Skip to content Skip to main navigation Report an accessibility issue

Dr. Veerle Keppens




keppens_3As magical as that sounds, thermoelectric power is not a silver-bullet solution to recovering energy from the waste-heat produced by combustion engines or industrial machinery. Not only do waste-heat scenarios vary, thermoelectricity itself is not fully defined.

“Efficient thermoelectric materials require the unusual combination of poor thermal conductivity and good electrical conductivity,” says JDRD team leader Veerle Keppens.

“Electrons alone carry electricity; both electrons and the vibrations of atoms in a crystal’s lattice structure—known as phonons—carry heat. A good thermoelectric material is efficient at scattering phonons, which impedes the transfer of heat from one side of a material to the other while allowing the electrons to pass right on through,” Keppens says.

But what, exactly, hinders heat transfer is still unclear.

Keppens, PhD student Lindsay VanBebber, and ORNL LDRD team leader Olivier Delaire are looking into the influence ferroelectric properties have on atoms moving within a crystal’s lattice structure.

Ferroelectric materials have the ability to spontaneously polarize—and change their conductivity—in the presence of an electric field at specific temperatures.

keppens_2“As a whole, these materials are neutral; they have no actual charge,” Keppens says. “But,the way the tiny units inside the material are distributed makes it a little more negative on one end and a little more positive on the other”—in other words, polarized.

But, Keppens explains, this can only happen if the structure lacks symmetry.

“In a perfect crystal at absolute zero degrees, the atoms occupy well-defined positions. Raise the temperature and the atoms start to move, some differently from others. Because of that you can create peculiar lattice dynamics, depending on the structure and the atoms you put into a material.”

keppens_4In year-one the JDRD team synthesized single crystals containing a combination of lead, tin, telluride, molybdenum, and antimony and studied their physical responses with Resonant Ultrasound Spectroscopy (RUS). RUS shows when the structural change actually occurs in the material—information that adds insight into the mechanisms linking lattice dynamics with suppressed thermal conductivity.

In year two Keppens’ team will increase the temperature range of the experiments. Delaire’s team will collect neutron scattering data from crystals grown in Keppens’ lab. Together the ultrasound and neutron scattering results will help them build realistic computer simulations of the unusual lattice dynamics found in thermoelectric materials.


JDRD project:
Ferroelectric instabilities in thermoelectric materials (year 2)
Veerle Keppens, UT Materials Science and Engineering Department

LDRD project:
Improving energy efficiency in thermoelectric materials by integrating neutron scattering with supercomputing and modeling
Oliver Delaire, ORNL Materials Science and Technology Division


Desktop DeviceAs promising as this sounds, Veerle Keppens says efficient thermoelectric materials require the unusual combination of poor thermal conductivity and good electrical conductivity.

“Typically, materials with good electrical conductivity also have good thermal conductivity,” she says.

Electrons alone carry electricity; both electrons and phonons (the atomic vibrations from atoms in a crystal’s lattice structure) carry heat. So, a good thermoelectric material is efficient at scattering phonons, which impedes the transfer of heat from one side of a material to the other while allowing the electrons to pass right on through.

Working in tandem, Keppens’ JDRD team, with PhD student Lindsay VanBebber and an LDRD team led by Olivier Delaire of ORNL’s Materials Science and Technology Division, are examining the influence ferroelectric properties have on lattice dynamics—in other words, how these properties affect the atoms moving within the crystal. Their goal is to gain a better understanding of the microscopic origins of suppressed thermal conductivity.

The two teams will explore how a material’s ferroelectric ability to spontaneously polarize under specific conditions correlates with thermoelectric performance.

Temperature Chart“Ferroelectric material as a whole is neutral; there’s no actual charge,” Keppens says. “But the way the tiny units inside are distributed makes it a little more negative on one end and a little more positive on the other. This can only happen if there’s a lack of a certain symmetry in the structure.

“In a perfect crystal at absolute zero degrees, the atoms occupy well-defined positions. Raise the temperature and the atoms start to move, some differently from others. Because of that you can create peculiar lattice dynamics, depending on the structure and the atoms you put into a material.”

For their study, the JDRD team is growing single crystals of Pb1-xSnxTe (lead, tin, telluride)—a new combination, cousin to lead-telluride materials, foremost in thermoelectric power generation for applications above room temperature. They will use Resonant Ultrasound Spectroscopy (RUS) to measure the crystal’s physical response to ultrasonic signals at varying temperatures. Keppens says RUS should show when structural changes actually occur in the material—information that will add insight into the mechanisms linking lattice dynamics with suppressed thermal conductivity.


JDRD project:
Ferroelectric instabilities in thermoelectric materials
Veerle Keppens, UT Department of Materials Science and Engineering

LDRD project:
Improving energy efficiency in thermoelectric materials by integrating neutron scattering with supercomputing and modeling
Oliver Delaire, ORNL Materials Science and Technology Division