While the frequencies are a tad too high for anything currently discussed on SED, the principle is familiar.
The following are likely behind a paywall, but all major libraries carry Science. And drafts of the articles may be findable.
"Room-temperature mid-infrared detector", REUVEN GORDON, SCIENCE 2 Dec 2021 Vol 374, Issue 6572 pp. 1201-1202 DOI:
10.1126/science.abm4252. This is the summary, and points to the two articles discussed, which are in the 3 Dec '21 issue of Science Magazine."Continuous-wave frequency upconversion with a molecular optomechanical nanocavity", Continuous-wave frequency upconversion with a molecular optomechanical nanocavity", WEN CHENPHILIPPE ROELLI et al, SCIENCE 2 Dec 2021 Vol 374, Issue 6572 pp. 1264-1267 DOI: 10.1126/science.abk3106.
The basic scheme is a single gold nanosphere (the ball) resting on one (flat) two (V-grooves) gold surfaces, with a monolayer of Biphenyl-4-thiol molecules in between ball and surface at the contact point or points. The Biphenyl-4-thiol molecules act as a parametric converter, allowing a near-IR pump beam to upconvert a Mid-IR signal up to visible, where it is easily detected.
As I understand it, in this parametric converter, no electron current flows. This is not a diode.
Case 1: The signal is at 32 THz (9.3 micron). The pump is far higher.The output is around 437 THz.
Case 2: MIR is 8.5 to 15 microns, from a Quantum Cascade laser. The pump is 785 nm, with a Acousto-Optical Modulator.
I don't fully understand the mechanism, but they talk of Stokes and Anti-Stokes sidebands, which sounds like a big clue. A form of four-wave mixing?
Joe Gwinn