Humidity sensing has been the subject of a great deal of study over recent years. The increased use of microelectronics in everyday life means that the importance of sensor technology in general is growing and will continue to do so. But the increasing miniaturisation of electronics means that associated circuitry is becoming cheaper and therefore the cost of the sensor is now becoming a more significant proportion of the cost of the overall sensor/actuator system. Therefore, the thrust of sensor research now is towards developing a better, cheaper sensor.

The main motivation behind this work is the need which has arisen in the automotive industry for a new low-cost, reliable humidity (r.h) sensor which would be suitable for placement in cars. The car environment is quite a demanding one and suitability for it puts special restrictions on this proposed sensor. The aim is to produce a sensor which fulufils all of the requirements of sensitivity, fast response, small hysteresis, small temperature co-efficient, long-term stability, durability and resistance to contaminants. Such a sensor would have a much broader range of applications in domestic life and in the workplace. There is presently a huge range of humidity sensors available, either commercially of as prototypes, catering to different markets. But none of them fulfils the above requirements satisfactorily. 

The problem of particular concern to us is the fogging which frequently occurs on car windows. This problem arises because of the differential in temperature between the glass and the car interior. The interior of a car is generally warmer than the exterior. This is partly due to the occupants, if any. But, during daylight, it is largely due to the heating effects of the sun on the trapped air in the car. Thus, while the relative humidity in the car may be reasonably low, at the cold windows a layer of air cools beyond the dew-point and so water vapour condenses onto the windows.

As an example, consider a mass of air in a car at 10^oC with a r.h. of 70%. This (ideally, at least) provides no condensation inside the car. However, if the glass temperature is 4^oC then the r.h. of that thin layer of air next to the glass is actually 100%. Thus some moisture will condense onto the glass. The greater the differential between the interior and exterior temperatures, the greater the degree of fogging. This poses at least an inconvenience to the driver and, in some circumstances, can prove dangerous, due to the loss of visibility it entails.

Present solutions involve the use of hot air blowers and rear window heaters to re-vapourise the water after it has condensed. The object of this is to detect the conditions which will lead to condensation before it happens. This will involve measuring the relative humidity (and, implicitly, the temperature) inside the car, and the temperature of the glass interface, since the degree of condensation depends on both these factors. These parameters must then be correlated and the range of values established for which condensation will occur. 

The measurement of the glass temperature presents no difficulties. Temperature measurement is very straightforward and there are many suitable (small, accurate, reliable) sensors available. However, the measurement of relative humidity is not as well established and sensors for humidity measurement are not as reliable as those for temperature measurement. Hence the need for this new sensor.

The new sensor proposed here has, like most humidity sensors, an electrical response to changes in humidity. It is proposed to use polyimide as the sensing material and silicon nitride as the sensor substrate. The use of a silicon substrate allows for the fabrication of a smart sensor. Associated electronics may be incorporated on the same chip, leading to enhanced miniaturisation and greater reliability and durability.

There are many reasons for selecting polyimide as the sensing material. It displays excellent thermal and electrical stability, with a low dielectric constant and a low equilibrium moisture content. It is resistant to irradiation, mechanically strong and chemically stable in the presence of most common contaminants, including automotive lubricants, coolants and greases. Also, polyimide processing is fully compatible with standard electronic processing procedures, an important consideration for cost control. All of these are important advantages which polyimide possesses over other moisture-sensitive materials.

A capacitive sensor using these materials has been produced. Interdigitated electrodes were fabricated on the silicon nitride substrate and these were the coated with the polyimide, as in Fig.1. Preliminary tests indicate that this device does have the potential for use as a humidity. The capacitance of the structure increases humidity. This response is more marked at the wet end of the scale. There is no observable temperature dependence. Hysteresis continues to be a problem and methods of reducing this are being examined. There is also scope for improving the sensitivity of the scale. Further work in these areas in now ongoing.