[Paper Review] LanTraP: A code for calculating thermoelectric transport properties with the Landauer formalism

Xufeng Wang, Evan Witkoske+2 more|arXiv (Cornell University)|Jun 23, 2018

Advanced Thermoelectric Materials and Devices32 references3 citations

TL;DR

LanTraP is a computational code that calculates thermoelectric and electronic transport properties using the Landauer formalism, enabling efficient band-counting algorithms for rapid screening of DFT band structures. It achieves good agreement with Fourier-based interpolation methods, offering a faster alternative for high-throughput materials screening in thermoelectrics.

ABSTRACT

A code for calculating the semi-classical thermoelectric and electronic transport properties is described. It uses the Landauer transport theory, which is equivalent to the Boltzmann theory, by introducing a central quantity-the distribution of modes. Its usage enables the so-called band-counting algorithm that can speed up the calculation and offers the potential to rapidly screen DFT band structures. Good agreements are found when comparing the results obtained using band-counting and established Fourier-based interpolation methods.

Motivation & Objective

• To develop a computationally efficient method for calculating thermoelectric transport properties in materials.
• To enable high-throughput screening of DFT band structures by reducing computational cost.
• To implement the Landauer formalism using a mode distribution approach for semi-classical transport calculations.
• To validate the band-counting algorithm against established Fourier-based interpolation techniques.
• To provide an open-source tool for researchers in thermoelectrics and materials science.

Proposed method

• The code employs the Landauer formalism, which is mathematically equivalent to the Boltzmann transport equation, using a mode distribution as the central quantity.
• It introduces a band-counting algorithm that aggregates transport contributions across energy levels without full band structure interpolation.
• The method avoids expensive Fourier interpolation by directly computing transport coefficients from discrete energy levels and group velocities.
• It uses semi-classical transport theory to compute electrical conductivity, Seebeck coefficient, and power factor.
• The algorithm is designed to be compatible with standard DFT band structure outputs, enabling rapid post-processing.
• The implementation is optimized for performance, allowing fast evaluation of transport properties across multiple doping levels and temperatures.

Experimental results

Research questions

• RQ1 Can the Landauer formalism with band-counting achieve accurate thermoelectric transport predictions comparable to Fourier-based interpolation?
• RQ2 How does the computational efficiency of the band-counting method compare to traditional interpolation techniques?
• RQ3 To what extent can the band-counting algorithm accelerate high-throughput screening of thermoelectric materials from DFT data?
• RQ4 What is the accuracy of LanTraP in predicting key thermoelectric properties like the power factor and Seebeck coefficient?
• RQ5 How robust is the method across different materials with varying band structures and carrier concentrations?

Key findings

• The LanTraP code achieves good agreement with established Fourier-based interpolation methods in predicting thermoelectric transport properties.
• The band-counting algorithm significantly reduces computational cost, enabling faster screening of DFT band structures.
• The method maintains accuracy across a range of materials and doping levels, as validated against reference calculations.
• The Landauer formalism implementation provides a reliable semi-classical framework for transport property prediction.
• The code is efficient and scalable, making it suitable for high-throughput materials discovery in thermoelectrics.
• The results demonstrate that mode-based aggregation via band-counting is a viable and accurate alternative to full band interpolation.