These can assist further in damping oscillations as the second electrode moves to its relaxed state.
The second electrode can be clamped between elastic members around a portion of the second electrode which defines the outer periphery of the first and second gas chambers. Again, this can be used to provide damping of the movement of the second electrode.
The second electrode can comprise a metal electrode layer, a piezoelectric layer and a circuit for dissipating electrical energy generated by the piezoelectric layer. This arrangement uses the conversion of mechanical to electrical energy as a further possible way to provide damping of the movement of the second electrode. A circuit can then be provided for controlling the actuation of the piezoelectric layer to provide active damping.
The device of the invention can for example comprises a MEMs switch (either a capacitive switch having a dielectric layer between the first and second electrodes, or a galvanic switch in which the second electrode is movable to make and break contact with the first electrode).
The invention also provides a method of controlling the movement of an electrically movable electrode of a MEMS device towards and away from a fixed electrode, wherein the movable electrode is at a spaced position from the fixed electrode in a relaxed state in which no voltages are applied to the electrodes, the method comprising:
providing a first gas chamber provided between the electrodes with a first pressure; providing a second gas chamber provided on an opposite side of the movable electrode facing away from the fixed electrode with a second pressure which is higher than the first pressure.
Examples of the invention will now be described in detail with reference to the accompanying drawings, in which:
Figure 1 shows a MEMS device which combines an electrically tunable dielectric and a MEMS switch function;
Figure 2 shows how the movable electrode moves in a known MEMS switch using atmospheric pressure between the electrodes;
Figure 3 shows how the movable electrode moves in a known MEMS switch using a partial vacuum between the electrodes; Figure 4 shows a first example of MEMS device of the invention;
Figure 5 shows a second example of MEMS device of the invention;
Figure 6 shows a third example of MEMS device of the invention;
Figure 7 shows a fourth example of MEMS device of the invention;
Figure 8 shows a fifth example of MEMS device of the invention; Figures 9 to 15 show an example of process sequence for forming a cavity between switch electrodes;
Figures 16 and 17 show two cross sections of resulting device structure; and
Figure 18 shows an example of device layout in plan view.
The same reference numbers are used in different figures to denote the same components. In Figures 9 to 18, the description of layers is not repeated when the same layers appear in different figures.
The invention provides a MEMS device in which the movable electrode provides a barrier between two gas chambers. A first gas chamber is between the movable electrode and a fixed electrode, and is at a low pressure so that the movable electrode can move quickly towards the fixed electrode. A
second gas chamber is on the opposite side of the movable electrode, and this acts as a damping device, to remove oscillations when the movable electrode returns to its relaxed position.
Figure 4 shows a first example of MEMS device of the invention, comprising a lower substrate assembly which includes a lower first electrode 42. This example is a capacitive switch, and the lower substrate assembly includes a dielectric layer 44 over the first, fixed electrode 42.
The second, movable electrode 46 is suspended over the first electrode, and is supported by a closed support 48 so that a first sealed gas chamber is defined between the electrodes. The second electrode is in the form of a deformable membrane, and it provides a hermetic seal of the first gas chamber. This chamber is provided with a low pressure, in particular a partial vacuum. A second gas chamber 52 is defined above the second electrode, by the materials of the device package, and this is at a higher pressure, such as atmospheric pressure.
The lower gas chamber has a height g1 , and the upper has chamber has a height g2.
The height g1 is between 50nm and 50μm, for example between 100nm and 10μm, more preferably between 100nm and 5μm. The height g2 can be larger as it is used for mechanical damping only and not for electrostatic actuation. For example g2 can be between 1 μm and 200μm, for example between 10μm and 100μm.
By applying a high pressure in a small gap above the switch, and a low pressure below the switch, the so-called squeeze film damping forces are small when the switch is closed, but they become significant when the switch is in the open position. By tuning the gap size g2 and the pressure in the second chamber, the required damping can be achieved.
In this way, a small damping is provided when the gap z between the electrodes is small, to prevent slow opening as explained with reference to Figure 2, but large enough damping is provided when the switch is opened to prevent oscillations as explained with reference to Figure 3.
Figure 5 shows a second example of MEMS device of the invention, which adds one or more stops 60 to the design explained with reference to Figure 4. These stops prohibit the top electrode 46 from making large upward deflections. When the electrode touches the stop 60, this absorbs energy and damps the oscillations. Moreover, the small air gap between the stop 60 and the top electrode 46 provides increased squeeze film damping when the relaxed position of the second electrode is reached.
Figure 6 shows a modification to the design of Figure 4, in which the top electrode 46 is clamped in an absorbing visco-elastic material 62 (plastic, resist, polymer). This can absorb the mechanical energy and also reduce oscillations.
Figure 7 shows a fourth example of MEMS device of the invention in which active damping electronics 70 is used to provide damping. This controls the voltage profiles applied to the electrodes. This may not be a preferred implementation because high voltages are required and complicated control circuitry needs to be added. However, electrical damping of the oscillations can be achieved in more simple manner using the arrangement shown in Figure 8.
In Figure 8, a piezoelectric layer 80 is used to convert the mechanical energy of the switch to electrical energy which is dissipated in a resistor 82 (a different circuit can be used). This energy dissipation will again reduce the oscillations shown in Figure 3.
In this arrangement, the second electrode 46 comprises an assembly of a piezoelectric layer 80 sandwiched between conductor layers 84. Electrodes 86 are also shown for piezoelectric actuation. Piezoelectric actuation can be used for active damping of the oscillations in the switch by inducing bending in the electrode layer which tends to reduce the bending/oscillations which have been induced.
The use of piezoelectric layer can enable active damping to be implemented more easily.
From the various examples above, it can be seen that the invention enables a decrease in the switching time of a MEMS switch, by creating a
partial vacuum between the plates of the switch to reduce squeeze film damping. This would normally be considered to introduce the problem of undesired oscillations which increase the opening time of the switch. These oscillations are then damped by providing a higher pressure controlled gap above the top electrode.
Additional measures can then be employed to improve the damping, such as stops to limit the amplitude of the oscillations and absorb their energy, clamping of the electrode in a viscoelastic material, active damping electronics and piezoelectric damping via a resistor or active damping electronics. The forming of the low pressure cavity will be routine to those skilled in the art. In particular, MEMS devices typically require a hermetic package, and the semiconductor processes are performed in an evacuated chamber, so that the required low pressure encapsulation is not difficult to achieve.
An example of manufacturing method is shown in Figures 9 to 17. Figure 9 shows the substrate 90 with a thermal oxide passivation layer
92, for example 1 μm thick. It also shows the patterned bottom electrode 94 for example formed from 30nm Cr and 250nm Au. This patterning uses a first mask.
Figure 10 shows a sacrificial layer 96, for example 150nm Mo which has been structured using mask number 2a to form opening 98.
The opening is used to enable an electrode shape to be defined which has bumps on the underside. These can be used to act as electrical contacts.
As will become evident from Figure 17, these bumps will be positioned over the bottom electrode, but this is not shown in Figures 9 to 15 in order to keep the diagrams simple.
A second sacrificial layer 97 has been deposited in Figure 11 , again of 150nm Mo, so that the opening 98 is partially filled.
The double sacrificial layer is then patterned with a mask step number
2b, and a top electrode 100 is deposited, resulting in the structure shown in Figure 12. The top electrode can comprise 250nm of Au and 30nm Cr, namely the same composition as the bottom electrode. This top electrode is then patterned, using mask 3.
A structural layer 102 is then deposited over the structure, for example
20nm SiO2 and 400nm SiN by PECVD deposition, resulting in the structure shown in Figure 13. This layer 102 provides an insulating connection between the different parts of the layer defining the second electrode, so that these parts all move up and down together.
Sacrifical layer holes 104 are etched with mask 4, and these are used to remove the Mo sacrificial layer 96 by a wet etch, followed by drying, to give the arrangement of Figure 14.
A plug layer 106 is shown in Figure 15, for example 600nm SiN or TEOS in order to plug the sacrificial layer holes. Figure 15 also shows bondpads having been opened with a mask number 5.
Figures 16 and 17 show cross sections of a structure which can be made using the processing concepts explained above. The position of these cross sections can be seen from the plan view of Figure 18. Figure 18 shows an RF input 110, an RF output 112, bottom electrode contact pads 114 and a top electrode contact pad 116. The bottom electrode is patterned as two sections, one on each side of the line of the RF input and output lines, because they are all formed from the same bottom metal layer. The top electrode is a single continuous electrode, spaced above the bottom electrode by the cavity formed using the sacrificial etching method as explained above.
A central top electrode area 120 functions as the contact between the input and output lines, and is connected to the other top electrode by the structural layer 102 so that all regions of the top electrode move together. Thus, when the main electrode areas are brought together, the area 120 makes electrical contact between the RF input and output lines. The area 120 has contact bumps on the under surface.
Figure 17 shows that these contact bumps are in fact provided over the lower electrode, which is not shown in the simplified diagrams of Figures 9 to 15.
The steps shown in Figures 9 to 17 demonstrate how a cavity can be formed between the top and bottom electrodes. The top cavity above the top
electrode is not shown, and this can be made either by placing a cap above the switch membrane or by a sacrificial process similar to that described above.
In the figures, the structure of the device is shown only schematically. The top electrode and structural layer is designed to seal the lower chamber. The detailed implementation will be routine to those skilled in the art of MEMS devices.
The example above shows capacitive switches, but the invention can also apply to galvanic switches or other MEMS devices in which mechanical contact is made and released.
Various modifications will be apparent to those skilled in the art.