NERSC Initiative for Scientific Exploration (NISE) 2011 Awards

Phonon, thermodynamic and diffusion properties of lithium-ion batteries from first-principles calculations

ShunLi Shang, Pennsylvania State University

Associated NERSC Project: First-principles investigations of fundamental properties for Ni-base alloys (m679)

The fundamental properties of thermodynamics and diffusion are crucial for the development of Li-ion batteries, e.g., (i) the voltage of a battery relates to the chemical potentials of Li-ion in cathode and anode, (ii) the Li-ion electrode discharging-charging process is enabled by Li and electron diffusion into (out of) the active cathode particles; and (iii) they are the basic inputs for other simulations such as phase-field simulations. However, most of these fundamental properties are still lacking for Li-ion battery materials due to the daunting amount of experimental works. In the present proposal, an integrated first-principles calculations and computational thermodynamics (i.e., the CALPHAD modeling) approach will be used to predict the critically needed thermodynamic and diffusion properties of Li-ion battery materials in the Li-Mn-Fe-Co-Ni-O system, with the main focus being the high-energy layered LixNiyMn1-2yCoyO2.

In order to gain the thermodynamic and diffusion properties as a function of composition and temperature for Li-ion batteries by CALPHAD modeling, the key fundamental properties are needed, herein, from the first-principles phonon calculations due to the dearth of experimental data. The complex structures, the dipole-dipole interactions, and the partially disordered nature of the Li-ion battery materials offer a number of challenges (barriers) for the first-principles phonon calculations, but they are tractable as shown in the next paragraph.

In the present proposal, first-principles based phonon calculations will be used to predict the temperature-dependent thermodynamic and diffusion properties. Using our newly developed Yphon approach for phonon [J. Phys. Cond. Mat. 22, 2010, 202201], the dipole-dipole interactions can be treated accurately for the polarized materials (e.g., the Li-ion oxides). Using first-principles transition state theory with inputs of the migration energy and phonon properties, the diffusion coefficients for both diffusion barrier and pre-factor can be predicted as demonstrated firstly by us for fcc Al [Phys. Rev. Lett. 100, 2008, 215901]. For the partially disordered phases (e.g., LixNiyMn1-2yCoyO2), the degree of randomness needs to be considered. Here, we propose to use the special quasirandom structures (SQS) approach as well as the partition function method developed recently by us [Phys. Rev. B 82, 2010, 014425].

When the project is completed successfully, both the computational approach (first-principles phonon and CALPHAD) for the complex Li-ion battery materials and the thermodynamic and diffusion databases of the Li-Mn-Fe-Co-Ni-O system will be delivered. These databases can be used, e.g., (i) to predict the phase stability of Li-ion battery materials under different process conditions, (ii) to find optimizing route to synthesize Li-ion batteries with good performances, (iii) as input for phase-field simulations.