Question

3.7. W hy is the cutoff wavelength of a photovoltaic detector independent of the energy gap of the impurity levels (unlike an extrinsic photoconduc tor)?

Answer 1

In a photovoltaic detector photoexcitation of electron-hole pairs occurs near a junction when radiation of energy greater than the band gap is incident on the junction region. Extrinsic photoexcitation is rarely used in photovoltaic photodetectors. The internal energy barrier of the junction causes the electron and hole to separate, creating a potential difference across the junction. Other types of structure are also used, such as p − i − n, Schottky-barrier (a metal deposited onto a semiconductor surface) and heterojunctions. The p − n and p − i − n structures are the most commonly used. All these devices are commonly called photodiodes. Important photodiodes include silicon for detection of radiation between 0.1 and 1.1 µm and detectors based on the InGaAs(P) system for the region between 0.9 and 1.7 µm, which encompasses the important fiber optical communication wavelengths of 1.3 µm and 1.55 µm. Other photodiodes with more specialized applications include germanium between 0.4 and 1.8 µm, indium arsenide between 1 and 3.8 µm, indium antimonide between 1 and 7 µm, lead-tin telluride between 2 and 18 µm, and mercury-cadmium telluride between 1 and 12 µm. These spectral response regions are not all necessarily covered by a detector operating at the same temperature; for example, InSb responds to 7 µm at 300 K but to wavelengths no longer than 5.6 µm at 77 K. The wavelength response of PbSnTe and HgCdTe depends also on the stoichiometric composition of the crystal. All these photodiodes have very high quantum efficiency, defined in this case as the ratio of photons absorbed to mobile electron-hole pairs produced in the junction region. Values in excess of 90% have been observed in the case of silicon.

When a photodiode detector is illuminated with radiation of energy greater than the band gap, it will generate a voltage and can be operated in the very simple circuit. However, it is much better to operate a photodiode detector in a reverse-biased mode, where positive voltage is applied to the n-type side of the junction and negative to the p-type. In this case, the observed photo signal is seen as a change in current through the load resistor. A photodiode responds much more linearly to changes in light intensity and has greater detectivity when operated in the reverse-biased mode. Ideal operation is obtained when the diode is operated in the current mode with an operational amplifier that effectively holds the photodiode voltage at zero–its optimum bias point. The p−i−n structure is most commonly used in these devices because its performance, in terms of quantum efficiency (number of useful carriers generated per photon absorbed) and frequency response, can be readily optimized. These devices have very low noise and fast response.

In practice, the limiting sensitivity that can be obtained with them will be determined by the noise of the associated amplifier circuitry. If the reverse bias voltage on a photodiode is increased, photo induced charge carriers can acquire sufficient energy traversing the junction region to produce additional electron-hole pairs. Such a photodiode exhibits current gain and is called an avalanche photodiode (APD). It is in some respects the solid-state analog of the photomultiplier. Avalanche photodiodes are noisier than p − i − n photodiodes, but because they have internal gain, the practical sensitivity that can be achieved with them is greater.