Microactuators

Microactuators are required to:

drive the resonant sensors, above, to oscillate at their resonant frequency.
produce the mechanical output required of particular microsystems:

moving micromirrors to scan laser beams,

to switch them from one fibre to another
to drive cutting tools for microsurgical applications
to drive micropumps and valves for microanalysis or microfluidic systems
to stimulate nervous tissue in neural prosthesis applications using microelectrodes

A variety of methods for achieving microactuation are briefly outlined:

electrostatic,
magnetic,
piezoelectric,
hydraulic, and
thermal.

Piezoelectric and hydraulic actuators currently look most promising, although the others
have their place.
Electrostatic actuation runs a close third, and is possibly the most common and well
developed method, but it does suffer a little from wear and sticking problems.
Magnetic actuators usually require relatively high currents (and high power), and on the
microscopic scale, electrostatic actuation methods usually offer better output per unit
volume (the limit is somewhere in the region of going from 1cm cubed devices to a few
mm cubed - depending on the application).
Thermal actuators also require relatively large amounts of electrical energy, and the heat
generated also has to be dissipated.

 When dealing with very smooth surfaces, typical of micromachined devices, sticking
or cold welding of one part to another can be a problem. These effects can increase
friction to such a degree that all the output power of the device is required just to
overcome it, and they can prevent some devices from operating at all.

Careful design and selection of materials can be used to overcome these problems;
but they still cause trouble with many micromotor designs.

Another point to be aware of is that when removing micromachined devices from

wet etch baths, the surface tension in the liquid can be strong enough to stick parts
together.

ELECTROSTATIC ACTUATORS
For a parallel plate capacitor, the energy stored, U, is given in equation 1 (where C is
the capacitance, and V is the voltage across the capacitor).

1
U CV 2
2

(1).

When the plates of the capacitor move towards each other, the work done by the
attractive force between them can be computed as the change in U with distance (x).
The force can be computed by equation 2.

V 2 C
Fx
2 x

(2)
.

Note that only attractive forces can be generated in this instance. Also, to generate large
forces (which will do the useful work of the device), a large change of capacitance with
distance is required. This has lead to the development of electrostatic comb drives
(figure 1-a).
2

Comb Drives:
These are particularly popular with surface micromachined devices. They consist of
many interdigitated fingers (figure 1-a). When a voltage is applied an attractive force is
developed between the fingers, which move together (figure 1-b). The increase in
capacitance is proportional to the number of fingers; so to generate large forces, large
numbers of fingers are required. One potential problem with this device is that if the
lateral gaps between the fingers are not the same on both sides (or if the device is
jogged), then it is possible for the fingers to move at right angles to the intended
direction of motion and stick together until the voltage is switched off (and in the worst
scenario, they will remain stuck even then).

Figure 1

Wobble motors:
are so called because of the rolling action by which they operate. Figure 13a,b shows a
surface micromachined wobble motor design. The rotor is a circular disk. In operation
the electrodes beneath it are switched on and off one after another. The disk is attracted
to each electrode in turn; the edge of the disk contacting the insulator over the electrode.
In this manner it rolls slowly around in a circle; making one revolution to many
revolutions of the stator voltage. Problems can arise if the insulating materials on the
stator electrodes wear rapidly, or stick to the rotor. Also, if the rotor and bearing aren't
circular (this is possible since many CAD packages draw circles as many sided
polygons), then the rotor can get stuck on its first revolution.

Figure 2

A problem with surface micromachined motors is that they have very small vertical
dimensions, so it is difficult to achieve large changes of capacitance with motion of the
rotor.
LIGA techniques can be used to overcome this problem - for instance the wobble motor
shown in figure 2-c,2 and 2-d, where the cylindrical rotor rolls around the stator.
4

Magnetic actuators
Microstructures are often fabricated by electroplating techniques, using nickel. This is
particularly common with LIGA. Nickel is a (weakly) ferromagnetic material, so lends
itself to use in magnetic microactuators. An example of a magnetic microactuator is the
linear motor shown in figure 3. The magnet resting in the channel is levitated and driven
back and forth by switching current into the various coils either side of the channel at
the appropriate time.

Figure 3
From figure 3, one common problem with magnetic actuators is clear: the coils are two
dimensional (three dimensional coils are very difficult to microfabricate). Also, the
choice of magnetic materials is limited to those that can be easily micromachined, so the
material of the magnet is not always optimum. This tends to lead to rather high power
consumption and heat dissipation for magnetic actuators. In addition, with microscopic
components (up to about mm dimensions), electrostatic devices are typically stronger
than magnetic devices for equivalent volumes; whereas magnetic devices excel for
larger dimensions.

Piezoelectric actuators
The piezoelectric effect mentioned previously for use in force sensors also works in
reverse. If a voltage is applied across a film of piezoelectric material, a force is
generated. Examples of how this may be used are given in figure 4. In figure 4-a, a layer
of piezoelectric material is deposited on a beam. When a voltage is applied, the stress
generated causes the beam to bend (figure 3-b).

Figure 4

The same principle can be applied to thin silicon membranes (figure 4-c). When a
voltage is applied, the membrane deforms (figure 4-d). This, when combined with
microvalves, can be used to pump fluids through a microfluidic system.

One problem with piezoelectric devices is making the films thick enough that high
enough voltages can be applied without dielectric breakdown (sparks / short circuits
across the film).

Thermal actuators
Thermal microactuators are commonly either of the "bimetallic" type, or rely on the
expansion of a liquid or gas.
In figure 5-a, a beam is machined from one material (e.g. silicon), and a layer of
material with a different coefficient of thermal expansivity (e.g. aluminium). When the
two are heated, one material expands faster than the other, and the beam bends (figure
5-b). Heating may be accomplished by passing a current through the device; heating it
electrically.

Figure 5

Figure 5-c shows a cavity containing a volume of fluid, with a thin membrane as one
wall. Current passed through a heating resistor causes the liquid in the cavity to expand,
deforming the membrane (figure 5-d).
Whilst thermally actuated devices can develop relatively large forces, the heating
elements consume quite large amounts of power. Also, the heated material has to cool
down to return the actuator to its original position; so the heat has to be dissipated into
the surrounding structure. This will take a finite amount of time, and may affect the
speed at which such actuators can be operated.
7

Hydraulic actuators
Despite problems associated with leaky valves and seals (a problem in many
microfluidic systems), hydraulic actuators have considerable potential as quite a lot of
power can be delivered from an external source along very narrow diameter tubes. This
has potential in areas such as catheter tip mounted microsurgical tools.
LIGA techniques can be used to fabricate turbines (as in figure 6), which can deliver
power to cutting tools.

Figure 6

Microstimulators

One further method of actuation is illustrated by the use of microelectrode devices to

electrically stimulate activity of nerves and muscles. Common designs for these devices
have already been discussed in the section on chemical sensors (see figure 9 on the
chemical sensors page). The use of microelectrode devices facilitates highly specific
stimulation of individual nerve fibres compared to other methods of stimulation; this
would allow finer control of the stimulation provided enough electrode sites can be
inserted into the tissue.
As relatively large stimulating currents have to be passed through the electrode sites,
microelectrodes for stimulation generally have geometrically larger electrode sites than
those for recording (500um.sq up, c.f. 16um.sq up). This is necessary otherwise the
currents involved will damage the electrode sites.
One area in which silicon microengineering is being applied in the hope that it will
result in a considerable improvement over more conventional electrodes is the area of
visual prosthesis - providing rudimentary vision for the blind. One project currently in
early stages of research involves a "forest" of silicon needles which will be inserted in
the visual cortex.
Early visual prosthetic devices involved an array of electrodes placed on the surface of
the visual cortex (brain). When activated, blind volunteers could see points of light
(phosphenes). These devices required relatively high currents to operate, however, and
the image was distorted by afterdischarges and interactions between groups of neurons.
This lead to the suggestion that a method for more selective stimulation of neurons
within the visual cortex was required to provide any functional form of vision. So this is
an area where microengineering technology has recently begun to be applied.

References and further reading

Introductory
Joe McEntee. Start Making Microsensors, Physics World, December 1993, pp 33-37.
J Bryzek, K Petersen, W McCulley. Micromachines on the March, IEEE Spectrum, 31 (5), May 1994,
pp 20-31.
A Heuberger. Silicon Microsystems, Microelectronic Engineering, 21, 1993, pp 445-458.

Introductory technical
IEE Colloquium on: Microengineering - the Future, held at the IEE, Savoy Place, London, 13 October
1993. Digest No: 1993/182.
IEE Colloquium on: Microengineering in Instrumentation, held at the IEE, Savoy Place, London, 16
November 1993. Digest No: 1993/218.
IEE Colloquium on: Medical Applications of Microengineering, held at the IEE, Savoy Place, London,
31 January 1996. Ref: 96/019
P Horowitz, W Hill. The Art of Electronics, 2nd edn., Cambridge University Press, 1989.
EA Parr (ed.). Newnes Electronics Pocket Book, 5th edn., Heinemann - Newnes, 1986.
JW Gardner. Microsensors: Principles and Applications, John Wiley & Sons, 1994.

Technical
DV Morgan, K Board. An Introduction to Semiconductor Microtechnology, John Wiley & Sons, 1985.
Notes from a two day short course on: Micromachining of Materials, held at the University of
Southampton, 25 & 26 March 1992. University of Southampton Institute of Transducer Technology.
L Ristic (ed.). Sensor Technology and Devices, Artech House, 1994. AEG Cass (ed.). Biosensors: A
Practical Approach, IRL Press at Oxford University Press, 1990.
EAH Hall. Biosensors, Open University Press, 1990.
RS Muller, RT Howe, SD Senturia, RL Smith, RM White (eds.). Microsensors, IEEE Press, 1991.

10