This is more physics curiosity than anything else but since it uses semiconductor fab line technology in its manufacture is of interest. A passive device that uses layers of silicon and dielectric to make a mirror surface that becomes cooler when exposed to direct sunlight by careful management of its absorption and thermal IR emissivity profile.
Looks for all the world like a mirror.
It doesn't actually cool itself so much as not get as hot as it surroundings.
Hi Bill, It radiates strongly in the IR, and so "sees" the temperature of the sky at 8-11 um. (Which reading between the lines is less than 300 K.) I think it need to see the sky to work. (But that's just my reading of it.)
One of the Labs (LLNL?) invented an incandescent lamp filament with a multilayer coating that had high emmisivity in tne visible but low in far IR, which made it really efficient. It hasn't been commercialized, so I guess it isn't practical.
The thing above probably isn't practical either. Six inches of roof foam would work as well and be a lot more practical.
Selective-emissivity stuff is interesting. Paint would be cool. Or a really good window.
But hey, scientists should keep inventing press releases.
"They found that on a sunny day, the device cooled to between 4 and 5 degrees below the surrounding air temperature."
That sure sounds like real cooling to me. But not much. They didn't discuss the power density potential.
It would make slightly more sense to pump water into a black array during the day (use it for pre-heating domestic hot water) and, at night, pump cool water for tank storage, for air conditioning. But the power transfer density up to the sky is low, day or night.
It's hilarious that they tuned the wavelengths to avoid atmospheric absorption that might cause global warming.
Selective coatings of the other sort (black in the visible, reflecting in the IR) have been used for a zillion years to improve thermal solar collectors.
You have to be careful reading this sort of stuff--the second law of thermodynamics applies wavelength-by-wavelength.
Cheers
Phil Hobbs
I've seen things like that. The usual problem is that the special feature (coating, microstructure, etc.) isn't stable in service.
Cheers
Phil Hobbs
Cool! Do not really need silver; alternate selected-thickness layers can reflect any selected wavelength; multiple "stacks" can make that rather efficient, and a sharp wavelength "knee" can be easily and cheaply produced. Never mind that is already done and the price for low-pass,high-pass and bandpass filters is rather high. I know someone that made those in his basement at a cost of a few dollars each (time, materials, power all accounted for).
That seems very logical.
That's what I suspected. The surface of an incandescent filament isn't a very friendly environment.
Radiation into the near-absolute-zero night sky must be modest power density, or we and our plant friends would freeze in the dark.
Doesn't being lower temp than its surroundings equate to cooling itself?
In order for the Earth's surface temperature to be stable, the net radiation has to equal
(1 - albedo) * total insolation at the top of the atmosphere
The Earth's albedo is about 0.3, so the total radiation has to be about
0.7 * 1.3 kW/m**2 = 900 W/m**2.Cheers
Phil Hobbs
The PDF that I linked to cites some real-life surface-level cases, like 99 w/m^2 at low humidity, 25 for overcast conditions. Given that radiation falls off as T^4, you don't wind up with much of a cold-plate power source. And low delta-Ts are hard to extract power from.
The typical clear-sky night radiative temperature was calculated as
3.9 degrees C... NOT 3.9 degrees K!
It's amusing that the energy density of the core of our sun is estimated to be less than 300W/m^3. Similar order of magnitude to that of a human body- the sun is just really, really big.
Best regards, Spehro Pefhany
Depends on the wavelength. At wavelengths where most of the light gets through the atmosphere, you see more of the cold. At wavelengths well below the Planck peak, the radiation per hertz is linear in T. The other three powers of T in the Stefan-Boltzmann law come from the peak frequency going up linearly in T. That gives you one factor of T from the energy per photon, and two more from the increased volume in k-space.
Thus at a wavelength where the atmospheric transmittance is 50%, you get an effective sky temperature
T_sky = (2.1 K + 280 K) /2 = 141 K.
Cheers
Phil Hobbs
That's the energy production rate. The energy _density_ is pretty big--the sun's core temperature is about 15 million kelvins, iirc.
Cheers
Phil Hobbs
Not necessarily high! The high prices go with lab-certified compliance to (possibly difficult) requirements. It's easy to go to any glass shop in town and get a panel of 'low-e' glass, and there are mass-production facilities for plastic films with this kind of character. It's not always visible, you might be gazing through such a coating now as you admire the snow...
The net energy exchange at the surface of the earth averages about 340W/m^2
I attended the 2014 World Metrology Day Awards ceremony at NMI where Dr Bruce Forgan received an award for his work increasing the accuracy to which this has been measured - now down to under 1% error. They still don't know what contribution the nuclear-thermal contribution of the earth's core is, so they can't determine the balance of absorption and radiation exactly yet. You might find Forgan's work online somewhere though.
Clifford Heath.
Except for the factor of 1/4 for the intercepted area (pi r**2) vs surface (4 pi r**2), of course. (Oops.)
Cheers
Phil Hobbs
Have something to add? Share your thoughts — no account required.
Ask the community — no account required