https://www.sciencenews.org/article/device-harnesses-cold-ni...
(They just painted aluminium black...)
----
EDIT: No, wait, that's a different team that the one I had in mind ? I seem to remember them using a complex material that specifically took advantage of this "infrared window" ??
----
EDIT2 : Ok, same guy(s), slightly different device :
https://www.asme.org/topics-resources/content/new-solar-ener...
> A thin wafer of germanium had the right properties: It is fairly opaque at visible wavelengths, absorbing most incoming sunlight, while being generally transparent at the mid-infrared.
> Because most of the energy in the solar spectrum is in the visible and near-IR range, Fan said, germanium could capture solar energy for use in thermal or photovoltaic applications, while allowing mid-IR energy to escape for radiative cooling.
> The Stanford team tested the concept with an experimental device that placed a germanium wafer in front of a mid-infrared emitter.
> As reported in a recent paper in the journal Joule, the wafer absorbed enough sunlight to warm up by 24 degrees Celsius, while the emitter sent enough radiation through the infrared “window” to cool itself by 29 degrees Celsius below ambient temperature.