Guest Post By Willis Eschenbach [See two Updates at the end]
Here’s an oddity. Some very clever folks have invented a plastic film that cools surfaces by as much as 10°C. From Science magazine:
Cheap plastic film cools whatever it touches up to 10°C
Here’s the innovative part, according to the article. The tiny glass spheres act as resonators for the infrared emitted by the underlying surface. By choosing the right size spheres, the frequency of the resonators is tuned to be that of the so-called “atmospheric window”. This is the band of frequencies that is not significantly absorbed by any of the greenhouse gases. Infrared (IR) at that frequency pretty much slides right past the water vapor, the carbon dioxide, the methane, the ozone, it misses everyone and goes straight out to space.
In other words, it dodges the greenhouse effect …
Now, I’m left with some questions.
First, is it possible-to frequency-shift infrared radiation in this manner?
Next, what does the emission curve for this material look like? As an example, here’s a typical curve from MODTRAN showing the absorption of upwelling longwave radiation:
The smooth colored lines in the upper right panel show the Planck blackbody emission curves for various temperatures. The uppermost green curve is the warmest, 300 kelvin. The lowest yellow curve is 22oK. The “atmospheric window” is the area from wavenumber 750 to 1250, interrupted in the middle by the ozone absorption band just above wavenumber 1000.
As you can see, the warmer it is, the more the peak of the Planck curves (smooth colored lines) is shifted to the right. Now, with the resonator the peak radiation is supposed to be shifted by the resonators to a wavenumber of around 1000. That’s just below the ozone absorption band.
So I’m very curious about the shape of that curve. If the peak shifts towards the right it would have the characteristics of a warmer surface … can you mess with the Planck curve like that, shift the peak? Not saying it’s impossible, metamaterials have bizarre properties, I’m just out of my wheelhouse here.
Finally, what can this be used for? Well, I had a scheme a while ago for solar distillation of water. This would have been very useful to cool the condensing side of the still.
More directly it seems like it could cool buildings. A coating that could cool a large building by even one degree would translate into big savings in air conditioning. Ten degrees would be marvelous.
Anyhow, that’s what I’m calling a reverse greenhouse effect … it concentrates the radiation on the band where there is minimum atmospheric absorption by greenhouse gases.
Best to all,
w.
My Usual Request: If you comment please QUOTE THE EXACT WORDS YOU ARE DISCUSSING. That way we can all understand your subject.
[UPDATE] Thanks to a tip from the commenter Johanus, the underlying paper is here. It has what I asked for above, the actual emissivity curve in the thermal IR range. Fascinating. Here’s a preview, a graph of the temperatures throughout the day:
Now, that is a beautiful thing for a couple of reasons.
One is that I love real data. It is so much more interesting that a computer model of the same thing. Facts. Observations. If I stick to the facts I know I can’t go far wrong.
Next, look at the photonic radiative cooler. Throughout the day it is running cooler than the ambient air temperature by something on the order of 5°C … so for all the folks who said it was impossible, there’s an old Soviet joke about a Political Commissar berating someone and saying “Yes, yes, Comrade, you’ve proven that it works in practice … but it will never work in theory!” …
[UPDATE 2] After many helpful comments I’m finally understanding what’s happening. It’s not so much related to the selective emission of longwave radiation (thermal infrared). Instead, Kirchoff’s law says that frequency by frequency, emissivity equals absorptivity. So selective emission in a narrow band also means selective absorption in the same band.
The selective absorption is important because the “atmospheric window” also means that there is very little downwelling radiation in that window. Here’ MODTRAN again, showing the downwelling radiation from the viewpoint of the surface looking up:
Now, we can see that as expected, we have a lot of downwelling radiation. With the given parameters shown at the left, it’s shown at the top right as “Iout”, about 260 W/m2.
But notice … almost none of that is in the atmospheric window. The photonic material selectively absorbs mainly in that window … but there’s almost nothing in that window to absorb.
This is how they get the large temperature differences shown in the underlying papers. The material simply absorbs poorly where the incoming longwave radiation is, and absorbs well in the window where there’s little radiation.
At least that’s my current understanding …
w.
At first glance, this material would appears to violate the 2nd law of thermodynamics. Of course it doesn’t, and after some thought I believe I understand why.
Imagine a house with photovoltaic cells on its roof. The electricity from these cells is used to power an air conditioner, which cools the interior of the house.
Obviously this is perfectly doable, and does not violate any law.
This material might do something analogous, in that it uses energy from photons to pump heat. It uses a completely different mechanism — it does not generate electricity or run an A/C — but the end result is the same.
How is this material not an example of Maxwell’s demon? This material needs photons to work. As soon as the sun goes down, it stops cooling the underlying material. Maxwell’s imp required vanishingly little energy to sort fast particles from slow, and thus was a thermodynamics scoff law.
Killer Marmot February 13, 2017 at 6:39 pm
How is this material not an example of Maxwell’s demon? This material needs photons to work. As soon as the sun goes down, it stops cooling the underlying material.
Not true, read the paper, it continues to cool even at night time, the photons it uses are IR photons from the surface.
“As soon as the sun goes down, it stops cooling the underlying material. ”
Not necessarily. If the film allows radiation from the underlying surface to excite the resonance of the glass beads then, overall, radiation through the air is increased compared with the surface in the absence of the film. However, the silver film might well prevent that occurring. It needs numbers to find out. Clearly, there are limits to the conditions over which it works.
“can you mess with the Planck curve like that, shift the peak?”
Yes by changing the temperature of the emitter you change the curve. It is the plastic film that’s emitting IR not the original surface. Presumably the film has different temperature than the surface.
Painting a surface white will change its emissivity and reflect all the spectra of visible light. It has cooling effect.
Thanks to a tip from the commenter Johanus, the underlying paper is here.
That link is to a letter to Nature about 2 years ago, it is a precursor to the paper referenced in the head post, which was:
Y. Zhai et al., Science 355, 6325 (9 February 2017)
The link to the paper is:
http://science.sciencemag.org/content/early/2017/02/08/science.aai7899
I have a chance to read the paper thanks to Johanus. As an academic paper I am impressed with the use of computer models. As an engineer, I can not see any practical application. It is Standford after all.
Quotes are from the paper.
“We demonstrate the performance of the photonic radiative cooler on a clear winter day in Stanford, California…”
With a few wind leaps of logic we put it on a roof in Phoenix.
“We assume a twenty-year lifespan,…”
On a roof in Phoenix?
“similar multilayer coatings for low-emissivity windows and other surfaces. ”
Window are different than roofs.
“which are below aggressive levelized cost projections for both rooftop and utility-scale photovoltaics”
Everything is better than PV.
When comes to saving energy, there is clearly a case of diminishing returns. After doing all the cheap things, there is not much saving left. For example, reducing AC energy consumption $500/month to 100/month leaves only $100/month to work with. The materials for my do it yourself project cost $200 and saved 10% of $10/month. Less than a two year payback period.
A couple years ago, I researched metal roof coating material. I selected a more expensive white coating not to save energy but to seal pin hole leaks in a storage shed. It still is not leaking. I used the same coating on an old camping trailer that is used a few weeks a year. I could have saved $100 using the product I used on my camping trailer 30 years ago. While clearly I will not save enough on energy to make up the difference. Now I just have to live long enough to see if last as long as claime.
The point is a roof is a sturdy structure in a harsh environment to prevent expensive water damage. Saving relatively small amounts of energy is only important to idiots at places like Stanford.
The academics do the research then other people make something of it. No reason to think these academics have the best ideas on how to use it. Look at lasers. (No, on second thoughts, don’t look at lasers – bad for the eyes!)
will be great for beer mugs
Ice cream scoops are made with some sort of aluminum alloy that are the opposite of this. – i.e. the scoop warms the ice cream as it cuts through more easily than a scoop made of some other material.
Cool!
Just good materials science. Seems there is still room for improvement. If they can dope the mix to skew the peak of absroptivity/emissivity out of the ozone bands, it will work even better. The cross section of the material in the paper is far more elaborate (and expensive) and bears no resemblance to the “pop culture” headline image.
gymnosperm February 14, 2017 at 8:41 am
Just good materials science. Seems there is still room for improvement. If they can dope the mix to skew the peak of absroptivity/emissivity out of the ozone bands, it will work even better. The cross section of the material in the paper is far more elaborate (and expensive) and bears no resemblance to the “pop culture” headline image.
That headline image is from the actual paper, it’s figure 4, as is the second image shown in the original post (Figure 1A). I suspect you are reading the wrong paper, the link given by Johanus is not the paper referred to in the Science article but a precursor to it from a couple of years ago, the actual material is described at the link I gave above at
https://wattsupwiththat.com/2017/02/13/a-reverse-greenhouse-effect/#comment-2425306
Thanks.
I’ve seen papers that compare satellite equilibrium temperatures when painted or covered with various materials and the coating or covering does affect operating temperature and this effect depends on the balance between albedo and IR emissivity. So an increase in IR emissivity in “clear window” wavelengths while preserving high albedo should have an effect.
It can work, but only with clear sky. Clouds would stop the effect. To work it needs a clear wiev to the cold space. The bottom of clouds would be only slightly colder than the roof and would radiate too much energy back in the same wavelenths.
Play with that http://hyperphysics.phy-astr.gsu.edu/hbase/quantum/radfrac.html#c1
The sun would dissipate ~10W/m2 in the active band (5 to 10um), and in the same band at 293K it would radiate ~100W/m2.
It is not exotic except that the material is very wavelenth selective.
The part that I misunderstood at first was I thought that it was cooling by radiating more in the range where it’s not absorbed by the atmosphere.
But I see now that it is the other way around. Emissivity equals absorptivity on a line-by-bline bases. So in addition to selectively radiating in the frequency band of the atmospheric window, it also ABSORBS SELECTIVELY in the atmospheric window …
… and under a clear sky, there’s very, very little in that window to absorb. So it is radiating the same amount of IR … but it is absorbing much less IR.
Best to you all, it’s a sunny bright day here, I can only wish the same for everyone. Well, unless you need rain …
w.
Willis,
I got and read the original paper. You are correct that it is the use of a film that absorbs/emits strongly in the 8-13 micron range, but not in the cisible range which is key. Turns out the polymer film they used just happens itself to have pretty strong absorbance in that wavelength range while perfectly transparent at visible wavelengths. Still, at 50 microns film thickness the plain film would have some transmission between 8 and 13 microns, and so less than perfect emissivity. The added “glass microspheres” (actually amorphous SiO2, or silica) have tremendously strong absorbance in the same wavelength range, so adding these (independent of their size) increases net emissivity. The only ‘interesting’ thing about the paper is the selection of the size of the particles. In general, interaction of particulate materials at or near the wavelenght of incident light maximizes the interaction of the light with the particles. The claim made by the paper is that the selected size of the silica particles is key to the emissive properties of the film. But neither the paper itself nor the supplimental information provides any real evidence of the importance of the particle size. They do provide a spectrum showing high emissivity in the 8-13 micron range for the film with ‘resonant’ particles, but provide no reasonable ‘control samples’ for comparison. The specific data they should provide but fail to:
1) No measured infrared spectrum of the plain film (without particles)
2) No measured infrared spectra of the same type of film with silica particles at a few different ‘non-resonant’ sizes.
In other words, they don’t actually measure the contribution of ‘resonant particles’ to the emissivity of the film, they only do calculations. Had I been a reviewer, I would have asked for these measurements.
Based on the published infrared spectrum for the polymer they used, along with the IR spectrum of silica, my guess is that you could accomplish exactly the same result without using ‘resonant’ particles, even if there is a significant contribution due to ‘resonance’, by just increasing the volume fraction of silica, increasing the thickness of the film, or both. A commercially produced film that is twice as thick actually is less expensive… probably because calendaring a very thin film is more difficult.
The University of Colorado says the group is applying for a patent… it will be interesting to see what the patent looks like, because this is a pretty well plowed field.
That was my reaction as well. Without a control test of shiny film without “resonance spheres” the results are not very interesting.
When solar radiation, with a 5800 K blackbody spectrum encounters the device, the device first reflects most of the incoming solar (albedo) and thermalizes the rest at ambient temperature, heating it. But instead of re-radiating as a blackbody which would be mostly reflected downward by greenhouse gases, it modifies the spectrum via a resonance so more can escape to space via the “radiation clear window,” hence the cooling effect.
Of course, the device will not work inside because in an enclosure such as an interior room or an oven or under dense clouds, all objects come to the same equilibrium temperature (Kirchoff’s radiation law) regardless of albedo, emissivity absorptivity. You need a high temperature source to activate the effect.
pochas94 February 14, 2017 at 2:31 pm
Thanks, Pochas. I was with you up to the last sentence, viz:
Actually, it will also work at night. Remember that it emits the same amount of IR … but it absorbs less. So all it needs is an unobstructed view of the sky to end up cooler than the surroundings.
The underlying principle is quite old. In the desert areas of the US Southwest, back in the day people would put shallow flat trays of water on a low rack on the flat adobe roofs common in that area. Even though night-time air temperatures don’t go below freezing, In the morning the trays are filled with ice.
This utilizes the same idea of an “atmospheric window” to get the water below air temperature, but in this case the “window” is due to the dry desert air. With no water and a clear sky, you’re basically exposed to outer space … add to that a slight bit of evaporative cooling (not a lot because of low temperature) and you get ice. Take it off the roof, bury it in sawdust to keep it frozen … old school.
Regards,
w.
I agree – it does work at night.
As the paper shows the cooling effect is stronger at night. See Fig 4:
” (C) The continuous measurement of radiative cooling power over three days shows an average cooling power > 110 W/m2 and a noon-time cooling power of 93 W/m2 between 11am – 2pm.”
============
Even though night-time air temperatures don’t go below freezing, In the morning the trays are filled with ice.
This utilizes the same idea of an “atmospheric window” to get the water below air temperature, but in this case the “window” is due to the dry desert air. With no water and a clear sky, you’re basically exposed to outer space …
============
Why does the ice drop below zero, but not the dry desert air?
Evaporate cooling potential is maximum when air is dry.
Could that be part of the equation that produces ice when ari temperature is above zero?
One of the most important lessons I learned from my tutors as a grad student, was when they chose to say nothing to a speaker with grand claims.
michael hart February 15, 2017 at 11:20 am Edit
Thanks, Michael, but I don’t understand this. It seems that you think someone is making “grand claims” … who is making said claims, and where? Me? The authors of the first paper? The authors of the press release? Some random commenter? The comment immediately above yours? WHO ARE YOU MUMBLING ABOUT AND WHAT DID THEY SAY???
This is why I ask people to QUOTE WHAT THEY ARE DISCUSSING.
In this case, you appear to be saying that you have some special insight that lets you divide the world into something you call “grand claims” and some other category. Since you have not identified the other category, I’ll call them the “not-so-grand claims”.
Then, bizarrely, you seem to think that the proper response to “grand claims” (whatever they may be) is to “say nothing”, but apparently you’re willing to discuss “not-so-grand claims” … say what? How on earth does staying schtumm help anyone? How does that improve understanding? How does that move science forwards?
I have no clue what you are referring to in this post as being a “grand claim”, or what a “grand claim” is on your planet, or why you think allowing them to render you speechless is the best way to deal with them …
Sorry, but you are making no sense at all. What am I missing?
w.
Won’t work for buildings, too much mass, the interior temperature change would be undetectably small at the new equilibrium.
Might be fun for humans, if it’s flexible.
At the very least, the contribution from the resonant spheres (as opposed to the film) is probably undetectably small due to insulation beneath the alleged “100W cooling effect.” Think about it for a minute: would you heat your house by putting a heater on top of the roof?
Also, the whole “beams IR directly to space, saving YOU big $$$ on cooling!” doesn’t really make any sense. What do you care if the IR is absorbed a hundred feet away or a million? It’s still not coming back.
@Willis @Micro6500 @Killer Marmot @george E. Smith
There is a fluorescence phenomenon known as anti-stokes radiation.
It can be used for cooling since it extracts phonons from its matrix, and essentially adds the phonon energy to the absorbed radiation for re-radiation as fluorescence. It can be observed with aggregates of dyes that efficiently fluoresce.
https://www.rp-photonics.com/optical_refrigeration.html
http://www.sciencedirect.com/science/article/pii/S0022231306000214
talldave2 February 15, 2017 at 1:14 pm
Thanks, Dave. Let me repeat what I said above.
The part that I misunderstood at first was I thought that it was cooling by radiating more in the range where it’s not absorbed by the atmosphere.
But I see now that it is the other way around. Emissivity equals absorptivity on a line-by-bline bases. So in addition to selectively radiating in the frequency band of the atmospheric window, it also ABSORBS SELECTIVELY in the atmospheric window …
… and under a clear sky, there’s very, very little in that window to absorb. So it is radiating the same amount of IR … but it is absorbing much less IR.
Average downwelling longwave radiation on a 24/7 basis is on the order of 340 W/m2. This film ABSORBS LESS of that downwelling radiation.
I’ll add another update to the head post explaining this,
Best regards,
w.
Thanks Willis, I did see that, and I agree, and I was pretty sure you understood too — my comment was directed more at the many commenters who did seem to think beaming IR to space was key.
The absorption claim is more plausible, but I won’t believe the effect is actually measurable until they do the control with just the shiny film. It just seems unlikely that equilibrium condition is going to change much, nature abhors a non-Maxwellian thermal distribution.
I’ve added the following to the head post:
[UPDATE 2] After many helpful comments I’m finally understanding what’s happening. It’s not so much related to the selective emission of longwave radiation (thermal infrared). Instead, Kirchoff’s law says that frequency by frequency, emissivity equals absorptivity. So selective emission in a narrow band also means selective absorption in the same band.
The selective absorption is important because the “atmospheric window” also means that there is very little downwelling radiation in that window. Here’ MODTRAN again, showing the downwelling radiation from the viewpoint of the surface looking up:
Now, we can see that as expected, we have a lot of downwelling radiation. With the given parameters shown at the left, it’s shown at the top right as “Iout” at about 260 watts per square metre (W/m2).
But notice … almost none of that is in the atmospheric window. The photonic material selectively absorbs mainly in that window … but there’s almost nothing in that window to absorb.
This is how they get the large temperature differences shown in the underlying papers. The material simply absorbs poorly where the incoming longwave radiation is, and absorbs well in the window where there’s little radiation.
At least that’s my current understanding …
w.