Understanding what influences Li-ion diffusion barriers.

/home/fri/projects_2021/lionDiffusions Studying kinetics is expensive as it takes both a lot of time and attention and a lot of computing power.
This project aims at automating the process as much as possible, while also looking for ways to bypass computation-intensive studies.
Here are a few ways to get involved in this project:

1. Peoplel: Rupin
   Process existing calculations, identify calculations with errors, complications, and label jobs that are finished/unfinished.

   Around 100 bands were attempted in the past. They can be found in : /home/fri/projects/lionbands2021
   These bands are modeling a few different types of cathode materials for Li-ion batteries.
   These bands are valuable data to train both humans and algorithms. It is, however, un-labeled at the moment.
   To correctly label all bands, you need to manually look at each band as well as the initial and final images used to determine:

   1. If the band has one of the following types of error: Inconsistent ENCUT, ISPIN, in the INCARs between ini, fin and band; Inconsistent POTCARs between ini, fin and band. Unoptimized initial and final image (force significantly greater than the EDIFFG setting in INCAR). If any is present, document the path to the job and the error type.
   2. If the band need or have had special treatments that is present in the existing MEP gallery
   3. If the band need or have had special treatments that is not already documented in the gallery above.

   
   
2. Continue the work of the last cohort and increase the degree of automation for running NEB calculations.

   1. People: Tristan Wang
      A very nice report written by Jackson Morris(2021) documented very nicely a few shell scripts he put together. The report can be found here: /home/fri/projects/lion/Fall_2021_Lithium_Diffusion_Report.docx
      The script was based on foundations laid by Graham Stockton(~2020), Andrew Chen(2021) and Sydney Suthar(2021).
      It can be found at /export/home/jjm5676/process_testing/BandRunner.py and BR_helper.sh
      This script does a few automated checks on common mistakes, convergence of the band, and if the band needs one of the known special treatments in the gallery.
      This script (BR_helper) is still a bash script. To make it easier to maintain, we will rewrite it with python.
   2. People: Hanjing
      BandRunner must be ran where the band is located. We aim to remove this inconvenience through automation.
      I have written a script /home/fri/projects/automagician.py that can remember a job and its location once a job is submitted through it. This script will automatically check and rerun all known jobs wherever the user runs it.
      Automagician only works on relaxation jobs for now. It needs abilities to recognize the need and the existence of NEB bands, then Automagician needs to absorb BandRunner's ability to check and run NEB jobs.

   
   
3. People: Dennis
   Find correlations between diffusion barriers and characteristics of critical points.

   Find all converged bands, then perform BCTA on its ini structure. Tabulate and report the density, Laplacian, eigenvectors and their corresponding eigenvalues for all critical points along the diffusion pathway. A tutorial on how to do Bader Charge Topology Analysis (BCTA) can be found here.