An alternative for next generation cooled LWIR detectors is the antimonide superlattice detector made of alternating thin layers of InAs/GaInSb. This new material combines the performance of bulk detectors with the easy processing of III/V material and QWIP. With its possibility to reduce dark current and raise operation temperature and at the same time maintain high detectivity and short integration times, antimonide superlattices will significantly enhance the performance of LWIR detectors and as a consequence lead to a cheaper camera system.
Physical background
A superlattice is the same as multiple quantum wells with thin barrier layers
separating them. The barriers are so thin that the wave function of subsequent
quantum wells overlap and cause energy perturbation and level shifting. Electrons
can tunnel through the barriers and the barrier thickness gives an additional
degree of freedom in tailoring the electro optical properties of the material.
In
a type-II superlattice, the electrons and holes are spatially separated from
each other, see figure 1. In InAs, the conduction band edge has a lower energy
level than the valence band in GaInSb. But if the layers are made very thin
(typically 15-50 Å), the ground level of the InAs conduction band shifts
up and the valence band ground level in GaInSb shifts down and a bandgap is
created.
The structure is epitaxially grown on (in our case) GaSb substrates and the
individual layers are alternating tensile and compressively strained. The
average strain in the structure should however become zero. This is achieved
by using the correct layer thicknesses and In-content in GaInSb.
The
tension in the GaInSb valence band causes a separation of heavy and light
hole bands, a separation that often is larger than the superlattice band gap
itself. This reduces the so called Auger recombination, an otherwise troublesome
contributor to the noise. The tunneling probability between the valence- and
conduction band in the depletion region is also expected to be lower and hence
the noise and dark current will be reduced.
A photovoltaic
detector
Antimonide superlattice materials are, in contrast to QWIP and QDIP, preferably
used as photovoltaic detectors. These are pin-diodes where incoming IR light
with higher energy than the bandgap of the material excites electron - hole
pairs. If the absorption takes place at or close to the pn-junction, the electrons
and holes are separated by the built-in electronic field. Electrons are driven
towards the n-side while the holes go to the p-side, and this creates an electromotive
force, se figure 2.
When
the p-and n-sides are connected by an external circuit, a short-circuit current,
that is proportional to the light intensity, flows through the component and
this is used as the signal from a photovoltaic detector.
Processing challenges
InAs/GaInSb is a III/V material and standard lab processes can be used to
fabricate detectors from this material. Since the material absorbs isotropically,
no light coupling grating is needed and the number of process steps can be
reduced.
There
are, however, two known challenges when it comes to producing antimonide superlattice
detectors; removal of the substrate and passivation. The substrate removal
can be handled by leaving a small amount of the substrate, thin enough to
let the IR light through, but the passivation is trickier.
MWIR
antimonide superlattices can be passivated using standard processes, for example
Si02, but LWIR detectors are more difficult to passivate. The LWIR-material
has a lower band gap and the purpose of the passivation is that a material
with a high band gap will stop leak currents from going out through or being
lead by the mesa sidewalls, see figure 3.
Outlook
With IRnova’s high detector competence and long experience in III/V
processing we have the right ingredients to make antimonide superlattices
the next generation of cooled high sensitivity LWIR detectors.