60/348,201, filed 11 January 2002, the entire content of which is incorporated herein.

FIELD OF THE INVENTION The invention is directed to an electrostatic actuator wherein the conventional air gap found between the dielectric and one of the capacitor plates is replaced with a conductive material.

BACKGROUND "Micro-Electro-Mechanical Systems" (MEMS) is an umbrella term given to a wide range of devices that integrate mechanical elements (such as sensors and actuators) and electronics on a common substrate. MEMS devices are fabricated using microfabrication technology. As a general proposition, the electronic portions of MEMS devices are fabricated using the same processes and materials used to fabricate integrated circuits (e. g., CMOS (Complementary Metal-Oxide Semiconductor), Bipolar, or BiCMOS (Bipolar Complementary Metal-Oxide Semiconductor) processes). The micromechanical components are fabricated using compatible"micromachining"processes that selectively etch away parts of the substrate or add new structural layers to form the mechanical and electromechanical aspects of the device. MEMS devices are very promising because they bring together silicon-based microelectronics with micromachining technology, thereby, making possible the realization of extraordinarily small and complete electro- mechanical systems-on-a-chip.

In its most basic form, the sensor (s) of a MEMS device gathers desired information from the environment through measuring, for example, mechanical, thermal, biological, chemical, optical, and magnetic phenomena. The associated

electronics of the MEMS device then process the information derived from the sensor (s) and, through a pre-programmed or pre-determined decision making capability, directs the actuator (s) to respond by moving, positioning, regulating, and/or directing the mechanical portion of the MEMS device, thus controlling the environment for some desired outcome or purpose (such as accurately positioning a workpiece).

SUMMARY OF THE INVENTION As described in complete detail hereinbelow, a first embodiment of the invention is directed to a capacitor-type actuator. This first embodiment of the comprises a first conductive plate and a second conductive plate defining a space between the plates. One of the first or second conductive plates is movable, while the other of the plates is not movable. A dielectric is disposed in the space between the first and second conductive plates. The dielectric is adhered (permanently or removably) to the conductive plate that is movable. Lastly, an electrically- conductive material is adhered the conductive plate that is not movable, at a location within the space defined by the two conductive plates.

The electrically-conductive material may be a liquid, a gas, or a solid. A solid material is preferred. The most preferred solid material are single-or multiple-walled carbon nanotubes, carbon fullerenes, silicon fullerenes, germanium fullerenes, and conducting polymers. It is preferred, although not required, that the electrically-conductive material has a relative dielectric greater than 100, more preferably greater than 1,000, and more preferably still greater than 10,000.

BRIEF DESCRIPTION OF THE FIGURES FIG. 1 is a cross-section schematic of a parallel plate capacitor showing the force experienced by a dielectric that is partially disposed between capacitor plates 10 and 10'.

FIG. 2 is a cross-section schematic showing a flexural stage motion hinge incorporating a dielectric actuator according to the present invention.

FIG. 3 is a perspective schematic of a prior art parallel plate dielectric actuator having an air gap 15.

FIG. 4 is a perspective rendering of a dielectric actuator according to the present invention, where the air gap has been replaced with a low-friction, conducting layer 17.

DETAILED DESCRIPTION OF THE INVENTION Referring to Figs. 1 and 3, these figures illustrate a conventional parallel plate capacitor (Fig. 1) and a prior art parallel plate dielectric actuator having an air gap 15 (Fig. 3). Briefly, when the capacitor shown in Fig. 1 is charged via circuit"V"the dielectric 12 will experience a force as shown by the arrow in Fig.

1. This force can be harnessed to move one of the capacitor plates by introducing an air gap 15 between one of the capacitor plates 10 or 10'and the dielectric 12.

See Fig. 3. Thus, as shown in Fig. 3, where capacitor plate 10 is fixed, capacitor plate 10'is movable, and there is an air gap 15 between plate 10 and dielectric 12, when the capacitor is charged, dielectric 12 will experience a force that can be used to move both the dielectric 12 and the capacitor plate 10'to which it is attached.

A significant problem, however, exists with the conventional actuator shown in Fig. 3. In this conventional arrangement, the low dielectric constant of the air gap 15 will dominate the electric field of the actuator. In short, the actual capacitance of the actuator will depend greatly on the dimensions of the air gap 15.

Because the dielectric constant of the air gap is so much lower than that of the dielectric material of layer 12 (by several orders of magnitude), the air gap 15 will always have an undesirable impact on the actuator's performance. The only way to minimize the dominance of the air gap is to make the gap extremely small.

However, the dimensions required of the air gap 15 for a practical MEMS device are well beyond the capabilities of standard processing. Thus, to make an actuator that takes full advantage of the dielectric attraction, the air gap must be eliminated.

Previous attempts to solve this problem have focused entirely on minimizing the distant"g"in Fig. 3, the air gap thickness. To make a nominally functional actuator, the air gap must be on the order of 10 nm thick. Fabricating such a tiny

and consistent air gap width is not only a serious challenge for production, but also entails difficult problems with attractive forces and viscous drag from the air between the plates. In any event, surmounting the air gap problem by minimizing the dimensions of the gap can only, at best, minimize the limitations inherent in the design. Minimizing the air gap thickness cannot eliminate the inherent limitations of an actuator that includes an air gap within the capacitor configuration.

The present invention, however, solves the problems present in the air gap configuration by replacing the air gap with a conductive layer, thereby yielding a very powerful, exceedingly small dielectric actuator that does not have an air gap.

In the present invention, the air gap is replaced with a conductive, lubricating layer 17 (see Figs. 2 and 4). Using another solid or liquid dielectric to replace the air gap is inherently limiting because any dielectric other than the piezoelectric material of the dielectric 12 can only reduce the overall efficiency of the actuator. This holds true even if the dielectric constant of the lubricant 17 is higher than that of the piezoelectric material within the dielectric 12.

In the present invention, the inventor has taken a different approach, which is to use a conducting layer between plate 10 and dielectric 12 ro replace the air gap 15.

Using a conducting lubricant in layer 17 has a distinct advantage in that it virtually eliminates the detrimental effect of the lubricating layer on the overall performance of the actuator. Using this approach, the lubricating layer 17 can be considered an extension of the conductive plates (i. e. plates 10 and 10'of Fig. 2 and plate 10 of Fig. 4).

The first (although not particularly preferred) choice for the conductive layer 17 is a conductive liquid or a gas. (For the discussion that follows, the term "liquid"is deemed to encompass both non-compressible liquids and compressible <BR> <BR> gases. ) While possible to configure, a conductive liquid for the layer 17 is not preferred because the conducting layers of the capacitor (10,10'and 10"of Fig.

2) must be isolated from one another. If a liquid is used for the layer 17, the geometry required for the actuator to maintain this separation is necessarily intricate

and requires physical conductive barriers to isolate layer 17 and prevent it from migrating to other parts of the actuator.

Therefore, in the preferred embodiment of the invention, conductive layer 17 is a lubricating solid, and most preferably a layer of carbon nanotubes or carbon, silicon, or germanium fullerenes (i. e. ,"buckyballs") affixed to one plate of the capacitor.

Thus, in the preferred embodiment of the invention, the layer 17 is made of conductive carbon nanotubes (either single-walled carbon nanotubes (SWNT's) or multiple-walled carbon nanotubes (MWNT's) ). As shown in Figs. 2 and 4, the conducting layers 17 are depicted as arrays of conductive nanotubes, each individual nanotube being anchored to the conductive plates (10'and 10"of Fig.

2,10 of Fig. 4) at one terminus of each nanotube, the other terminus being unattached and floating free. This configuration solves all of the problems inherent in the air gap configuration.

For example, a potential problem when using a solid lubricant as layer 17 (in place of an air gap 15) is that the solid lubricant might cause a large amount of slip-stick friction, which is highly undesirable. In short, because of the small physical dimensions and forces involved, there could be bonding between the layers 12 and 17. If such bonding were of sufficient strength, the actuator would not be able to generate enough force to move once it sticks. Nanotubes, however, are flexible and very slippery, so the slip-stick problem is significantly reduced and, depending upon the overall area presented between the two layers 12 and 17, eliminated entirely. Additionally, because the nanotubes are physically and permanently attached to one capacitor plate of each capacitor plate pair, it is easy to isolate the two conductive plates. The nanotubes may be physically attached either to the fixed plate of the actuator or the movable plate.

The overriding reason that MEMS flexure nanopositioners are not available today is because of the lack of an adequate actuation mechanism. The electrostatic comb actuator, commonly used in MEMS, does not produce sufficient force to drive the stage, unless it consists of a large number of teeth and/or the spring constant of flexures is sufficiently low. Both of these conditions (multiple teeth in

the actuator, low spring constant), however, tend to make the actuator structure too fragile, fragile to the point of being unsuitable to carry a useful amount of load.

These conditions also combine to lower the mechanical resonant frequency of the device.

To take maximum advantage of miniaturization, what is needed is a MEMS-compatible actuator with much stronger driving force than the conventional comb actuator. As noted above, one possible technique is to use a high dielectric material in the gap of the actuator to take advantage of the force exerted on the dielectric material in an electric field.

Equations (1), (2), and (3) predict the force that such an actuator can produce neglecting fringe fields and the air gap. The relative dielectric, Er, is a factor of the material between the electrodes. For silicon and most standard dielectric materials used in MEMS,, is on the order of 10 or lower. For some ferroelectrics, however, Er is quite large: £r ~ 103 for PZT (lead-zirconium-titanate) and c, > 104 for barium-titanate. Thus, one could generate quite a large force with a low operating voltage.

Eliminating the air gap, however, not only solves the problems inherent in the air gap itself, it also yields a very powerful actuator. The primary purpose of the air gap is to allow movement between the dielectric and one of the capacitor plates. Ideally the dielectric constant of layer 17, as noted above, would be much higher than the dielectric constant of the dielectric material 12. Thus, if the dielectric constant of layer 17 approaches infinity, layer 17 can be arbitrarily thick (or thin) without impacting the actuator's performance.

An array of conductive nanotubes is relatively simple to build. Because they are conductive, the gap width (see Fig 3, reference number 15) can be much larger, thereby making the tolerances much less stringent. The nanotube layer 17 can be several micrometers thick and still yield only minimal impact on the operation of the actuator.

Alternatively, spherical fullerenes, preferably fullerenes comprised entirely of carbon atoms or comprised predominately of carbon atoms, and preferably C60, can be used in place of carbon nanotubes. Fullerenes comprised of carbon and other atoms selected from the group consisting of silicon and/or germanium may also be used as the conductive layer 17. Alternatively, conducting polymers can be deposited on the surfaces of the plates 10 and 10'via plasma-assisted chemical vapor deposition.

Fig. 4 depicts an actuator according to the present invention and Fig. 2 depicts the actuator integrated into a flexure-guided motion stage. Referring to Fig.

4 specifically, the actuator includes a parallel plate capacitor having upper conductive plate 10, lower conductive plate 10', dielectric block 12, and conductive layer 17, preferably an array of carbon nanotubes as described above, affixed to

plate 10. In the configuration shownin Fig. 4, conductive plates 10 and 10'can be fixed in place, in which case the dielectric block 12 is the only element of the device that moves relative to the other elements. Or, one of plates 10 or 10'could be movable and the other fixed, in which case the force exerted on the dielectric block 12 upon charging of the plates 10 and 10'would exert a force sufficient to move the movable plate.

Fig. 2 depicts the actuator shown in Fig. 4, integrated into a flexure-guided motion stage. In Fig. 2 is shown a parallel plate capacitor having upper plate 10 and two distinct lower plates 10'and 10". A dielectric block 12 overlaps both of the lower plates 10'and 10"Each of lower plates 10'and 10"are coated with a conductive layer of carbon nanotubes 17. The actuator is disposed within a housing 18 that includes one or more flexure hinges 14 attached to a stage area 16. As shown in Fig. 2, the upper plate 10 and dielectric 12 of the capacitor are affixed to the lower surface of the stage area 16. In this arrangement, the stage area 16 can be made to move by charging one or both of plates 10'and 10"via circuits"VL" or"VR". Using programmable circuitry (not shown), the stage area 16 can be accurately and precisely translated to the right or the left.

The housing 18, stage area 16, and flexure hinges 14 of Fig. 2 can be fabricated by any means now known in the art or developed in the future. Two preferred means are wire electric discharge machining (wire EDM) where the substrate material is conductive, and reactive ion etching (RIE) where the substrate is conductive or non-conductive. The housing 18, stage area 16, and flexure hinges 14, can be made of any suitably stiff material, either electrically conductive, semi- conductive, or non-conductive. Suitable materials include metal and metal alloys of any description. One such alloy is Invar. Silicon, silicon carbide, silicon nitride, silicon borides, and the like are also suitable. This list is exemplary and non-limiting. The capacitor plates 10,10'and 10"can be made of any suitable, electrically-conductive material. The dielectric 12 likewise can be made of any suitable dielectric. A piezoelectric dielectric material, such as quartz, is much preferred.

Wire EDM is a method to cut conductive materials with a thin electrode that follows a programmed path. The electrode is a very thin wire. Typical diameters range from roughly 10 item to 30 item, although smaller and larger diameters are available. The hardness of the work piece material has no detrimental effect on the cutting speed. There is no physical contact between the wire and the part being machined. Rather, the wire is charged to a voltage very rapidly. The wire and the work piece are surrounded by de-ionized water. When the voltage reaches a sufficiently high level, a spark jumps the gap between the wire and the work piece and melts a small portion of the work piece. The de-ionized water is circulated and serves to cool the work piece and flush away the small particles from the gap. Wire EDM is generally accurate to approximately 0. 0001 cm. The process is well known and widely employed in the manufacture of parts requiring exacting dimensions and tolerances.

Reactive ion etching (RIE) is used to form shapes on work pieces such as semiconductor wafers. Like wire EMD, the RIE process is well known and widely employed in the manufacture of parts requiring exacting dimensions. In a typical RIE process, radio frequency (RF) or microwave power is used to excite a gas to form a plasma. In conjunction with mask works covering the work piece, the plasma is then used to etch desired shapes into the work piece. Reactive ion etching uses reactive species in the plasma to remove materials selectively. For example, gases such as SF6, CHF3, and 02 are commonly used in RIE. Etching products remain in gas phase and are pumped out of the system immediately and continuously.

As noted above, the attached nanotubes can be multi-wall nanotubes (MWNT) or single-wall nanotubes (SWNT). The fabrication of such nanotubes is well known to those skilled in the art and will not be discussed in any detail herein.

See, for example, WO 98/05920, published 12 February 1998, incorporated herein in its entirety. Note also that this reference describes attaching carbon nanotubes to a substrate. To attach the nanotubes to the capacitor plate, an electrically conductive glue can be used or electrically conductive double-sided carbon tape.

These items are available commercially from a host of national and international

suppliers, including Holland Shielding Systems BV (Dordrecht, The Netherlands), Structure Probe, Inc. (West Chester, Pennsylvania), and Ernest F. Fullam, Inc.

(Latham, New York).

Alternatively, the carbon nanotubes may be grown directly on the capacitor plate itself. See for example, E. W. Wong, P. E. Sheehan and C. M. Lieber, "Nanobeam Mechanics: Elasticity, Strength, and Toughness of Nanorods and Nanotubes,"Science 277, pp. 1971-1975,1997 ; and J. H. Hafner, C. L. Cheung and C. M. Lieber, "Growth of nanotubes for probe microscopy tips,"Nature 398, pp. 761-762,1999. Carbon nanotubes can also be grown directly on the plates via chemical vapor deposition or plasma-aided chemical vapor deposition.

A host of fullerenes are available commercially from the Aldrich Chemical Company, Milwaukee, Wisconsin, including C60, C761C80, and C84. Likewise, a host of different types of carbon nanotubes can also be purchased from Aldrich, see catalog nos. 40,607-4 ; 41,300-3 ; 41,299-6 ; and 41-298-8. When using fullerenes as the conductive layer 17, it is preferred that they by attached to the capacitor plate via conductive glue or double-sided carbon tape, or any other means whereby the fullerenes are adhered to the capacitor plate.

A wide range of electrically-conductive polymers are known in the art and can be used in the invention. An exemplary, non-limiting list of suitable electrically-conductive polymers includes emeraldine-based polymers (such as Panipol-brand polymers, available commercially from Panipol Ltd, Porvoo, Finland), polyaniline-based conductive polymers (such as Ormecon-brand polyaniline polymers, available commercially from Zipperling Kessler & Co., Ahrensburg, Germany), polypyrrole-based polymers (available commercially from Milliken Research Corporation, Spartanburg, South Carolina), polythiophene-based polymers and polyethylenedioxythiophene-based polymers (available commercially from Bayer Corporation, Pittsburgh, Pennsylvania), and poly (p-phenylene vinylene) -based polymers (available commercially from Covion, Frankfort, Germany).

The actuators of the present invention are useful in any application requiring the precise, accurate, and reliable positioning of a workpiece in space. The

actuators are also useful as a physical switch to actuate a desired action in any given circumstance. In short, wherever an actuator is required, the present invention can be used. Two or more of the actuators according to the present invention can be disposed in a cooperative fashion to yield linear or rotary motors that comprise a plurality of operationally-linked actuators as described herein.