Skip to Content

Turning Waste Heat into Power

Research shows that silicon is as efficient as pricier materials.
January 14, 2008

Silicon, in the form of photovoltaic cells, is good at generating electricity from sunlight. New research shows that it could also make a good thermoelectric: a material that converts heat into electricity and vice versa. Since silicon is more abundant than the leading thermoelectric materials and has a vast manufacturing infrastructure behind it, it could eventually yield cheap devices for generating power from engines’ waste heat or from solar heat.

Cool customer: This image, produced by a scanning electron microscope, shows a rough silicon nanowire bridging two heating pads–one serving as a heat source and the other as a sensor. Researchers have found that 50-nanometer-wide silicon nanowires have drastically lower heat conductivity than bulk silicon but retain their electrical conductivity. Thus the nanowires show potential as thermoelectric materials–ones that convert heat into electricity and vice versa.

In this week’s Nature, University of California, Berkeley, chemistry professor Peidong Yang and his colleagues report having fabricated silicon nanowires that generate electricity when a temperature differential is applied across them. Until now, silicon has been considered a bad thermoelectric material. But according to Yang, “the performance of the nanowires is already comparable to the best existing thermoelectric material.”

Thermoelectric devices have been around since the early 1960s, usually made from either bismuth telluride or lead telluride. They are used mainly for cooling: when a voltage is applied across a thermoelectric material, it gets hotter on one side and cooler on the other. Thermoelectric coolers are popularly used in portable picnic coolers and cooling car seats.

But more exciting applications lie in energy efficiency and energy generation. Thermoelectrics could be used to convert waste heat generated by car engines into electricity. Even more attractive is the idea of thermoelectrics’ harnessing the sun’s heat to create electricity. But bismuth telluride and lead telluride are not efficient enough, so devices made from them are costly as well as bulky, because they require more material.

Thermoelectrics would have to be at least twice as efficient as they now are to be used for cheap power generation, says Mildred Dresselhaus, a thermoelectrics pioneer and physics and electrical-engineering professor at MIT. Using nanoscale structures instead of bulk crystals of the materials can increase their efficiency, she says. Nanostructures block the flow of heat but allow electrons to flow easily. But processing and nanostructuring bismuth telluride is not easy.

Silicon, on the other hand, “is much easier to process, has a big processing infrastructure behind it,” Yang says. “Silicon also has a much lower cost than bismuth telluride.” The problem with silicon is that it is a bad thermoelectric. A good thermoelectric needs to be two things: a good electrical conductor and a bad heat conductor. Silicon conducts both heat and electricity very well.

Yang and his colleagues reduced silicon’s thermal conductivity by using silicon nanowires. They fabricated an array of silicon nanowires that are between 20 and 300 nanometers in diameter. Nanowire synthesis often involves liquefying a nanoparticle and inducing it to grow, much like a hair. But that produces nanowires with smooth surfaces. The chemical etching method that Yang’s team uses results instead in nanowires that have rough surfaces. The researchers found that wires that are about 50 nanometers wide retain electrical conductivity but have only one-hundredth the thermal conductivity. This results in a thermoelectric efficiency close to that of some commercial bismuth telluride materials.

No current theory explains why the nanowires’ thermal conductivity goes down so drastically. One of the reasons, Yang believes, is that the nearly one-dimensional nanowires and the wires’ rough edges block the flow of phonons, which are particles that carry heat. But the complete picture remains unclear.

Ali Shakouri, an electrical-engineering professor at the University of California, Santa Cruz, says that researchers will have to understand how the physics works before they can improve the technology enough to produce commercial devices. Furthermore, using nanowires for energy conversion and power generation has its own limitations, Shakouri says. Such applications require large arrays of nanowires, but in the Nature paper, Yang and his colleagues measured the electrical properties of individual nanowires. The researchers will have to make sure that those properties translate to entire nanowire arrays, says Shakouri: “Variations and interactions between nanowires could take away some advantages.”

Still, he says, “this is important work that could have a big impact.” Shakouri points not only to the demonstration of silicon’s potential as a thermoelectric, but also to the unique engineering that the researchers used to make rough nanowires. “The new way of playing with material properties is very interesting,” he says. “It could open up a way to improve thermoelectrics that could be applied to other materials.”

Yang and his colleagues, meanwhile, are already thinking about how to improve their nanowires’ performance. They plan to reduce the size of the nanowires and make their surfaces rougher than they already are. That should enhance their thermoelectric properties, Yang says. The researchers also plan to make and test an actual thermoelectric device using silicon nanowires.

Keep Reading

Most Popular

Large language models can do jaw-dropping things. But nobody knows exactly why.

And that's a problem. Figuring it out is one of the biggest scientific puzzles of our time and a crucial step towards controlling more powerful future models.

The problem with plug-in hybrids? Their drivers.

Plug-in hybrids are often sold as a transition to EVs, but new data from Europe shows we’re still underestimating the emissions they produce.

Google DeepMind’s new generative model makes Super Mario–like games from scratch

Genie learns how to control games by watching hours and hours of video. It could help train next-gen robots too.

How scientists traced a mysterious covid case back to six toilets

When wastewater surveillance turns into a hunt for a single infected individual, the ethics get tricky.

Stay connected

Illustration by Rose Wong

Get the latest updates from
MIT Technology Review

Discover special offers, top stories, upcoming events, and more.

Thank you for submitting your email!

Explore more newsletters

It looks like something went wrong.

We’re having trouble saving your preferences. Try refreshing this page and updating them one more time. If you continue to get this message, reach out to us at with a list of newsletters you’d like to receive.