Putting the Heat on Inefficient Electronics
Posted by Lauren Rugani on December 28, 2009
Most of you have probably heard somewhere or another that light bulbs release 90 percent of their energy as heat and only 10 percent as light. Considering that the primary purpose of light bulbs is to provide light, not warmth (with the exception of the Easy Bake Oven), Edison’s claim to fame seems to be appallingly inefficient. There are, of course, more efficient alternatives to lighting – fluorescent bulbs, LEDs and even quantum dot lighting – but these aren’t relevant to the rest of the modern technologies that lose large amounts of energy as excess heat.
One solution lies in a phenomenon known as the thermoelectric effect, or converting a temperature difference between two surfaces into electricity. Scientists and product manufacturers are no strangers to the idea, however. Thermoelectric devices exist in the form of portable refrigerators and electronic component coolers, and the technology is gaining traction in the transportation and power plant industries. There is a theoretical limit to just how efficiently a machine can convert heat to electricity, but current applications reach only about ten percent of this limit – faring no better than the 130 year-old light bulb.
New research out of MIT suggests that, with existing technology, thermoelectric systems can reach up to 40 percent of that limit, with the possibility of reaching 90 percent as technologies continue to develop. The problem facing devices now is a trade-off between high efficiency and high throughput – either a system produces very little power very efficiently, or a lot of power a lot less efficiently and a lot more expensively. Peter Hagelstein, an associate professor of electrical engineering at MIT, says his design can provide both now that technological developments have caught up to the theory he began working on in 2002.
His design is based around a type of semiconductor where electrons and holes – the individual charge carriers – are tightly confined in three dimensions. By controlling each aspect of the device they got a better understanding of how to ideally convert thermal energy to electric energy, and found that the key adjustment was to reduce the separation between the hot surface and the converter.
As the necessary devices become commercially available over the next few years, Hagelstein guesses that the first applications of his theory will be in computer chips, providing cell phones with longer talk times and lap tops with better battery lives. Ultimately, it could find a place in cars and airplanes, or be incorporated into power plants to help capture and reuse a significant portion of wasted energy.