A Plastic Film That Might Replace Your Air Conditioner

19 Comments Read/write comments

A plastic film with amazing radiative cooling capacity

When it's hot out, we want cold. At night, we like to be able to turn on the lights. During the daytime, it can be hard to find the darkness. All these things — hot and cold, light and dark, day and night — can seem like opposites. Chinese philosophy suggests, however, that these opposing forces, known collectively as the yin and the yang, aren't separate. And science has proved it. Let me tell you about the latest yin and yang science and how it could revolutionize air conditioning.

Willis Carrier's air conditioner

For a hundred years we've been using these mechanical systems called air conditioners to remove heat and humidity from buildings. It was one of many revolutionary technologies to come out of the twentieth century. Because of air conditioning, places like Orlando and Phoenix have far higher populations than they would have otherwise.

An air conditioner works by cycling the indoor air through a box that contains a fan to move the air and a cold coil to cool and dehumidify the air. It pulls in air from the home, sends it over the cold coil, and then sends that cooler, drier air back into the house. Wonderful!

Air conditioners make us comfortable!

But what happens to the heat? The heat from the indoor air goes into that cold coil. The coil is filled with a refrigerant, a material that can get cold. After it absorbs the indoor heat, the refrigerant travels to the outdoor unit (in a typical split system). The outdoor unit uses a compressor and another coil to dump the heat into the outdoor air. Other types of air conditioners exist, but the vast majority are of the type I just described: split-system, air-source air conditioners. (For more detail on how they work, see my article, The Magic of Cold, Part 1 - How Your Air Conditioner Works.)

The compressor is the the big energy hog in an air conditioner. It creates a high pressure and circulates the refrigerant. They've gotten better over the years, but the process is still energy intensive.

The power of radiative cooling

Every bit of matter radiates energy away to its surroundings. Matter also absorbs energy from its surroundings. Whether the net energy flow is into or out of a particular object depends on two factors:  temperature and emissivity. The Stefan-Boltzmann Law in physics puts all this together.

We've discussed this before. I stated it a bit differently then, using as an example a naked person jumping on a bed. (Perhaps you recall my article Naked People Need Building Science?) The surface temperature of his body is higher than the surroundings, especially the single pane window. Thus, the net flow of heat is out his naked body and into the surroundings. 

Net flow of radiant heat depends on temperature

Back in the 1970s, when the energy crises spurred lots of energy innovation, radiative cooling was one of the ways people tried to save energy. In the solar energy class I took in grad school, I learned about one method that required a pool of water on the roof. The water was thermally connected to the indoors during the day while being covered and insulated from direct solar gain and ambient heat. At night, the cover came off and the water could radiate the heat it picked up from inside the house during the day,  sending it out to the cold night sky.

Nice idea, if you don't mind having all that water overhead. But it was limited by being able to get rid of that heat only at night. So we have a day and night problem. Light and dark. They got the yang to the yin but not the yin to the yang. How do we get both?

A metamaterial film that provides free cooling*

The peer-reviewed journal Science this month published an article with a convoluted title, Scalable-manufactured randomized glass-polymer hybrid metamaterial for daytime radiative cooling (behind a paywall; sorry), but a simple message. The researchers have developed a material that can radiate heat away at night and during the daytime. And it does so with an impressive cooling capacity.

The photo at the top of the article shows the material, a translucent film. It's basically a plastic film with tiny silicon dioxide spheres embedded in it. (Silicon dioxide is what quartz is made of, the main component of a lot of the world's sand, and used to make glass.) The spheres play a critical role in tuning the material to emit infrared radiation while not absorbing any of solar radiation that hits the material during the daytime. The paper goes into the the physics, including a discussion of phonon-enhanced Fröhlich resonances of the microspheres and extinction cross-sections, but I'll let you read you that those details if you choose.

The film and microspheres, together called a metamaterial, make up one part. It doesn't absorb solar radiation, instead letting those electromagnetic waves pass right through. So they put a reflective backing on the film to reflect the sunlight back out. That prevents whatever is behind the film (which would be a building in our case) from heating up.

How much for a ton?

What could turn out to be a really exciting discovery for the building community is their measured cooling capacity. They set up the film outdoors and measured how much heat it could radiate to the open sky. They used an electric heater to pump heat into the film and adjusted the rate to keep the film at the same temperature as the surrounding air. By adding just enough heat to keep the temperatures equal, the researchers say the "total radiative cooling power is therefore the same as the heating power generated by the electric heater."

They did this experiment in Cave Creek, Arizona in autumn. The solar irradiance at noon was more than 289 BTU/hr per square foot (900 W per square meter). Over three days, they found the cooling capacity to be:

  • 30 BTU/hr per square foot (93 Watts per square meter) at noon
  • 35 BTU/hr per square foot (110 W per square meter) over the whole 3 days

That's pretty amazing. What it means is that you could get a ton of cooling capacity (12,000 BTU per hour) with about 400 square feet of this film at their noon conditions. At night, you'd get more than a ton from those same 400 square feet. A typical new house these days has a cooling load of about one ton per 1,000 square feet of conditioned floor area. It shouldn't be hard to find enough roof area to get all your cooling needs met if this material turns out to be as good as it looks now.

About that asterisk

Yes, the title of the section before last is correct. The film itself does provide cooling with no energy input. But to cool a building, you'll definitely need to use some energy. Otherwise, you won't be able to get the heat in the building to the film on the roof. One good way to do this would be to use hydronic panels to pick up heat inside the house and circulate it up to where the film is on the roof.

So it's not a free lunch. Sorry.

What's new here?

Other researchers have looked at similar methods of radiative cooling. (See this article, for example.) What seems to be different here is that the previous materials were difficult to make and expensive to scale up. This one is a plastic film with embedded microspheres. The research team claims it will be much easier and cheaper to produce. Hence the first word of their title: scalable.

Yes, there are still a lot of questions. They did this work in a dry climate. What happens when you put that film on a roof in a humid climate? How well do they work on cloudy days? Will there be enough space on the roof for both this cooling film and photovoltaic modules? How long will the material last? What's the best way to get the building's heat up to the roof?

But the big question is this: Will this metamaterial eventually replace traditional air conditioners? I think it holds a lot of promise but it remains to be seen what happens. The best technology, even after being fully developed, doesn't always win in the marketplace. (Remember VHS versus Beta?)

One thing we do know, though, is that those Chinese philosophers had a lot of insight. So, too, do the University of Colorado researchers who led this study. Their names are Yin and Yang. Really!


Thanks to Lloyd Alter for bringing this article to my attention. You can read his article about it on Treehugger: New plastic wrap might keep buildings cool even when the sun is shining.

Related Articles

The Magic of Cold, Part 1 - How Your Air Conditioner Works

Naked People Need Building Science

Air Conditioner Sizing Rules of Thumb Must Die


Photo of cooling film by University of Colorado at Boulder. Photo of man in front of air conditioner by Qfamily, used under a Creative Commons license. Image of naked man and window by Energy Vanguard, with naked man by Pixel Addict on flickr.com, used under a Creative Commons license.


NOTE: Comments are moderated. Your comment will not appear below until approved.


Feb 21 2017 - 7:12am

Allison, are these the same microspheres that are contained in the coatings that we use in lieu of paint on our concrete buildings? I have been using these coatings for several years now and found them to be extremely durable and resistant to mildew in our humid climate. I had also read some claims that the microspheres were supposed to somehow reflect heat but discounted the concept.

Feb 21 2017 - 7:23am

Good question, Thomas. I don't know much paint and coatings so I don't have an answer for you. If the paint microspheres are the same type as in this new metamaterial, they don't reflect heat. They absorb it and then reradiate it.

Feb 21 2017 - 9:05am

We have a number of examples from South Carolina to south Texas and around the world of using passive cooling - though we help clients build super-insulated building envelopes over a concrete slab. Leveraging the phenomenon of the earth maintaining close to average annual temps of the environment (San Antonio and Columbia SC are in the low 70s and low 60s, respectively) and eliminating thermal bridging from traditional stud frame construction - which typically overpowers the cooling effect - we are able to create very comfortable environments in what can otherwise be very uncomfortable - and high energy using - locations.

I do look forward to further advances in new composite materials, however. As a company whose focus is on composite materials, it's quite exciting. I think I saw this article last week on an MSN news feed.

Feb 21 2017 - 11:09am

This seems as magical as cold fusion. I eagerly wait to see how this develops, if a commercial product can be developed, it sounds amazing. Your question about how a humid climate might affect it, is insightful.

As economic progress spreads to Mexico, India, Africa and other poor and hot parts of the world, this implies a major new technology alternative to Carrier's AC design and gigawatts of additional electric demand.

Feb 22 2017 - 3:36pm

Doesn't sound magical to me. What is being described is a high-E window, as opposed to a low-E window. A normal window to let light into the building has the objective of being low-E so that emission from what heat it absorbs into the house is subdued. In this case, you want the opposite. One question the article did not answer is what the temperature of the material is in full sunlight at, say 100F. Air circulation based on the usual AC setup is, after all, based on cooling 74F indoor air to 55F. One would expect the material's resting temperature to be a function of ambient temperature and insolation, and this would dictate the airflow through the device necessary to deliver heat to the device at a rate which would enable it to emit sufficiently. If there are conditions in which the resting temp of the material starts getting above 70F, you're going to be looking at which is less odious of an energy and mechanical design demand, supplemental AC or a boatload of airflow.

Feb 21 2017 - 1:25pm

Most environments require heating so I don't see this as really anything more than a 'special use' type application. If anything maybe a way to reduce the air conditioner's tonnage requirements.

But complications would result from say having a home with a 5 ton furnace (100,000 BTU) but only needing say a 3 ton AC because of a reduced load due to the film.

So what could happen is the film is used the AC is still 5 ton rated, but uses more speeds to off set the reduction in the load. If the higher load does occur at some point, the AC just merely ramps up to a higher speed.

If you still need heating then I really don't see this changing much except maybe for areas that don't call for much heating.

In addition to that... R22 Freon production ban goes into effect January 1, 2020. But the Montreal Protocol that was signed on Sept. 16, 1987 to ultimately ban the production of R22 happened 30 years ago.

Don't hold your breath... you probably won't see it in your lifetime.

Feb 21 2017 - 7:21pm

I was thinking the exact same thing, if it cools great, but at the same time in many parts of the world we need winter heating. Seems this technology would be useless where winter heating is required.

Feb 21 2017 - 4:22pm

"fully transparent to the solar spectrum"
"infrared emissivity greater than 0.93 across the atmospheric window."
Something odd here. Call it 0.93.
Solar spectrum includes UV, visible and IR. If the material has IR emissivity of 0.93 then its IR reflectance is 0.07 and its IR absorption should be 0.93. It will be absorbing 93% of arriving IR (directly from the sun and as radiated and reflected by objects nearby (sun-heated ground, etc.)). Then it's 93% emittance for radiating that absorbed IR away. Why wouldn't that have a net gain in heat over time? How can it have a net loss in heat over time?
Then if it's transparent towards wavelengths in the other direction (UV and visible spectrum), and a shiny metal backing is ~0.97 reflective and ~0.03 absorbing (and more likely 0.92 & 0.08, or worse), then the metal backing is taking that 0.03 in as heat but can only emit it (front and back) at 0.03 but should conduct it to the plastic film much more efficiently, so a token source of heat.
Then they say it's net is losing heat, and so much so that they can input heat from an electric heater at a rate that has the film staying at ambient air temperature.
It would sure be interesting to know how they're inputting that heat from the electric heater. And how much heat from the film can conduct through and off of the metal film backing; is it insulated?
The magic would be if it is IR transparent for absorption yet emissive at 0.93.
Darn. I may for the first time end up paying to see beyond a paywall...

Feb 21 2017 - 7:37pm

Hey Allison, April 1st is still a little over a month away.

Feb 21 2017 - 9:54pm

Well, Ron, I do occasionally throw in something wacky and untrue at other times of year. This one's true, though.

Feb 21 2017 - 8:59pm

I think the typical challenge of cooling with hydronics still stands. That is to pull the heat out of the house you still need cool pipes. Those can lead to condensation and the mirage of problems associated with that. However a building with low cooling demand by way of an effecient envelope could do the trick. Very neat stuff though, and a lot of potential if they can commercialize it. This would also be fantastic for facilities rejecting mild or highly exergetic waste heat such as from industrial processes Thank you Mr. Bailes for bringing this to my attention.

Feb 22 2017 - 1:35am

Not to be a spoil-sport but at 400ft2 per ton + hydronics or other system to move heat back and forth, I can't imagine this technology will ever be competitive in a high performance home. I have clients in CZ-3 that cool their entire house with an 800 watt mini-split, at less than $100/yr.

Let's see what these systems actually cost if and when they become real. In cooling dominated climates, anything-hydronics rules out most residential installers and supercharges mechanical costs.

Feb 22 2017 - 7:58pm
Oh, come on, David. You are, too, being a spoil-sport! Look, any type of compressor-based air conditioning system is going to have to move heat to the outside. Whether mini-splits, conventional split systems, or package units, you've got to move a fluid carrying the heat to the outdoors. Let's stick with split systems here.

The indoor coil picks up heat and transports it the outdoor unit in the refrigerant. The compressor is what moves it. With this film, you don't need a compressor. You need a circulator to move a hydronic fluid (probably not just water since it's going to the roof). Hydronic circulators are much more efficient than compressors. Your house with $100 of annual cooling costs might use only $50 with this system.

Yes, you'll still need some dehumidification here in latent-land and that'll add to the cost. But it's hard to build a high-performance house in hot humid-climates and even mixed-humid climates now without supplemental dehumidification.

Feb 22 2017 - 8:36pm

> Your house with $100 of annual cooling costs might use only $50 with this system.

So after 10 years, I've saved $500, and after 20 years, $1,000. Just how much do you imagine this system might cost compared to the 1-ton baseline mini-split in my baseline high performance home that might cost $7k or 8k, depending on ducting.

You're focusing on energy savings while ignoring what appears to be a very expensive heat transport and removal system. I'm conditioned to hear 'ka-ching whenever I see anything mechanical that's outside the norm, especially if hydronics are in play.

I'll eat my hat if this technology can ever compete in terms of life-cycle cost with a conventional compression-based system, at least not in my lifetime or yours. And as others have pointed out, you still gotta have the mini-split or some other system to heat the house.

I prefer to stick with conventional equipment and design approaches done well. Ironically, as we make our homes more efficient, the more difficult it becomes to justify the additional cost of the highest efficiency equipment. You can quote me on that :-)

Feb 22 2017 - 9:06am

Good day gentlemen, just came across you bog by accident. Reminds me of the "Tromb Wall" paSsive heating and cooling but not requiring any energy input. Wondered if you chaps had come aross it. Brian.

Feb 23 2017 - 9:10am

Why does it need to go on the roof? Can't this be a wall panel system?

What effect would this have as a passive wall panel system in a "perfect wall". Is it better to have this film on the exterior, or inside of the wall exposed to the air gap?

Any idea on cost? Could charities afford this in poor countries?

How does it hold up to dust, wind, rain, snow?

Feb 23 2017 - 10:12am

I know we're not there yet...but I keep wondering: practically, how would you install it? Laminate it to the roofing or something? Roll it out when needed like a curtain then roll it up again when not needed? (sort of partially kidding..) Among other considerations already mentioned, like how do you move the heat to it from the house (and how do you do that without drastically altering how roofs/attics are constructed, or do you drastically alter how roofs/attics are constructed?), if it's sheet plastic, what about it's vapor diffusion properties?

Mar 1 2017 - 8:31am

Allison...fascinating, thanks for posting, but wondering if I misunderstood something. I would like to look at the main paper more closely as I'm having the most trouble with this sentence: "The spheres play a critical role in tuning the material to emit infrared radiation while not absorbing any of solar radiation that hits the material during the daytime". I agree with Michael, something does seem a bit off here. A = E, but what these values are varies considerably by the material type and wavelength. Would be interested to know what part of the spectrum they're dealing with here as well, but as far as I'm aware, you can't have something that has a high emissivity and a low absorptivity at the same time at a specific wavelength.

Mar 6 2017 - 12:50pm

Very interesting. I'm not surprised. It seems there are always ideas and advances in every field out there. I'm not sure this would be good for us HVAC Contractors though! lol ;) We would either have to evolve with these new inventions or suffer some loss of business, possibly.

Add new comment