Abstract The conventional doping strategy for thermoelectric materials generally focuses on a shallow donor/acceptor model with the energy level close to the band edge as for electronic devices. However, thermoelectric devices operate over a large temperature difference, and the optimal carrier concentration increases with increasing temperature. A shallow level cannot meet the requirement over a large temperature range. Here, an innovative strategy of introducing an intermediate level is proposed. Such an intermediate level introduces more carriers with increasing temperature, consistent with the trend of increasing optimal doping concentration with temperature, enabling larger power factor over a broader temperature range. Furthermore, the intermediate level typically requires more impurities, leading to increased phonon scattering. This strategy allows simultaneous optimization of carrier concentration over a wide temperature range and suppression of thermal conductivity via stronger point-defect phonon scattering. Experimental results from heavily-doped ZrCoSb employing shallow, intermediate, and deep levels successfully corroborate this strategy, where simultaneously improved power factor and figure of merit are obtained by introducing an intermediate level. Our work indicates that the performance of known thermoelectric materials should be reevaluated by introducing an intermediate level to unleash their full potential.

中文翻译：

---

同时优化功率因数和声子热导率以提高热电品质因数的中间级掺杂策略

摘要 热电材料的传统掺杂策略通常侧重于能级接近带边的浅施主/受主模型，如电子器件。然而，热电器件在很大的温差下工作，最佳载流子浓度随着温度的升高而增加。浅水平不能满足大温度范围的要求。在此，提出了引入中级的创新策略。这种中间水平会随着温度的升高引入更多的载流子，这与最佳掺杂浓度随温度升高的趋势一致，从而在更宽的温度范围内实现更大的功率因数。此外，中间水平通常需要更多的杂质，导致声子散射增加。该策略允许在宽温度范围内同时优化载流子浓度并通过更强的点缺陷声子散射抑制热导率。采用浅、中和深能级的重掺杂 ZrCoSb 的实验结果成功地证实了这一策略，其中通过引入中间能级同时提高了功率因数和品质因数。我们的工作表明，应该通过引入中间水平来重新评估已知热电材料的性能，以释放其全部潜力。深层次成功地证实了这一策略，通过引入一个中间层次，同时提高了功率因数和品质因数。我们的工作表明，应该通过引入中间水平来重新评估已知热电材料的性能，以释放其全部潜力。深层次成功地证实了这一策略，通过引入一个中间层次，同时提高了功率因数和品质因数。我们的工作表明，应该通过引入中间水平来重新评估已知热电材料的性能，以释放其全部潜力。