2 Form and function, method and material
One aspect of micro and nano scale engineering that distinguishes it from many other forms of manufacturing is the way it involves building both the devices themselves and the very materials from which they are made, in one place and at more or less the same time. In general, MEMS are made from thin layers of new material produced, and then shaped in some way, on the surface of a silicon wafer. The devices contain several different materials, and have a three-dimensional structure that is defined by building and shaping these layers one by one. This means that form and function, method and material are intimately interlocked. Imagine the complexities that would be added to motor-car manufacture if the iron had to be smelted and converted to steel alongside the spot-welding robots, the window glass moulded, toughened and bonded in situ, etc.
Silicon's electronic properties make it a suitable material for integrated circuits, and it is manufactured in large numbers in the form of highly polished circular wafers, sliced from a single crystal.
To illustrate the MEMS combined ‘form and functionalise’ route without having to describe the operation of some complex system, I'll use a miniaturised hot-wire sensor as an example. This type of sensor might be found in a number of applications where some aspect of its environment affects the transport of heat to or from the wire. A hotwire anemometer measures air flow by measuring the cooling effect of a flow of air over the wire – the cooler the wire, the faster the flow must be. The device I'm going to describe was designed specifically to be a Pirani sensor (see Box 1) for gas pressure measurement, though it could also be used as an air-flow sensor. Figure 2 shows one particular design.
The sensor's most important component is a thin metal track that acts as a heating element. The track is supported on a stiff beam that spans a pit in the substrate. The cooling of the element is chiefly by conduction into the gas, which is able to get all around the heated structure. Heat flow along the beam should be negligible.
Box 1 Pirani sensors
Pirani sensors are used for measuring gas pressure in the range from atmospheric (105 Pa) down to a rough vacuum (say, 1 Pa). The principle of the sensor is that a heated wire is cooled by the gas surrounding it. The ability of the gas to do this depends strongly on its pressure. By passing a fixed electrical current through the wire and measuring its temperature or, alternatively, by measuring the amount of power needed to maintain the wire at a constant temperature, the pressure of the gas can be deduced.
In a conventional Pirani sensor, the hot wire forms one resistor in a Wheatstone bridge. This arrangement allows the resistance of the wire to be measured with great accuracy. The relationship between resistance and temperature of the wire is almost linear, so resistance is a direct indication of the average temperature of the wire. The measurement is then effectively one of voltage (or, equivalently, resistance).