Collection |

Thermoelectric materials by design

Thermoelectric materials are attractive options for recovering energy that would otherwise be lost, based on the principle of an electrical potential being generated when there is a temperature difference across a piece of material. Enhancing the thermoelectric performance is a significant challenge however because the determining factors typically correlate with each other and are therefore difficult to tune independently. Deep understanding of electrical and thermal transport properties does make it possible, however, to design optimized material structures across multiple length scales. In particular, machine learning is a powerful tool for optimizing compositions, and the combination of calculations and experiments has accelerated the development of high performance thermoelectric materials. This collection brings together recent works published in npj Computational Materials that contribute towards the design of high performance thermoelectric materials.

Articles

High accuracy predictions of materials properties can be obtained using Bayesian optimization (BO). A team led by Priya Vashishta at University of Southern California developed a Gaussian regression model capable of predicting the band gap value and thermoelectric properties of three-layered van der Waals heterostructures of transition metal dichalcogenides. A BO model further allowed identification of optimal heterostructures using a minimal number of ab initio calculations. BO models were computed to find either heterostructures with maximum band gap or heterostructures with a band gap value closest to 1.1 eV, relevant for optoelectronic and thermoelectric applications. BO was found to identify nearly optimal materials configurations with high probability, whilst significantly reducing the computational cost of discovering ideal structures using regression models.

Article | Open Access | | npj Computational Materials

Experimental carrier concentration can serve as the basis for a model to understand and predict high performance thermoelectrics. Carrier concentration is instrumental in controlling properties. Despite significant experimental progress, establishing guidelines towards the desired performance through doping remains challenging. Now, a team from Northwestern University, Colorado School of Mines, and National Renewable Energy Laboratory in USA have predicted the dopability ranges of several diamond-like semiconductors, based on data from experimentally reported doping limits for 127 compounds. Several materials that combine simultaneously promising thermoelectric quality factor and complementary dopability are singled out. Apart from shedding light on what drives dopability in this family, the model also suggests that a number of less-studied compounds deserve more attention.

Article | Open Access | | npj Computational Materials

A phonon scattering process on the surface of phononic crystals can explain their ultra-low thermal conductivity. Two-dimensional silicon phononic crystals are promising for thermoelectric applications, as the periodic arrangement of holes allows for significant reduction of their thermal conductivity. A team from Hunan University of Science and Technology and Xiangtan Universities in China, and the Institute of High Performance Computing in Singapore, manage to model the values reported experimentally by incorporating a phonon scattering mechanism, rooted to the bond imperfection on the surface of the nanostructure. As the bonds towards the surface grow shorter and therefore stronger, they perturb the local potential, and suppress the high-frequency phonons. Low-frequency phonons, on the other hand are suppressed by normal boundary scattering. The authors conclude that these two actions lead to the ultra-low thermal conductivity values, and predict that further reduction can be achieved by roughening the hole walls in the phononic crystal.

Article | Open Access | | npj Computational Materials

Accurate modeling of defects in PbTe lies on the combination of hybrid functionals with spin-orbit coupling and Green function theory. Dopability of thermoelectric materials (that can convert thermal into electrical energy and vice versa) is crucial for their improvement; however most modeling approaches fail on an atomic level, especially if spin-orbit coupling effects (that modify the band position) are present. In this work, Vladan Stevanovic and coauthors manage to accurately reproduce the electronic structure of PbTe with first-principles based density functional theory. They prove that the only approach that can model the intrinsic defect chemistry, in agreement with experimental results, is the combination of hybrid functionals (based on a screened Coulomb potential) with spin-orbit coupling and the band edge shifts calculated through the G 0 W 0 approximation (which calculates the Green function theory that models excited-state properties of extended systems). Such level of understanding is necessary to predict the dopability of PbTe.

Article | Open Access | | npj Computational Materials

Researchers have established how missing atoms and ‘nano-grains’ inside thermoelectric oxides can optimize harvesting of waste heat. Materials that transform temperature gradients into electricity are increasingly being made from metal oxides instead of alloys due to issues of cost and stability. Theoretical simulations by Yuan-Hua Lin from Tsinghua University and co-workers show that reducing the flow of heat through oxide frameworks—a key factor in improving energy conversion—is possible with a phonon scattering strategy. Atoms misplaced from normal lattice positions, known as point defects, deflect heat-carrying phonon waves in both alloys and oxides. But through mathematical modelling and comparison with experimental data, Lin’s team predicts that oxides can access additional scattering routes with appropriate levels of oxygen vacancies and nanoscale crystal grain sizes, leading to enhanced thermoelectric efficiency.

Article | Open Access | | npj Computational Materials

A simple method for determining a material’s thermoelectric properties is developed by researchers in the United States and Belgium. Jeffrey Snyder from Northwestern University and his co-workers’ model could simplify the search for materials that efficiently generate electricity from waste heat. Even though the environment of an electron in a solid is very complex, the way an electron moves through a solid’s lattice of atoms can be treated as if it is moving in free space. However, because of the influence of its environment an effective mass, not its true mass, is used to model the movement of electrons and that material’s properties. But this effective-mass can be defined in several ways depending on which material property is being modeled. Snyder et al. determine that the ratio of two different effective masses, as computed from different electronic properties, could be a good method to identify novel thermoelectric materials and can be associated with the “complexity” of the electronic structure.

Article | Open Access | | npj Computational Materials

Reviews