Tuned radiative cooling for PV solar panels?!

I read a similar article about Stanford. Seems pretty much sci-fi. Although stranger things can be true.

It is just hard for me to believe someone found a material that is clear enough to pass the spectrum of light that a pv cells like yet reflect the longer IR waves.

More details at

Definitely more details.

They still need to figure out how to either block the wavelengths that generate heat or a way of quickly radiating the heat that is collected.

I am still not sure what they propose will work under all sunlight conditions or what the cost would be for each solar panel.

Still trying to eliminate heat which is a detractor of pv cell efficiency is a step in the right direction.

OK, I think I get it now.

Roofing materials are rated for solar reflectance and thermal emissivity.
That latter number just means how well it radiates heat.
eg gold and some other metals have low emissivity, hence films of them can be used as radiant barriers.

Ideally you want the solar panel's cover to efficiently radiate heat.
As the paper mentions, glass does already; they're just trying to improve on that a bit.

See also this book about spacecraft power


The references to the cold of space still sound kind of silly, though.

I like the idea of the silica pyramid that would in effect become some type of heat "radiator".

What I am trying to get my head around is how hard would it be to create that type of surface, how much does it cost and will the additional cost be justified when compared to estimate improved performance of a cooler pv cell?

Still sounds like a pipe dream of researches in the Lab. I know. I've been there and done that.

This silica pyramids are not acting like a radiator. A radiator does not radiate heat: as my old high school physics teacher used to say, they should really be called convectors. The purpose of fins or other surface variation in a radiator is to increase the surface area over which heat is conducted between the radiator and the surrounding fluid. The silica pyramids are actually increasing the rate at which the panels radiate heat. But I don't fully understand their explanation of how that works!

says (about a laser, not a photocell, but mechanism may be similar?):
"High emissivity is obtained using inverted pyramids on the top surface, formed by anisotropic etching. These pyramids not only reduce reflection, but, more importantly, also increase emissivity by trapping weakly absorbed light within the cell."

That's gibberish to me (how does trapping light increase emissivity?), but then, I'm not a physicist.

OK that one I actually do understand. Obviously trapping more light increases the absorptivity of the material, and by Kirchhoff's law the absorptivity must equal the emissivity for any given frequency/temperature, so it also increases the emissivity.

The explanation for the pyramids in the paper you linked was a bit different: "the pyramids provide a gradual refractive index change to overcome the impedance mismatch between silica and air at a broad range of wavelengths, including the phonon–polariton resonant wavelengths." I do not know what a phonon or polariton is and I'm pretty vague about the use of the term impedance mismatch in this context; so yeah, gibberish!

Ah, that makes sense, thanks.

So the holy grail would be a material that has very high emissivity for IR wavelenths longer than those from the sun, which is exactly what the paper is about.

It must be fun to work on that stuff.

So, here is the physicists explanation in a nutshell, in hopes that it will make it easier to understand what the science publicists may not be describing well or accurately.

1. A so-called black body radiator will absorb all radiation that hits it and emit light with a spectral distribution which is based entirely on its temperature.
The classic approximation of a black body which is a good reference for practical measurements is a large cavity inside a block of metal that is connected to the surface of the metal by a small hole.
All light hitting the hole goes in and regardless of how reflective the inside might be (and you make it as unreflective as possible) it will bounce enough times to be absorbed before it chances to be aimed back out the hole.
Note that the surface of the metal does not behave like a black body, just the small area of the hole.
2. There are four characteristics of a surface which are frequency dependent:
a. emissivity (or how easy it is for radiation, based on the temperature of the material, to leave the surface,
b. reflectivity, which is the amount of incoming radiation at a particular frequency that bounces off the surface instead of being absorbed,
c. transmission/transmitivity, which is what passes through the material and comes out the other side without reaction.
d. absorbtivity which is just 100% minus the sum of the reflectivity and transmittivity (because any photon must do one or the other.) For non-transparent/translucent materials we can ignore transmission.

It is a physical fact which can be theoretically derived that emissivity and absorbtivity are locked together. You cannot increase one without increasing the other.

So, to get a material to absorb heat more strongly that it radiates it, or vice versa, you have to take advantage of the frequency spectrum in and out.

For example, a heat radiating surface will be subject to incoming radiation with a color temperature of around 10,000 degrees K.
If you can block the absorption (encourage reflection) of the higher frequencies (up into ultraviolet) which make up the majority of the sunlight it will not heat up as much.
If you then encourage absorption of infrared frequencies where most of the emission will take place for an ambient to solar heated temperature, you will emit light/heat very efficiently in that part of the spectrum where most of the black body radiation emission lies.

End result, you reject the majority of the incoming solar heat and still allow effective cooling by radiation.
That is where the "tuned" emissivity comes into play.

BTW, incoming photons whose energy is converted to electrical energy by moving electrons around does NOT heat the panel. But since the panel is only 20% efficient in converting incoming light it will heat up almost the same when open or short circuited as when under MPPT load.

Quite. The only remaining somewhat mysterious bits are

- is the reference to the "cold of space" a canard? Presumably it's good to emit in the atmosphere's "transparant region", 8 - 13 μm, as there's little ambient irradiation at that wavelength, so being an efficient absorber there doesn't heat up the cell as it would at other wavelengths? If so, why do they talk about using layers that emit at >= 4 um?

- how pyramids increase emissivity (is that phys 2? Did I miss a lecture?

- should one really get excited about two degrees more cooling than plain old silica, i.e. glass?

Maybe this'd make more difference in concentrating solar cells...

I think the significance of the 4 um is that below that wavelength increasing emissivity will cause more heat to be absorbed than is emitted so the panel will get hotter. Whereas above that wavelength the reverse will be true. As was pointed out above, this is a function of the different temperatures, and hence emission spectra, of the sky and the panel.

This stuff is a bit obscure, but it is certainly possible the next big leap in panel efficiency may come through reducing panel temperature. And it wouldn't surprise me if some installer knowledge/expertise becomes important in maximizing performance.

The big question is what is the cost of any enhancements to the pv cell and can the increase production from the cell improvement off set the cost of that improvement?

This is what has been researched in the pv industry for over 40 years. How can I get more output without driving up the cost of the cell production?

The two easiest paths to solve this problem is the reducing the cost of the materials that the cell is made out of or reducing the cost of production of the pv cell.

So most labs have looked into finding a solid state material that yields more electrons when exposed to a specific amount of sunlight.

The other direction is finding a faster & cheaper way to make the cell besides growing one from a silicon ingot. That is where thin film material and strip manufacturing became hot. That along with solid state paints (something DuPont looked into)

Yup. Hopefully those pyramids are easy to fabricate with standard photolithography, and make sense as part of a solar panel.

If not, there are probably a couple dozen other little tweaks being worked on at any one time, and one or so of them might make it out of the lab and into production in any given year, bumping up efficiency by a tiny percentage. (e.g. the smaller and more numerous busbars coming into use lately).