3 - Structures

Semiconductor devices in sensors use silicon or compound semiconductors for fabricating transducer device parts. Their sensitive layers may be insulating, semiconductor or conductive materials. In sensor devices based on the field-effect principle, the semiconductor surface potential is modulated by potential or electric charge variations elsewhere in the structure, usually called gate, which is actually an insulator/(metal)/(sensitive layer)/environment interface. The thin insulating layer (SiO₂ and/or Si₃N₄) and the covering (metallization and/or sensing) layer is called the gate, and it is this electrode that mainly controls the action of the transistor. In the conventional MOS transistor, a bias voltage is applied to the gate, which is called the controlling gate voltage. If an increasing negative bias is applied to the gate, a positive charge is first induced at the silicon surface, and electrons are repelled away from this region. A depletion region results, which will deepen as the bias increases. A point called the turn-on (or threshold) voltage, Vt is reached when inversion occurs, and movable free holes are induced at the surface; thus, the transistor channel starts to conduct between the source and drain. In the sensor FET devices, the gate metallization and/or coating environment interaction produces the controlling potential (through charge and work function condition changes), which is a function of the quantity to be measured and can be detected only as a threshold voltage variation. Practically, these devices are operated at the threshold point by an external regulation, and the gate voltage variations, which are sufficient for the compensation, can be followed.

Temperature-based

Semiconductor pn-junctions show an analogous effect as thermocouples. The different electron band structure at both sides causes a rectifying behavior of the device. This diode has a temperature dependent voltage-current characteristic:
where U_D is the diode voltage; I its current; k the Boltzmann-constant; e is the electron charge, and I₀ is the reverse current. If the supply is a constant current source, the voltage is a linear function of the absolute temperature. This ideal behavior is only disturbed by the temperature dependency of I₀.

Light-based

Photodiodes are based also on semiconductor pn-junctions. They apply the current-voltage characteristic shift of the device due to the electron-hole generation caused by the absorbed light. Practically, the reverse biased current of the diode is measured. The devices have a built-in field enabling them to operate in the photovoltaic cell mode but have much better performance in the photoconductive mode. Photodiodes apply Si or Ge pn-junctions and compound semiconductors often with heterojunctions. Devices operated within the visible range are based on the latter materials.