In this section we briefly describe the operational principles of the
NICMOS3 detectors. Figure 7.1
shows the basic physical structure of a photovoltaic HgCdTe detector.
An infrared detector is basically a photodiode, the core of which is a
p-n-junction created during the wafer processing. The Fermi-levels of the p- and n-type materials, i.e. the highest occupied energy state of the electron gas within the semiconductor material, must match, which effectively creates an electric field across the junction. The incident infrared photons free electron-hole pairs into the conductance band at or near the junction which are immediately separated by the electric field. The accumulated charge carriers cause a voltage change across the junction which can be detected and used as a measure of the incident light. One can think of the detector as a capacitor that is discharged by the infrared photons. In practice, the voltage change is monitored by a Si field effect transistor (FET), used as a source follower amplifier. Figure 7.2
shows the equivalent circuit diagram for the NICMOS3 “unit cell”.
In order to produce an “imaging” detector, a large number of such unit
cells, or pixels, are combined into an array. The photon-sensitive layer (HgCdTe in the case of NICMOS3 detectors) and the Si-multiplexer (which contains the array of FETs) are combined in a “hybrid” structure, connected via tiny indium bumps (Figure 7.3
). For better mechanical stability, the “hybrid” array structure is put on an infrared-transparent Sapphire substrate. Since each pixel contains its own FET, there is no “bleeding” along columns, as in CCD chips, and bad pixels do not block the rest of the column.