Development of thermoelectric (TE) materials is important, for energy saving via waste heat power generation , and IoT power sources . For high TE performance, we must find ways to overcome the traditional tradeoffs between the properties, namely, between Seebeck coefficient S and electrical conductivity s, and between the electrical and thermal conductivity k .
For the latter aspect, in addition to various nanostructurings, intrinsic low k mechanisms have been demonstrated. Materials informatics approach , doping leading to lattice softening , heterogeneous bonding from mixed anions , etc.
For overcoming the first tradeoff, we have found that magnetism can be utilized to enhance the Seebeck coefficient and overall power factor (PF). Coupling of the electrical carriers with magnetic moments can lead to magnon drag which was known from long ago in Fe, etc. to lead to the increase in S at low temperatures. However, the recent advancement is the demonstration that magnon drag can actually lead to high performance, i.e. high power factors at higher temperatures, namely for CuFeS2 chalcopyrite . Recently, magnon drag was also proposed as the origin  of the huge power factor in metastable Fe2VAl-based thin films .
We have also discovered TE enhancement in paramagnetic systems, namely, we show that in cases with a strong coupling of the electrical carriers with the magnetic moments, this interaction “drags” the carriers, leading to an increase in the effective mass which enhances the Seebeck coefficient. This will be detrimental to the mobility but overall, enhancements to the power factor have been able to be realised in high-performance TE systems, the first example demonstrated being CuGaTe2 . Later this interaction was named paramagnon drag. Magnetic ion doping enhancement has also been demonstrated for Bi2Te3 , for example.
Spin fluctuation was found to enhance the Seebeck coefficient in the Heusler alloy Fe2VAl . Spin entropy is also known to enhance S .
I will discuss all these magnetic TE enhancement phenomena.
In a final topic, an interesting dual effect of small amounts of Cu doping in Mg3Sb2 was revealed. Interstitial Cu doping lowered the phonon group velocity, while doping into the grain boundaries promoted grain growth and optimum interfaces leading to very high mobilities similar to single crystals, while being a polycrystalline material with low thermal conductivity. An initial realistic 8 pair module exhibited an efficiency of 7.3%@320°C, with estimated efficiency from the actual materials being ~11%! . Tuning toward room temperature yielded an initial realistic 8 pair module with an efficiency of 2.8% with temperature difference of 95 K from RT and cooling of 56.5 K . Recently, a modified single element device of Mg3Sb2 was able to achieve a TE efficiency ~12% .
 L. E. Bell, Science 321, 1457 (2008), JOM, 68, 2673-2679 (2016).
 Sci. Tech. Adv. Mater. 19, 836 (2018), MRS Bulletin, 43, 176 (2018).
 T. Mori, Small 13, 1702013 (2017), Energies, 15, 7307 (2022).
 Energy Environ. Sci., 14, 3579 (2021).
 Adv. Energy Mater., 11, 2101122 (2021).
 J. Mater. Chem. A, 9, 22660 (2021), J. Mater. Chem. A, 11, 10213 (2023) Hot article.
 Angew. Chem. Int. Ed. 54, 12909 (2015).
 Phys. Rev. B, 104, 214421 (2021).
 Nature 576 (7785) 85-90 (2019).
 J. Mater. Chem. A, 5, 7545 (2017).
 Mater. Today Phys., 9, 100090 (2019).
 Science Advances, 5, eaat5935 (2019).
 Sci. Tech. Adv. Mater., 22, 583-596 (2021).
 Joule, 5, 1196-1208 (2021).
 Nature Commun., 13, 1120 (2022).
 Advanced Energy Materials, doi: 10.1002/aenm.20230166 Selected as Front Cover Article.