[bardlehel]

I am curious to know if there are any molecules (proteins in particular) that react to voltage changes in such a way to emit photons (radio or microwave frequencies).

[Enthalpy]

If you hope to just inject DC electricity and harvest RF just by some magic behaviour of a molecule, my answer is "probably not".
Because creating an AC field requires some non-natural actions.
Because I'd probably have heard of it, as RF is one of my former professions.

Now, special molecules being part of more complex devices that produce RF: certainly.
Masers have been created using hydrogen, ammonia...
Atomic clocks use atoms or ions, like rubidium, caesium...
Infrared lasers use electric pumping of gas, optical pumping of dyes, chemical production of excited species...

In these cases, the molecules' aim is to define a precise frequency, not to produce efficiently any significant power, for which specialized devices are far better: magnetrons (microwave ovens) and other vacuum valves, semiconductors.

[408]

The heavy metal azides make a flash of light when a suitable potential is applied to a crystal... 

But in all seriousness...

Photons are easily emitted from organic and inorganic compounds....for example polydioctylfluorne is a blue light emitter when put under a DC voltage, and there are dozens of other polymer and non-polymer types.

Blue emitters are an intense area of research, due to higher energy transitions needing to exist to produce the light...  Microwave spectroscopy is also known (characterization of rotational modes), so presumably if a volatage could excite a rotational mode, it may be what you are looking for...

[Enthalpy]

Interesting!

I've already played with light emitting thin panels (probably OLED) but I considered the layer stacking was fundamental as it is in a semiconductor Light Emitting Diode. Hence my answer "not just by the material" as I suspect it needs more refinements.

Some modes are in the RF domain, sure, but...
- These modes are intensely excited by room temperature (300K = 10µm = 30THz >> RF). Would an electrical excitation add any desire by the molecule to radiate? This is a difficulty at Quantum Cascade Lasers, though they use well-known materials and emit in the infrared.
- In a liquid, the mean time between molecule collision is much shorter than a period of RF, meaning that these modes are so much broadened that they can't be seen on a spectrum. Even gas at moderate pressure shows a very indistinct spectrum: atmospheric humidity "resonating" at 2.45GHz (hence microwave ovens) has a very broad absorption. Can a device exploit such a degraded resonance?

In quantum devices (like LED) people generally consider that the working temperature must be much lower than the used frequency, which would mean <<0.3K for 30GHz and <<0.3mK for 30MHz. Are exceptions known in this area?

One device made to emit THz waves lets two Quantum Cascade Lasers produce IR and brings the slightly different frequencies to beat in a nonlinear crystal. This is the turnaround found to work at moderately cold temperature.

-----

People interested in devices creating or receiving THz waves could have a look at my inventions there:
http://www.physforum.com/index.php?showtopic=15617&st=0&#entry225095

I apologize for the huge mess there: it also contains ring lasers and partially related topics, and no drawing. And I contradict some early statements later.
Sorry for that. A better redaction may come some day, or maybe not. Anyway, I believe the future of THz is in that thread.