Thermal transistors

1. Introduction

Many readers may be unfamiliar with the term "thermal transistor". In fact, when asked about thermal transistors, even modern generative AI systems sometimes produce illustrations of fictitious devices that route heat using heat itself, reflecting the fact that the concept remains relatively unknown outside specialized research communities. Whereas a field-effect transistor controls electrical current through a gate voltage, a thermal transistor is a device that modulates heat flow in response to external electrical, magnetic, or thermal stimuli. Rather than merely dissipating or storing heat, thermal transistors aim to actively switch, regulate, and control heat transport on demand.

Heat is generated in a wide range of systems, including electronic devices, batteries, power electronics, and engines. Conventional thermal management has primarily focused on efficient heat dissipation. However, as modern electronic and energy systems become increasingly compact and powerful, simply removing heat is no longer sufficient. Advanced thermal management requires dynamic control of heat flow in both space and time, retaining heat when necessary and rapidly releasing it when desired. Achieving such functionality demands technologies capable of actively controlling heat transport, beyond conventional passive thermal conductors and heat sinks.

This article briefly reviews the historical development of the thermal transistor concept and summarizes the operating principles of several representative thermal-transistor architectures. Particular emphasis is placed on thermal transistors closely related to our own research. Readers interested in a comprehensive historical overview are referred to recent review articles [1-3]. We then introduce our all-solid-state electrochemical thermal transistor, which electrically modulates thermal conductivity through electrochemical reactions, and discuss its operating mechanism, materials design, device structure, and future prospects.

2. History of Thermal Transistors

Fig. 1. Vanadium dioxide based thermal transistor. Heat flow from the hot source to the cold drain is enhanced by Joule heating in the vanadium dioxide active layer, thereby increasing thermal transport between the source and drain.
P. Ben-Abdallah and S.-A. Biehs, Phys. Rev. Lett. 112, 044301 (2014).

2.1 Early Theoretical Proposals

The concept of the thermal transistor emerged in the mid-2000s. In 2006, Li and co-workers [4] theoretically proposed a three-terminal thermal transistor consisting of source, drain, and gate terminals. Analogous to electronic transistors, the device was designed to control and amplify heat flow. However, its operation relied on Joule heating generated by electrical current to modulate heat transport between the source and drain (Fig. 1), differing from the thermal-transistor concepts pursued today.

In 2014, Ben-Abdallah and co-workers [5] proposed a thermal transistor utilizing the metal-insulator phase transition of VO2 near 68 °C. In this design, Joule heating induced a phase transition that switched the thermal conduction state. Although VO2 exhibits dramatic changes in electrical conductivity across the phase transition, experimentally observed changes in thermal conductivity are relatively small [6, 7], and practical device demonstrations have remained elusive. These pioneering studies established the theoretical foundation of thermal transistors. However, because they rely on heat generated by electrical current, they differ from the modern vision of thermal transistors, which seeks reversible and substantial modulation of intrinsic thermal transport properties through external control signals.

Fig. 2. Electrochemical thermal transistor. When ionic vacancies are present, the crystal lattice is incomplete, resulting in low thermal conductivity. Upon electrical biasing, the vacancies are filled through ionic insertion, restoring a more ordered crystal structure and increasing the thermal conductivity.
J. Cho, D.G. Cahill et al., Nat. Commun. 5, 4035 (2014).

2.2 Electrochemical Modulation of Thermal Conductivity

A major advance was the development of electrochemical approaches that directly alter the thermal conductivity of the active material. Rather than converting electrical energy into heat, these devices use electric signals to modify the composition or crystal structure of a material, thereby changing its thermal transport properties (Fig. 2).

In 2014, Braun and Cahill [8] demonstrated that electrochemical extraction of Li ions from LiCoO2 reduces its thermal conductivity. Using a specially designed battery structure on a transparent sapphire substrate and evaluating thermal conductivity by time-domain thermoreflectance (TDTR), they provided one of the earliest experimental demonstrations of this concept. In 2018, Cui and co-workers [9] reported large thermal-conductivity modulation in MoS2 through Li-ion intercalation. This work introduced the term electrochemical thermal transistor and showed that reversible insertion and extraction of ions can modify crystal structures and electronic states, leading to substantial thermal-conductivity changes. Subsequent studies by Yildiz, Chen, and others [10] demonstrated thermal switches based on oxidation-reduction reactions and protonation-deprotonation processes in perovskite oxide thin films such as SrCoOx. By altering crystal structures, bonding states, and defect concentrations through ion insertion and extraction, these devices achieved significant thermal-conductivity modulation. Although large switching ratios are possible, challenges remain in ionic mobility, reversibility, cycling durability, and operating temperature.

Fig. 3. Antiferroelectric thermal transistor. In devices employing an antiferroelectric active layer, polarization states cancel each other in the absence of an applied electric field, leading to low thermal conductivity. Applying an electric field aligns the polarization, modifies phonon transport, and results in higher thermal conductivity.
C. Liu, Z. Chen et al., Science 382, 1265 (2023).

2.3 Electrostatic Control of Thermal Transport

Electrostatic approaches have also been proposed to modulate thermal transport without relying on ion migration. In 2023, Hu and co-workers [11] demonstrated electrostatic modulation of thermal boundary resistance between an active material and electrode by as much as 1300%. Rather than changing the thermal conductivity of the bulk material, this approach controls heat flow through interfacial thermal resistance.

In the same year, researchers at Harbin Institute of Technology [12] reported field-induced switching of thermal conductivity in antiferroelectric PbZrO3 thin films. Applied electric fields alter atomic arrangements and polarization states, thereby modifying phonon transport (Fig. 3).

Because electrostatic methods avoid long-range ion migration, they offer the possibility of very fast switching. However, electric-field effects are typically confined to nanometer-scale regions, making it difficult to modulate the thermal conductivity of thick active layers [13]. Consequently, these approaches are particularly promising for nanoscale thermal devices and integrated circuits.

Fig. 4. Airgap thermal transistor. When an air gap exists between the hot source and the cold drain, heat flow is strongly suppressed. Heating the gate mechanically connects the source and drain through a high-thermal-conductivity material, thereby establishing a thermal conduction path and enabling heat flow.
L. Castelli, G. Wehmeyer et al., Nat. Commun. 14, 393 (2023).

2.4 Mechanical Modulation Using Air Gaps

An alternative strategy is to mechanically alter the heat-conduction pathway itself rather than changing a material's intrinsic thermal conductivity. In these devices, low-conductivity air gaps are introduced into the thermal transport path (Fig. 4).

In 2023, Wehmeyer and co-workers [14] reported a magnetic thermal transistor that switches physical contact between source and drain using temperature-induced magnetic transitions. More recently, Liu and co-workers [15] demonstrated an origami-inspired thermal transistor that introduces air gaps through temperature-driven structural deformation, thereby reducing heat flow.

These devices physically open and close thermal pathways and offer a fundamentally different route to thermal switching. Future challenges include improving switching speed, durability, and device miniaturization.

3. Development of an All-Solid-State Electrochemical Thermal Transistor

To realize thermal transistors suitable for both miniaturization and large-area integration, we have focused on electrochemical thermal transistors. While conventional electrochemical devices can operate near room temperature, they typically require liquid electrolytes, raising concerns regarding packaging, leakage, and long-term stability. To overcome these limitations, we developed an all-solid-state electrochemical thermal transistor based on a solid electrolyte (Fig. 5). The device consists of a Pt electrode/metal oxide/YSZ (yttria-stabilized zirconia)/Pt electrode stack, where the metal oxide serves as the active layer, YSZ functions as an oxide-ion-conducting solid electrolyte, and Pt acts as the electrode. Upon voltage application at elevated temperatures (~280 °C), oxide ions migrate through YSZ, inducing oxygen insertion or extraction within the active layer. These electrochemical reactions alter transition-metal valence states, local coordination environments, and crystal structures, thereby modifying phonon scattering and thermal conductivity. Using this platform, we have investigated a variety of redox-active oxide thin films, including SrCoOx [16-18], LaNiO3−δ [19], CeO2−δ [20], and PrBaCo2O5+x [21]. Heteroepitaxial films were grown on YSZ substrates, and their thermal conductivities were evaluated by TDTR. The top Pt electrode simultaneously functions as an optical transducer, allowing thermal characterization without substantial modification of the device structure. For thermal transistors, performance evaluation should focus primarily on the minimum thermal conductivity achievable in the OFF state and the magnitude and reversibility of thermal-conductivity modulation. Parameters such as switching time, voltage, and current strongly depend on device dimensions, electrolyte thickness, and ionic conductivity, making direct comparisons between different device architectures difficult.

Fig. 5. Solid-state electrochemical thermal transistor based on a solid electrolyte. (Left) Schematic cross-sectional structure of the device. The device was placed on a heater maintained at 280 °C in air, and voltage was applied to electrochemically oxidize and reduce the SrCoOx thin film. (Right) Changes in thermal conductivity at room temperature after repeated oxidation and reduction cycles. The thermal conductivity was 3.8 W m−1 K−1 in the oxidized state (ON) and 0.95 W m−1 K−1 in the reduced state (OFF), yielding an ON/OFF ratio of 4.0.

Q. Yang, M. Yoshimura, H. Ohta et al., Adv. Funct. Mater. 33, 2214939 (2023).

4. Future Prospects

For thin-film all-solid-state electrochemical thermal transistors operating in the cross-plane direction, the thermal-conductivity modulation of the active layer is the key performance factor, since the thermal conductivities of the electrodes and solid electrolyte remain essentially unchanged during operation. Therefore, achieving large and reversible changes in the thermal conductivity of the active material is essential. In addition, materials exhibiting low thermal conductivity in the OFF state are particularly advantageous for thin-film devices. Because thermal resistance increases as thermal conductivity decreases, a low-conductivity OFF state enhances the contribution of the active layer to the overall device thermal resistance. Otherwise, thermal resistance arising from electrodes and electrolytes can dominate, reducing overall modulation efficiency. From this perspective, cobalt oxides are especially promising. For example, SrCoOx undergoes substantial structural changes during oxygen insertion and extraction while exhibiting thermal conductivity below 1 W m−1 K−1 in the OFF state. This combination of a low OFF-state thermal conductivity and a large, reversible thermal-conductivity modulation makes cobalt oxides attractive active materials for high-performance thin-film thermal transistors. Future research should focus on further improving thermal-conductivity modulation, switching speed, cycling durability, operating temperature, and energy efficiency. Optimization of device architectures, including electrodes and solid electrolytes, will also be crucial. These advances are expected to transform thermal transistors from laboratory research topics into practical components for active thermal management in electronic and energy systems.

References

[1] T. Swoboda, K. Klinar, A. S. Yalamarthy, A. Kitanovski, and M. M. Rojo, Adv. Electron. Mater. 7, 2000625 (2021).

[2] J. Jia, S. Li, X. Chen, and Y. Shigesato, Adv. Funct. Mater. 34, 2406667 (2024).

[3] C. Liu, C. Wu, Y. Zhao, Z. Chen, T-L. Ren, Y. Chen, and G. Zhang, Phys. Rep. 1058, 1−32 (2024).

[4] B. Li and L. Wang, Appl. Phys. Lett. 88, 143501 (2006).

[5] P. Ben-Abdallah and S.-A. Biehs, Phys. Rev. Lett. 112, 044301 (2014).

[6] C. N. Berglund and H. J. Guggenheim, Phys. Rev. B 185, 1022 (1969).

[7] S. Lee, K. Hippalgaonkar, F. Yang, J. Hong, C. Ko, J. Suh, K. Liu, K. Wang, J. J. Urban, X. Zhang, C. Dames, S. A. Hartnoll, O. Delaire, and J Wu, Science 355, 371 (2017).

[8] J. Cho, M.D. Losego, H.G. Zhang, H. Kim, J. Zuo, I. Petrov, D.G. Cahill, and P.V. Braun, Nat. Commun. 5, 4035 (2014).

[9] A. Sood, F. Xiong, S. Chen, H. Wang, D. Selli, J. Zhang, C.J. McClellan, J. Sun, D. Donadio, Y. Cui, E. Pop, and K.E. Goodson, Nat. Commun. 9, 4510 (2018).

[10] Q. Lu, S. Huberman, H. Zhang, Q. Song, J. Wang, G. Vardar, A. Hunt, I. Waluyo, G. Chen, and B. Yildiz, Nat. Mater. 19, 655 (2020).

[11] M. Li, H. Wu, E.M. Avery, Z. Qin, D. P. Goronzy, H. D. Nguyen, T. Liu, P. S. Weiss, and Y. Hu, Science 382, 585–589 (2023).

[12] C. Liu, Y. Si, H. Zhang, C. Wu, S. Deng, Y. Dong, Y. Li, M. Zhuo, N. Fan, B. Xu, P. Lu, L. Zhang, X. Lin, X. Liu, J. Yang, Z. Luo, S. Das, L. Bellaiche, Y. Chen, and Z. Chen, Science 382, 1265 (2023).

[13] R. Toida, P. Lotfalinezhad, M. Adachi, T. Matsunaga, T. Fujita, K. Hirata, M. Matsunami, and T. Takeuchi, Mater. Trans. 67, 92-97 (2026).

[14] L. Castelli, Q. Zhu, T.J. Shimokusu, and G. Wehmeyer, Nat. Commun. 14, 393 (2023).

[15] B. Tan, J. Lyu, F. Yang, K. Xu, S. Qiao, B. Song, L. Yang, and K. Liu, Nat. Commun. 17, 3177 (2026).

[16] Q. Yang, H.J. Cho, Z. Bian, M. Yoshimura, J. Lee, H. Jeen, J. Lin, J. Wei, B. Feng, Y. Ikuhara, and H. Ohta, Adv. Funct. Mater. 33, 2214939 (2023).

[17] Z. Bian, Q. Yang, M. Yoshimura, H. J. Cho, J. Lee, H. Jeen, T. Endo, Y. Matsuo, and H. Ohta, ACS Appl. Mater. Interfaces 15, 23512–23517 (2023).

[18] M. Yoshimura, Q. Yang, Z. Bian, and H. Ohta, ACS Appl. Electron. Mater. 5, 4233 (2023).

[19] Z. Bian, M. Yoshimura, A. Jeong, H. Li, T. Endo, Y. Matsuo, Y. Magari, H. Tanaka, and H. Ohta, Adv. Sci. 11, 2401331 (2024).

[20] A. Jeong, M. Yoshimura, H. Kong, Z. Bian, J. Tam, B. Feng, Y. Ikuhara, T. Endo, Y. Matsuo, and H. Ohta, Sci. Adv. 11, eads6137 (2025).

[21] H. Chung, Y. Mun, A. Jeong, H. Kong, M. Yoshimura, D. Cho, S. Kwon, B-J. Hwang, Y. Kim, S-J. Lee, S. Park, H. Ohta, and H. Jeen, Chem. Mater. 38, 7802–7810 (2026).

Page updated

Google Sites

Report abuse