6.5. Photodetectors 405
The photocurrent gain is then obtained by dividingisbyiγ. Hence we get
Gain =
is
iγ
=
τμE
d
=
τ
td
. (6.5.56)
Here we have assumed that the charges are produced at the surface and traverse
the thicknessdof the detector before being collected by the electrode in a time
td=d/vd=d/μE,withvdbeing the average drift velocity of the charges. tdis
generally called transit or drift time and is an important parameter for estimating
the gain of a photodetector. A well fabricated pn junction diode usually has a gain
of 1, which signifies the fact that the recombination is minimal. Avalanche diode
that we will discuss in the next section, may have a gain of 10,000 or more. The
reason why one would want the gain to be so high is not difficult to understand if
one considers the operation at very low incident photon fluxes. Even though most of
the photodiodes have a quantum efficiency of more than 80% throughout the visible
and near infrared regions, but since at nominal applied voltages there is no charge
multiplication process therefore their signal-to-noise ratio at the single photon level
is not large enough for obtaining meaningful results. Therefore, for low intensity
measurements, the signal must be somehow amplified. In avalanche photodiodes,
this is done internally in the bulk of the material through the process of charge
multiplication. This emerging technology has already started to take the place of
conventional PMTs in some applications and therefore deserves some attention. The
next section is devoted to avalanche photodiodes.
6.5.C AvalanchePhotodiodeDetectors(APD)............
Avalanche photodiode detectors exploit the process known asimpact ionization,
in which the initial electrons create more free electrons by imparting energy to the
molecules along their tracks. This process is very similar to the gas multiplication we
discussed in chapter 3. The primary electrons produced by the incident radiation
are made to attain high velocities under the influence of externally applied high
electric field. If the energy attained by an electron is high enough, it can free one or
more secondary electrons, thereby creating more charge pairs. It should be noted
that theoretically such a process is only possible if the incident electron gains energy
at least equal to the band gap energy of the material. However since an electron
also looses energy through non-radiative scatterings, on the average energy of the
electron should be much higher than the band gap energy. For most semiconductors
an energy difference of a factor of 3 is normally required. The secondary electrons,
being under the influence of the same electric field, produce tertiary charge pairs
and so on. Once started, this process of charge multiplication grows and eventually
causes avalanche multiplication of charge pairs. The large number of charge pairs
thus produced create an electrical current that is much higher than obtained in
conventional photodiode detectors. Since this current is proportional to the incident
particle energy therefore the detector can be used for spectroscopic purposes.