Are Ferroelectrics the Overlooked Thermoelectric Materials?

Controlling the transport and storage of thermal energy could potentially have a powerful impact on our lives. After all, everything you can think of has thermal energy. Atoms vibrate, molecules collide, heat energy radiates from the sun in the form of light waves. We use heat to create electricity through the phase change of water into steam in power plants.     

The scientific community has gotten very good at controlling the movement of electrons. We know how to make materials that slow their movement, that store their energy, and redistribute electrons at electrodes to do work in slow batteries or fast capacitors. But are there other classes of materials we could use to control heat much as we can control electrons? The answer might be a type of material that we have overlooked, a class of materials called ferroelectrics.

Penn State is widely known for the study of ferroelectrics, a class of crystalline materials that shows a spontaneous electric polarization that can be reversed when an electric field is applied. Ferroelectric materials are used for nonvolatile memory in computers, in pyroelectric applications for infrared sensing, and in piezoelectric applications such as sensors and capacitors, of which trillions are produced every year.

Clive Randall is one of the scientists who studies ferroelectrics at Penn State, following in the footsteps of a long line of highly regarded researchers and teachers who have made Penn State a leading center for ferroelectrics and the larger class of materials called piezoelectrics – materials that develop an electric charge in response to mechanical stress – and pyroelectrics – materials that develop an electric charge as a result of a change in temperature.

Ferroelectricity is analogous to ferromagnetism, which is used to read and write data in digital electronics. Where magnets have north and south magnetic poles, which can be flipped to store the zeroes and ones of binary information, ferroelectrics have coupled positive and negative charges that form dipoles that can also be collectively flipped. The two most commonly used ferroelectrics are barium titanate and lead zirconium titanate (PZT). Both of these materials were discovered more than 60 years ago, and belong to the perovskite family-a related material SrTiO3 is one of the highest performing thermoelectrics.

Several years ago, Randall began to wonder if there might be some way to use the deep knowledge they had gained about ferroelectric materials and apply it to the control of heat energy. Looking through some of the classic works on ferroelectricity, he was surprised to find that the telluride family of materials was among those listed as ferroelectrics.

“It didn’t seem appropriate,” he said of his revelation. “Usually, ferroelectrics are applied to insulators when you want to minimize electrical conduction. But a good thermal material has high electron conductivity, like a metal. When ferroelectrics become highly conductive, it destabilizes the long range dipolar ordering necessary for ferroelectricity. Is it even possible to have a metallic ferroelectric?”

There were little clues in the classic textbooks that hinted that the telluride materials had the kind of crystal symmetries that were consistent with ferroelectricity, but because they had such high electrical conductivity, they couldn’t switch polarization, a requirement of true ferroelectrics.

But, he reasoned, if a good thermoelectric material had some of the characteristics of a ferroelectric, wasn’t that a good enough path to follow to explore this new, unknown territory? Randall thought it was.

The figure of merit that indicates the strength of thermoelectricity in a material is known as zT. It consists of three elements: a high Seebeck coefficient (a thermal gradient across a sample creates a voltage), high electrical transport, and very low thermal transport. One of the problems with finding materials with high zT is that good conductors of electricity are usually good conductors of heat. If you have ever burned your hand on a frying pan, you know that metals are good conductors of heat, and metals also have lots of free electrons, making them good electrical conductors. Ferroelectrics, which are insulators, are poor conductors of electrons, but also have low thermal transport. This is because the packets of heat energy called phonons scatter off the internal polar nanostructures of ferroelectrics, which slows the spread of heat from the hot side to the colder side of a material; developing a phonon glass, within the crystal structure.

 “We started asking very fundamental questions about the nature of ferroelectricity in highly conductive materials. If there were such a thing as a metallic ferroelectric, could it have the high electrical conductivity and low thermal conductivity required for high zT? And how were the conductivity and the Seebeck coefficient being influenced by the phase transition behavior of a number of ferroelectric type materials,” Randall recalled.

Phase transitions are the changes in the properties or structure of a material that take place as heat energy is added or removed from the system. Randall and his team found that in many of these ferroelectric crystals at around the phase transition point a transition in the freedom of electron movement occurred from semiconductor to metallic, from low conductivity to high conductivity.

“We’ve seen this transition in a number of ferroelectric materials at over 1000 Kelvin (1340 °F). As we asked these basic questions about the interrelationship between thermal electricity and ferroelectricity, we turned up some very promising materials, such as the tungsten bronzes,” Randall said. In September 2014, his initial results were published in a journal article in Physics Review B with coauthors Jonathan Bock, Susan Trolier-McKinstry, and Gerald Mahan.

Applications: Cooling circuits, cooling buildings

Crystal grain size in a ferroelectric material can have a profound effect on thermal transport. In ferroelectrics, a region of material called a domain has a uniform magnetic polarization – the north/south poles point in the same direction rather than randomly. By switching the polarity of domains with an electric field, not only is the polarization controlled, but thermal transport can be manipulated as well. This could lead to the ability to turn thermal transport on and off like a pump. This could be useful for applications such as heating or cooling circuits in power electronics where the circuits operate most efficiently at a certain temperature. If the circuit gets too hot, thermal conduction is turned on with an electric field. If it is cooling down too much, thermal conduction could be turned off.     

There are numerous opportunities for utilizing thermal control, from the very small in microelectronics to the very large in building materials that buffer large thermal fluctuations and keep interior spaces at a uniform temperature or lower the cost of heating and cooling.  

“We hope this approach using ferroelectrics will spur a whole new area of research for materials,” Randall said.

At Penn State, Randall’s group specializes in bulk single crystal and polycrystalline materials, while Susan Trolier-McKinstry’s group specializes in thin film materials for thermal control. Other collaborators include Patrick Hopkins at the University of Virginia, and former Penn State post-doc John Ifeld, now at Sandia National Labs. Trolier-McKinstry is currently the co-director of the Center for Dielectrics and Piezoelectrics, a joint industry/university cooperative research center with North Carolina State University sponsored by the National Science Foundation. UVA and Sandia are also CDP members.

Contact Dr. Randall at