Rapid in-silico Battery Electrolyte Electrochemical Reaction Generation using 3T-VASP Multi-Scale Energy Minimization

This research uses 3T-VASP to start a lithium ion battery electrolyte reaction pathways, analyzing byproducts and SEI to tackle experimental challenges and set up workflows for new systems. Existing experimental and computational methods struggle to study process electrolyte reaction pathways due to the many types and hard to analyze complex SEI. Ab-initio methods, like DFT, are slow for finding real electrochemical byproducts, needing many steps.

This paper fills the gap with a fast, multi-scale approach that use far fewer DFT steps. To begin, create a new Conda environment named name 3T with Python version 3.11 by entering the appropriate command in the terminal. Activate the 3T environment using the activate command.

Then use conda to install git from the conda-forge channel. Clone the 3T-VASP github repository using the URL. Then enter the cloned repository directory named External_3T.

Use Conda to install mamba from the conda-forge channel. Then use mamba to install the required libraries for the 3T conda environment. Next, install the software GROMACS and InterMol within the 3T environment.

After installing the modified version of InterMol in the 3T environment, open the Python script file named calculator_3T_VASP. py located in the utils directory using a code editor. Scroll to the run_VASP function and locate the default OS system call used to execute the VASP software.

Then modify the command line to reflect the path of the user's VASP executable and specify the computing resources as needed. In the Linux terminal, verify that non-Python third-party utilities required by 3T are available by executing the commands for gmx, wget, unzip, packmol. To test if 3T-VASP has been configured correctly, use Python to run randomize_3T_bulk_electrolyte_reduction.

py from the terminal. Open another terminal to monitor the output file named default. log as the simulation runs.

Name the 3T-VASP lattice file using the vasp convention and save each file into an appropriate subfolder inside the input directory. Create override files for each periodic box lattice structure with the file extension.override. Place them in the same subfolder under the input directory as the corresponding VASP files.

Then save these files as JSON dictionaries with the required keys movable_group and atom_charge_proximity. To generate molecule structure files in the XYZ format, name them using the xyz pattern and place these files directly inside the input folder. Prepare the template VASP input files required for each 3T-VASP step, including INCAR, KPOINTS, and POTCAR files.

In the INCAR file, ensure the molecular dynamics step count parameter NSW is either omitted or explicitly set to zero. Next, prepare a 3T configuration file that references and integrates all the created input files. To perform a single 3T-FF or 3T-VASP trajectory generation, launch Python in the terminal.

Import the main function from the main run utils module. Execute the trajectory generation using a specific configuration file. Monitor the progress by viewing the default.

log file in a separate terminal. To perform large scale trajectory generation, write a short function to replace specific phrases in the config template file, generating new config files to produce different 3T-VASP trajectories. A gradual dispersion of electrolyte molecules was observed inside the periodic boundary condition box during the 3T-FF phase without any chemical reactions occurring, confirming a stable force field setup as seen across the three snapshots.

During the 3T-VASP phase, electrochemical reduction of ethylene carbonate into ethylene and carbonate anion was commonly observed. While a secondary reaction producing ethane-1, 2-diolate anion and carbon monoxide occurred less frequently.