Most photodetectors such as photodiodes, phototransistors, CCD sensors and phototubes produce a signal current which is more or less proportional to the incident optical power. However, even in the absence of any light input, there is often some tiny amount of DC current, which one calls the dark current. An also possible fluctuating thermal current with zero mean value is usually not called a dark current.
For many applications, the dark current is totally negligible, but in some cases it matters – for example, when extremely small optical powers need to be detected. One may in principle subtract the dark current from the obtained signal either with analog electronics or with software, but that works only to a limited extent, because the dark current can be substantially temperature-dependent (see below), and it also exhibits shot noise.
Origin of Dark Current
The dark current of a photodetector can have different origins. In photodiodes and other detectors with some p–n or p–i–n junction, it is often caused by thermal excitation of carriers – not necessarily directly from valence to conduction band, but possibly through defect states related to crystal defects or impurities. The rate of such thermal processes depends not only on the active area, but also critically on the temperature and on the band gap energy of the material, and also on the operation voltage.
For visible light detectors such as silicon-based photodiodes, the dark current can be very small (e.g. in the picoampere region) (even for significant bias voltages) and is then negligible for most applications. Germanium photodiodes exhibit much higher dark currents which is however mostly not due to their somewhat lower band energy. Indium gallium arsenide diodes, which also have a reduced bandgap energy compared with silicon, also exhibit a relatively low dark current.
For materials with substantially smaller band gap, dark current can be a serious problem and may thus enforce the operation at substantially reduced temperatures. Therefore, some mid-infrared cameras, for example, need to be equipped with a Stirling cooler for operation around 100 K or even lower.
For operation near the break-down voltage, the dark current can become far stronger than for lower voltages.
Dark currents may also be generated by some leakage currents which are not related to thermal excitation.
In any case, a dark current can normally not occur for operation with zero bias voltage, since there is no energy supply available for it – at least as long as the temperature of the device is uniform, excluding any Peltier effects. Therefore, one may operate a photodiode, for example, with zero bias voltage in cases where influences of a dark current must be avoided.
Of course, drifts of output signals may also occur in related electronics, for example due to bias drifts of operational amplifiers. Therefore, a non-zero output signal does not necessarily indicate a dark current of the detector.
If you like this article, share it with your friends and colleagues, e.g. via social media: