Abstract
Notwithstanding its relevance to many applications in sensing, security, and
communications, electrical generation of narrow-band mid-infrared light remains
highly challenging. Unlike in the ultraviolet or visible spectral regions few
materials possess direct electronic transitions in the mid-infrared and most
that do are created through complex band-engineering schemes. An alternative
mechanism, independent of dipole active material transitions, relies instead on
energy lost to the polar lattice through the Coulomb interaction. Longitudinal
phonons radiated in this way can be spectrally tuned through the engineering of
polar nanostructures and coupled to localized optical modes in the material,
allowing them to radiate mid-infrared photons into the far-field. A recent
theoretical work explored this process providing for the first time an
indication of its technological relevance when compared to standard thermal
emitters. In order to do so it nevertheless used an equilibrium model of the
electron gas, making this model difficult to inform the design of an optimal
device to experimentally observe the effect. The present paper removes this
limitation, describing the electron gas using a rigorous, self-consistent,
non-equilibrium Green's function model, accounting for variations in material
properties across the device, and electron-electron interactions. Although the
instability of the Schrodinger-Poisson iteration limits our studies to the
low-bias regime, our results demonstrate emission rates comparable to that of
room-temperature thermal emission despite such low biases. These results
provide a pathway to design a confirmatory experiment of this new emission
channel, that could power a new generation of mid-infrared optoelectronic
devices.