Arrays of HgCdTe infrared photodetectors have been developed to satisfy stringent performance requirements for use in NASA’s Tropospheric Emission Spectrometer. The design of these detectors improves (relative to prior such arrays) manufacturing yield, current-versus-voltage characteristics, and low-temperature performance. The design is of the doublelayer planar heterostructure (DLPH) type but differs from conventional DLPH designs in that it features a distinctive lateral current-collection configuration.
Figure 1 depicts the basic DLPH configuration of one of these detectors. An important aspect of the DLPH approach is a planar p-doped/n-doped device geometry that includes a wide-bandgap cap layer over a narrow-bandgap base layer (which is the active device layer). The layers are all initially grown n-doped in situ by molecular-beam epitaxy (MBE), thereby ensuring that the critical heterointerface is never exposed to the ambient contamination that has plagued HgCdTe photovoltaic infrared detectors in the past. Then, within the area selected for each detector, arsenic ions are implanted through the cap layer into an upper sublayer of the base layer (which becomes a p-doped sublayer) to form a planar p-on-n photodiode. The implantation step is followed by a two-step thermal anneal in Hg vapor, which is needed to activate the arsenic doping.
A thin passivating layer of polycrystalline CdTe is deposited over the cap layer, holes for ohmic contacts to the p-on-n diodes are etched in the CdTe, and then the ohmic contacts are deposited. The wafer is then overcoated with ZnS, which serves to increase the adhesion of the metal interconnections to be deposited subsequently. Holes are etched in the ZnS to expose the ohmic-contact metal, then the metal interconnections are deposited. Finally, the back side (the lower side in Figure 1) of the wafer is polished and antireflection coated and the wafer is diced into separate detector arrays.
For fabrication as described above, localized defects give rise to area-dependent detector performance, such that conventional DLPH detectors with areas <10–5 cm2 perform considerably better than do detectors with larger areas: this is because reducing the electrical p/n junction area of a diode reduces the probability that a given defect will coincide with the junction. By taking advantage of lateral current collection, which takes place within a few diffusion lengths of a junction, one can obtain, for each detector, a large optically sensitive area with one or more small junction(s). Figure 2 illustrates the difference between the conventional DLPH current-collection configuration and the lateral current- collection configuration.
With proper design, as in the present HgCdTe detector arrays, the optical collection area of a detector can be substantially larger than its electrical junction area. Instead of one diode, each detector pixel contains several diodes formed as small-area implants separated by a distance of the order of one minority-carrier diffusion length and electrically connected in parallel via their ohmic contacts. Following this approach to design and fabrication, optical detection areas can be an order of magnitude greater than electrical junction areas and the advantages of reduced dark current, relatively high resistance, and relatively low capacitance can be realized.
This work was done by Carl Bruce, W. McLevige, and K. Vural of Caltech for NASA’s Jet Propulsion Laboratory. For further information, access the Technical Support Package (TSP) free on-line at www.nasatech.com/tsp under the Electronic Components and Systems category.
This Brief includes a Technical Support Package (TSP).
Lateral-Current-Collection HgCdTe Infrared Detectors
(reference NPO-21002) is currently available for download from the TSP library.
Don't have an account? Sign up here.