Temperature dependence of pressure

The most frequently used calibration procedure is based on temperature dependence of pressure of saturated mercury vapour [19,39-41]. At 25°C this pressure is of 0.0018 mm Hg height it corresponds to the vapour density of 20 pg/1. To get in the measurement cell a mercury concentration of about 10 ng/1, the saturated vapour should be strongly diluted. Instead of dilution, a lower temperature can be used however, the density of saturated vapour of 10 ng/1 corresponds to the temperature of less than —40°C. Both dilution and temperature decrease can be realized easily in laboratory conditions but their incorporation into a miniaturized chemical sensor is rather complicated. An attempt to develop such a device is reported in Ref. [41]. An additional problem in application of these techniques in portable sensing devices with integrated calibration is the necessity to have a reservoir with mercury in the device it complicates recycling of these devices and does not correspond to modern trends in technology. 

These coefficients give p in Mbar when T is in volts (i.e., in units of 11,605.6°K). The expression for b rj) is an analytic fit to values of b obtained from the Thomas-Fermi-Dirac zero-temperature pressure. The electronic temperature dependence is calculated from the zero-temperature pressure by assuming the latter to be the pressure of a degenerate Fermi gas, and then determining the temperature dependence of pressure of a Fermi gas at any density of interest. This procedure is equivalent to obtaining b from temperature dependent Thomas-Fermi-Dirac calculations. 

In this paper, we present the structure and shape of the new thin-film pressure sensor and results of oil-film pressure distribution measured in an engine main betuing and in a piston pin-boss. In addition, a new sensor structure is reported, using aluminum-nitride to reduce the temperature dependence of pressure sensitivity in high temperature range. 

---