Most thermoelectric devices in use and under development rely on the longitudinal thermoelectric effect, where electricity is generated along the same direction as the heat flow across the device. These systems typically employ alternating layers of p-type and n-type semiconductors connected in series, with each type generating voltage in opposite directions when a temperature difference drives charge carriers from a hot side to a cold side. Although this architecture is well established, stacking multiple layers introduces many interfaces, and the associated electrical contact resistance at these junctions leads to energy losses that limit overall conversion efficiency.
Transverse thermoelectric devices offer an alternative approach by generating voltage perpendicular to the direction of heat flow, opening the door to simpler and potentially more efficient designs. Because transverse thermoelectric devices can be realized using a single material rather than a series of p-type and n-type elements, they avoid most interface-related losses, reduce contact resistance and simplify manufacturing. However, materials that exhibit a strong transverse thermoelectric effect are scarce, and identifying suitable candidates has been a major challenge.
A research team led by Associate Professor Ryuji Okazaki of the Department of Physics and Astronomy at Tokyo University of Science has now demonstrated promising transverse thermoelectric behavior in the mixed-dimensional semimetal molybdenum disilicide, or MoSi2. The team also included Ms. Hikari Manako, Mr. Shoya Ohsumi and Assistant Professor Shogo Yoshida from Tokyo University of Science, together with Assistant Professor Yoshiki J. Sato from Saitama University in Japan. Their results appear in the journal Communications Materials in a paper published on December 29, 2025.
"We wanted to explore new transverse thermoelectric materials. Recently, the presence of axis-dependent conduction polarity (ADCP) in a material has been recognized as an indicator for TTE generation ability," explains Okazaki. "Mixed-metal conductors like MoSi2 are potential ADCP candidates, but their thermopower generation ability has not been thoroughly investigated."
To assess the potential of MoSi2, the researchers combined experimental measurements with first-principles calculations to study its transport properties. They examined how its resistivity and thermal conductivity change with temperature, and they measured longitudinal thermopower along two different crystallographic axes of the crystal. These thermopower measurements revealed clear axis-dependent conduction polarity, and Hall resistivity measurements confirmed the presence of this behavior in the material.
The team then turned to electronic structure calculations to understand the origin of ADCP in MoSi2 at a microscopic level. They showed that the effect arises from a mixed-dimensional Fermi surface made up of two distinct Fermi surfaces with opposite polarities. The Fermi surface marks the boundary between filled and empty electronic states in a solid and plays a central role in dictating how electrons move and how charges and heat are transported through a material.
To directly probe the transverse thermoelectric response, the researchers applied a temperature gradient at a 45-degree angle to one of the crystallographic axes of MoSi2 and measured the resulting transverse thermopower. The experiments showed a clear and sizable transverse thermopower signal, demonstrating that MoSi2 can generate voltage perpendicular to the heat flow under appropriate conditions. The magnitude of this transverse signal exceeded that found in tungsten disilicide, or WSi2, another axis-dependent conduction polarity material previously investigated by the same group, largely due to differences in the way electrons are distributed in MoSi2.
The transverse thermopower observed in MoSi2 was also comparable to values reported for anomalous Nernst materials, a class of magnetic materials recognized for exhibiting strong transverse thermoelectric effects. This comparison suggests that MoSi2, despite being a mixed-semimetal system rather than a magnetic material, can deliver competitive performance for transverse energy conversion. As a result, the study positions MoSi2 as a strong candidate for transverse thermoelectric applications, particularly in low-temperature regimes where its properties are most favorable.
"These findings establish MoSi2 as an ideal material for TTE applications, particularly in the low-temperature range, thereby expanding the list of viable candidates," says Okazaki. "Moreover, both MoSi2 and WSi2 show that mixed-dimensional Fermi surfaces are important for the emergence of ADCP and therefore transverse thermopower."
Looking ahead, the team notes that using thin films of MoSi2 as the active layer in transverse thermoelectric devices could allow large-area heat sources to be covered and efficiently converted into electrical voltage. This capability would be valuable for industrial waste heat recovery, power generation on heat-emitting components and other low-grade heat sources that are currently underutilized. The work also demonstrates how focusing on mixed-dimensional Fermi surface structures and axis-dependent conduction polarity can guide the search for new transverse thermoelectric materials.
By highlighting mixed-semimetal systems such as MoSi2 and clarifying the role of electronic structure in their thermoelectric response, the study offers a new strategy for discovering and optimizing materials that can turn heat into electricity. Such advances may feed into the development of compact and efficient waste heat recovery systems that support a more sustainable energy landscape.
Research Report:Axis-Dependent Conduction Polarity and Transverse Thermoelectric Conversion in the Mixed-dimensional Semimetal MoSi2
Related Links
Tokyo University of Science
Powering The World in the 21st Century at Energy-Daily.com
| Subscribe Free To Our Daily Newsletters |
| Subscribe Free To Our Daily Newsletters |
