Thermoelectric energy conversion is one of the key technologies for energy harvesting devices to convert waste heat into electric power, and vice versa. In these devices, the Seebeck coefficient (= thermopower, S), which is the proportional constant to the voltage generation against an induced temperature gradient, is a significant factor in designing thermoelectric devices. Importantly, according to the Mott equation, Sis proportional to the energy derivative of the electronic density of states at around Fermi energy; therefore, it is critically important to control the energy band filling. Conventionally, the band filling is tuned by chemical doping and, in most materials, it is very difficult to establish precisely controllable doping methods.

Another important factor for thermoelectric energy conversion devices is the thermoelectric power factor S²·σ (σ is conductivity). It should be strongly emphasized that the power factor has to be optimized to maximize the electric power output of thermoelectric devices. Consequently, this is a key parameter for applications. However, it is widely known that there is a trade-off between |S| and σ in terms of carrier density, n. Although σ is almost linearly proportional to n, |S| decreases with increasing n. Therefore, it is necessary to maximize S²·σ by tuning n.

Here, we optimized thermoelectric energy conversion efficiency using the electrolyte gating technique. Importantly, owing to the high specific capacitance of electric double-layers (EDLs) (1−10 μF cm^-2), we can continuously control the carrier density up to 10¹⁴ cm^-2, which results in the tuning of the carrier polarity between p-type and n-type [1-6]. Based on this technique, we investigated the thermoelectric properties of transition metal dichalcogenide monolayers [7], single-walled carbon nanotubes [8,9], organic Mott insulators [10], organic molecules and organic polymers.

[1] J. Pu, L.-J. Li, T. Takenobu, et al., Nano Lett. 12, 4013 (2012).

[2] J.-K. Huang, T. Takenobu, L.-J. Li, et al., ACS Nano. 8, 923 (2014).

[3] J. Pu, L.-J. Li, T. Takenobu, et al., Adv. Mater. 12, 4013 (2016).

[4] H. Shimotani, T. Takenobu, et al., Adv. Funct. Mater. 24, 3305 (2014).

[5] Y. Kawasugi, T. Takenobu, et al., Nat. Commun. 7, 12356 (2016).

[6] Y. Yomogida, T. Takenobu, et al., Adv. Mater. 24, 4392 (2012).

[7] J. Pu, L.-J. Li, T. Takenobu, et al., Phys. Rev. B 94, 014312 (2016).

[8] K. Yanagi, T. Takenobu, et al., Nano Lett. 14, 6437-6442 (2014)

[9] S. Shimizu, T. Takenobu, et al., SMALL 12, 3388 (2016)

[10] Y. Kawasugi, T. Takenobu, et al., Appl. Phys. Lett. 109, 233301 (2016).