Breaking barriers in solar thermoelectric devices

Written by Denis Paiste | 27 January 2013

Faculty Highlight: Professor Gang Chen

MIT Professor Gang Chen focuses on finding better ways to generate electricity from heat and his research is broadening the fundamental understanding of the nature of heat transfer at the atomic level.

As director of the Solid-State Solar-Thermal Energy Conversion Center (S3TEC Center), and its principal investigator, Chen’s recent breakthroughs include:

• Developing an optical technique to map out the traveling distance in solids of different phonons, which are high frequency lattice waves.
• Experimental demonstration of coherent heat conduction in superlattices.
• Showing a theoretical increase in thermal electric power by using invisible dopants, with Mona Zebarjadi, Assistant Professor Rutgers School of Engineering and former MIT postdoctoral associate, and others.
• Showing a theoretical cloaking of core-shell nanoparticles from conducting electrons in solids, with Bolin Liao, a graduate student at MIT, and others.
• Demonstrating a prototype of a high-performance flat-panel solar thermoelectric generator with high thermal concentration with Daniel Kraemer, a graduate student at MIT.

Besides serving as Carl Richard Soderberg Professor of Power Engineering in MIT’s Department of Mechanical Engineering and director of S3TEC Center, a U.S. Department of Energy (DOE) Energy Frontier Research Center, Chen is director of the Pappalardo Micro and Nano Engineering Laboratories.
The potential market for solar thermoelectric generators is huge, with more than 140 million square meters of vacuum tube solar hot water systems already in use in China.

The solar thermoelectric generator prototype demonstrated an energy conversion efficiency of about 4.6 percent without optical concentration and jumped to 5.2 to 5.3 percent with about 1.5 times optical concentration.
“What we’re pushing now is we are trying is to actually double the efficiency of what we have achieved, and also thinking combined with heat storage so that we can really supply electricity continuously in completely different ways,” Chen said in a December 2012 interview.
Chen is a co-founder of GMZ Energy, which is trying to commercialize the technology in combination with other systems.

Doctoral student Daniel Kraemer, right, and Professor Gang Chen display a prototype of a flat-panel solar-thermoelectric generating device. Photo: Melanie Gonick

Detail of solar thermoelectric generator. Image: Daniel Kraemer

Challenges of Thermoelectric
Converting electricity from the heat in sunlight through thermoelectric materials is different from what is currently done with thermal-mechanical engines. The attraction is that the conversion devices are solid-state, comparable to photovoltaic cells. A core challenge is the material’s performance. “In terms of thermoelectrics, we actually have to deal with the two basic energy carriers. One is the charged electrons and the other is the lattice vibration (atomic vibration) that conducts heat. Although we really only want the electrons to do the heat to electrical energy conversion, we have to deal with lattice vibration that leaks heat. ” Chen said. The quantum name for the lattice vibration is phonon.
A recent Science paper by Chen’s group analyzed through quantum mechanical calculation whether the phonons propagate in different materials as a wave or a particle. The paper, “Coherent Phonon Heat Conduction in Superlattices,” produced with MIT graduate student Maria N. Luckyanova, postdoc Jivtesh Garg and others, found that heat could travel like waves through nanostructures.
“This is of fundamental importance in terms of controlling how we can reduce their ability to carry heat,” Chen said.

Crystal versus Glass
A thermoelectric device is a solid-state system that converts heat energy to electric power by using a temperature difference across the material to drive electrons from the hot side to the cold side or to carry heat from one side to another when an external current drives the electrons from one end to the other. So thermoelectrics can be used for either heating or cooling.
However, in a thermoelectric material, what’s good for electric conductivity, or electron flow, is often in conflict with what’s good for thermal, or phonon, flow. The ideal thermoelectric material will act like a crystal structure, in which atoms are highly ordered, toward electrons but at the same time act like glass, in which atoms are loosely ordered toward phonons, as the heat carried by phonons are wasted and reduce the energy conversion efficiency. “We want the crystal and amorphous material in the same material.  In the past, all thermoelectric materials used alloys to reduce thermal conductivity,”  Chen said in his Materials Day presentation at MIT last fall.
One approach Chen has followed is to create structural barriers that reflect heat-carrying phonons much like an echo chamber reflects back sound waves. The reflection helps meet the goal of reducing heat loss through phonons.  
Chen achieved the desired reflective characteristics in one semiconductor material, bismuth telluride, by crushing the material into a nanoscale powder and recompacting it into a bulk form creating more interfaces at nanoscale to reflect phonons, which led to some improvement in thermoelectric performance.

Mapping phonon paths
Researchers want to know how far phonons travel, i.e., their mean free path, in different materials so that they can design proper nanostructures to effectively reflect these heat carriers. While neutron scattering can potentially map out the phonon mean free path distributions in single crystals, Chen and his collaborator, MIT Professor of Chemistry Keith A. Nelson, developed an optical technique so that they can achieve the feat in an optical benchtop system, and is potentially applicable to more materials. “We use a laser focused into different diameters as the heat source in the material of interest and measure the thermal conductivity. We found that the measured

Diagram shows the 'probability flux' of electrons, a representation  of the paths of electrons as they pass through an 'invisible'  nanoparticle. While the paths are bent as they enter the particle,  they are subsequently bent back so that they re-emerge from the  other side on the same trajectory they started with — just as if  the particle wasn't there. Image courtesy Bolin Liao et al.

thermal conductivity depends on the laser spot size, and realized that this is due to ballistic phonon transport when the laser spot size becomes smaller than the phonon mean free paths. By systematically changing the heating spot size, we found that we can map out the distribution of phonon mean free paths and how much heat they carry in a solid,” Chen said in his Materials Day presentation.
Another Chen collaborator, Boston College Physics Professor Zhifeng Ren, reported a higher thermoelectric output figure for a group of metallic alloys called Half Heuslers using a nanostructure approach.

Invisibility cloak and dopants
Another idea, an invisibility cloak, a phrase borrowed from Harry Potter, would allow electrons to travel undetected through nanoparticles made of a core material and a shell of a different material. “I can create a situation, where the electron really goes through the particle without being scattered, so they do not see the particle. We don’t know how far we can push along this direction, but this as I said you need to think out of the box. The invisibility idea is a little bit crazy, but it gives us some inspiration to see how we can potentially explore this,” Chen said. In the paper with lead author Bolin Liao in Physical Review Letters, computer simulations show the electron could pass through the nanoparticle and continue on its path instead of being scattered as it normally would. In a subsequent paper published in Advanced Materials online in January 2013, lead author Mona Zebarjadi, with Chen, and MIT 
Institute Professor Emerita Mildred Dresselhaus, and others, showed it is possible to exploit such invisibility to create nanoparticle dopants that would increase thermoelectric power output. This approach thus provides a pathway to engineer both electrons and phonons to benefit thermoelectric energy conversion.

Phase change heat storage
To compliment the solar thermoelectric device, Chen is working on a heat storage system based on a phase change from solid to liquid and back again.
“We are actually developing high temperature phase change materials that melt at high temperature and then it cools down during the night and releases the latent heat during the solidification process,” Chen said. Current research involves metallic alloys that melt between 600 and 900 degrees Celsius (1,112 to 1,652 degrees Fahrenheit). By concentrating solar energy, a solar thermoelectric system can heat up to 1,000 degrees Celsius (1,832 degrees Fahrenheit), Chen said.
The long-term goal is to bring the cost of solar energy down enough to compete with fossil fuels. The cost of electricity from solar panels has dropped from about $3 to 60 cents in just three years, but the installed electricity cost is still 10 times higher than a typical Boston area utility's energy cost and four times the utility’s consumer charge. But the rapid drop in solar cost has also led to tumultuous times for the solar industry in both the U.S. and rest of the world. “Hopefully something good will come out,” Chen said.