Explained: Thermoelectricity

Turning temperature differences directly into electricity could be an efficient way of harnessing heat that is wasted in cars and power plants.

Thermoelectricity is a two-way process. It can refer either to the way a temperature difference between one side of a material and the other can produce electricity, or to the reverse: the way applying an electric current through a material can create a temperature difference between its two sides, which can be used to heat or cool things without combustion or moving parts. It is a field in which MIT has been doing pioneering work for decades.

The first part of the thermoelectric effect, the conversion of heat to electricity, was discovered in 1821 by the Estonian physicist Thomas Seebeck and was explored in more detail by French physicist Jean Peltier, and it is sometimes referred to as the Peltier-Seebeck effect.

The reverse phenomenon, where heating or cooling can be produced by running an electric current through a material, was discovered in 1851 by William Thomson, also known as Lord Kelvin (for whom the absolute Kelvin temperature scale is named), and is called the Thomson effect. The effect is caused by charge carriers within the material (either electrons, or places where an electron is missing, known as “holes”) diffusing from the hotter side to the cooler side, similarly to the way gas expands when it is heated. The thermoelectric property of a material is measured in volts per Kelvin.

These effects, which are generally quite inefficient, began to be developed into practical products, such as power generators for spacecraft, in the 1960s by researchers including Paul Gray, the electrical engineering professor who would later become MIT’s president. This work has been carried forward since the 1990s by Institute Professor Mildred Dresselhaus, Theodore Harman and his co-workers at MIT’s Lincoln Laboratory, and other MIT researchers, who worked on developing new materials based on the semiconductors used in the computer and electronics industries to convert temperature differences more efficiently into electricity, and to use the reverse effect to produce heating and cooling devices with no moving parts.

The fundamental problem in creating efficient thermoelectric materials is that they need to be good at conducting electricity, but not at conducting thermal energy. That way, one side can get hot while the other gets cold, instead of the material quickly equalizing the temperature. But in most materials, electrical and thermal conductivity go hand in hand. New nano-engineered materials provide a way around that, making it possible to fine-tune the thermal and electrical properties of the material. Some MIT groups, including ones led by professors Gang Chen and Michael Strano, have been developing such materials.

Such systems are produced for the heating and cooling of a variety of things, such as car seats, food and beverage carriers, and computer chips. Also under development by researchers including MIT’s Anantha Chandrakasan are systems that use the Peltier-Seebeck effect to harvest waste heat, for everything from electronic devices to cars and powerplants, in order to produce usable electricity and thus improve overall efficiency.

Topics: Energy, Explained, Materials science, Physics, Thermoelectricity


It's outstanding, it could be used to harvest wasted heat from magnets of an aneutronic nuclear fusion reactor improving overall electricity conversion efficiency to virtually 100 percent.

As we look around at our world, I am wondering if there might be other, natural temperature differences that could be tapped into...

Like the different temperature between an attic and the outside air, or between the desert sand on a hot day and the temperature 3-6 feet down (geo loops would probably be pretty easy to dig, in sand!), etc. I imagine that extracting energy would result in moderating temperatures on each side. If that were to occur, it seems that a homeowner might simultaneously expect to get electricity and a cooler attic in the summer out of the deal. Then, there are the ocean's currents, carrying huge volumes of cool or warm water into other ocean areas, and industrial processes that use lakes to cool down their systems could employ cooling/ energy recovery operations, too.

The "Johnson engine" was looking into this a couple years ago -- "paired" fuel cells, where electricity could be "invested" in the lower temperature environment to pressurize hydrogen, which was to be circulated to the higher temperature environment, where a greater amount of electricity could be harvested with another fuel cell, and cycled back to the cool environment.

We might not even need smart grids, if such readily available energy sources could be harnessed.

With the advancing state of the art in thermoelectrics, what overall conversion efficiency from heat to electricity has now been demonstrated in the lab, and how does that compare to theoretical carnot efficiency?

No more than 3 to 5% of efficiency. It is really low for commercial Peltier modules.

To achieve Carnott efficiency figure of merit of 3-4 is needed. I think nowadays is less than 1.

Effiency of 100% is impossible, remember the 2nd law of thermodynamics.


Back to the top