In your basic heat transfer class, you should have learned that there are three primary modes of heat transfer: conduction (heat transfer through solids or stationary fluids), convection (heat transfer by virtue of moving fluids), and radiation. As a general rule, in electronics cooling situations, radiation is the least important mode, and is often neglected in first-order system design – not necessarily because there isn’t much radiation heat transfer, but that the other modes typically transfer a much larger percentage of the total energy. Electronics cooling in outer space, for instance, requires consideration of radiation effects, because that’s all there is between a spacecraft and deep space. What might surprise you, however, is that in typical Earth-bound free-convection environments, radiation heat exchange can account for as much as 35-40 percent of the total heat loss. If you neglect this in your system design, you’ll likely get a conservative value for how much heat you can reject from your system – so much so that you might decide you can’t get away with free convection, so you add a fan and start moving the air. Then, of course, the moving air convection effects begin to dominate, and the error in neglecting the radiation effects becomes small, and your design works the way you expect it to.
But what if you could improve the radiation efficiency of your system? This has always been a dream of mine, but the bottom line is, it’s remained more of a science-fiction dream. Why? Because the natural world just doesn’t seem to provide many options to get around the basic physics limitations. You see, radiation exchange is a two-way street: Every object loses thermal energy via radiation, so two objects which can “see” each other, as it were, send radiation back and forth. If they’re at the same temperature, all else being equal, each radiates an equal amount to the other, and each absorbs an equal amount from the other. Net exchange: zero. But the basic law of radiation transfer says that as an object’s temperature increases, the amount of radiation it emits increases; thus, a hotter object in view of a colder object unbalances this exchange, the net effect being that more energy moves from the hot object to the cool object. That’s exactly what you expect (and want) to happen, and as I mentioned above, if your only other mode of heat transfer is free-convection, and you’re talking about objects in the range of perhaps 150 °C exchanging heat with surroundings at perhaps 50 °C, the net radiation exchange might be as much as 40 percent of the free-convection loss. But in light of the other modes of heat transfer, to get radiation up into the 90 percent range of the total, or higher, you need incredibly high temperatures and reasonable temperature differences. (In the case of space applications, where the receiving object is deep space at -270 °C, all you need is room temperature, and you’ve automatically got several hundred degrees of difference.) Typical electronics cooling applications in room temperature rooms, however, simply can’t get there – unless you could take advantage of a subtlety in the radiation law, which is that the radiation that’s emitted and absorbed by an object is actually wavelength dependent. Enter photonic crystals, or photonic coatings, which may allow you to tune the wavelengths at which radiation is emitted and absorbed. For instance (you might think), what if you could make the hot object refuse to absorb radiation at the wavelength the cold object emits? The “gotcha” is that whatever fraction of the so-called blackbody radiation an object emits at a given wavelength is also the fraction it absorbs at that same wavelength, and when you work it out, it seems that no matter how clever you are, you basically can’t do much better than what nature already does.
Keeping the foregoing discussion in mind, then, a NanoLetters paper published by Stanford scientists this past April blew my mind (Ultrabroadband Photonic Structures To Achieve High-Performance Daytime Radiative Cooling, Rephaeli, Rama, & Fan); you can read about the development at http://news.stanford.edu/news/2013/april/fan-solar-cooling-041513.html. Basically, these Stanford scientists have designed a photonic coating to apply to a mirror, to give it this totally amazing capability: the coating is essentially transparent to visible light, but it emits strongly in the infrared bandwidth near room temperature – at which bandwidth, serendipitously, the Earth’s atmosphere is mainly transparent. So what, you ask? First consider what happens with a normal mirror when it sits in the hot sun. It’s not a perfect mirror, so it reflects maybe 97 percent of the sunlight, which means that it absorbs 3 percent – which is actually quite a bit of heat – and it starts to heat up. But because it’s a normal mirror, it’s also a pretty poor emitter of infrared (at any wavelength). To oversimplify a bit, it absorbs 3 percent of the energy from a 5000 °K object (the Sun), and emits only 3 percent of its own energy at roughly 300 °K. So it really loses this battle and it heats up significantly. In fact, the reason it doesn’t heat up any more than it does is because additional heat is lost by free convection to the ambient air surrounding the mirror (and even more heat loss if there’s a breeze blowing, i.e. forced convection cooling). OK, now the magic with the specially coated mirror: Because the coating is transparent in the visible, the mirror still reflects 97 percent and absorbs 3 percent of the hot Sun’s radiation; but because the coating is carefully designed to emit strongly in the atmospheric-transparent-infrared bandwidth, it wins the radiation exchange battle with deep space! In fact, for a realistically good mirror with this special coating, the net radiation exchange battle is actually in the mirror’s favor, and it cools down even in bright sunlight. It cools until it’s far enough below ambient that the heat input from ambient convection balances the heat loss to space. (It may now occur to you that most normal objects left out to view the night sky should cool to below ambient temperature. Exactly! That’s how dew can turn into frost, even when the air temperature is well above freezing.) So you can think of it like this: to the specially-coated mirror, it’s always night! The obvious application is for passive solar cooling of large structures – your house, for instance. In places like Phoenix, where I live, wouldn’t that be a kick? Coat your house with these special mirrors, and bye-bye summer air-conditioning bills! But I’m thinking really large structures, such as the Earth itself – that is to say, the next most obvious application, in my mind, is as a real engineering work-around for global warming. Whereas I don’t think it’s realistic for humankind to avoid continually increasing our non-renewable energy consumption, and its burden on the biosphere, I now see a way to at least offset it. For me, this is the most stunningly hopeful technological development I’ve run across in years.
The down side is, I still can’t see a way to make this work for my typical electronics cooling problems, but I’ll keep looking at it!
Till next time, be cool!
