Apr 9, 2025
5:00pm - 7:00pm
Summit, Level 2, Flex Hall C
Hrushikesh Sahasrabuddhe1,2,Jiongzhi Zheng3,Geoffroy Hautier3,Anubhav Jain2
University of California, Berkeley1,Lawrence Berkeley National Laboratory2,Dartmouth College3
Hrushikesh Sahasrabuddhe1,2,Jiongzhi Zheng3,Geoffroy Hautier3,Anubhav Jain2
University of California, Berkeley1,Lawrence Berkeley National Laboratory2,Dartmouth College3
In this work, we describe our effort to build a harmonic phonon database (that ultimately also includes anharmonic phonons) that will be hosted on the Materials Project (MP). We use
pheasy – a compressed sensing lattice dynamics code that enables an efficient way to calculate the interatomic force constants by using the regularized linear regression technique of LASSO to fit the force constants to the sensing matrix. Our workflow can achieve a significant speedup as compared to the finite displacement technique when calculating 2
nd, 3
rd & 4
th-order FCs. The database, which is currently under construction, is composed of compounds of all the crystal symmetries ranging from simple cubic systems to the least symmetric triclinic systems. It stores the phonon density of states (DOS), phonon dispersion, Helmholtz free energy (F), entropy (S), and heat capacity at constant volume (C
v) calculated for compounds with less than or equal to 20 total sites in the primitive cell. This talk will present the progress thus far, overall high-level statistics of the database, and unit-level results of some interesting material systems.
In this work's second part, we study the anharmonic lattice dynamics behavior of lead-free double halide perovskites. Such materials are candidate thermoelectric materials due to their ultra-low experimentally observed lattice thermal conductivities (K
lat), high carrier mobilities, and large Seebeck coefficient. We will discuss our progress in calculating K
lat, elimination of imaginary phonon modes at higher temperatures (via the Stochastic self-consistent harmonic approximation (SSCHA) and statistical perturbation-operator renormalization (SPOR) techniques), and polymorphic phase transition temperature ranges for different chemistries within the same family of double halide perovskites.
This work is enabled by integrating
pheasy with the workflow orchestration software –
Atomate2, which ensures compatibility with the standard technology stack of the MP. In the future, we plan to develop a high-quality ab initio level anharmonic phonon database to assist in generating finite temperature phase diagrams and compiling K
lat. This will be highly beneficial for discovering new thermoelectric materials and other applications in which thermal properties must be carefully considered.