NCSI: DFT Simulation of Novel Lithium Rich Double Anti-Perovskites for Solid Electrolyte in Lithium Ion Battery

Status	Completed
Mentor Name	Kevin Brandt
Mentor's XSEDE Affiliation	Campus Champion and past Regional Campus Champion
Mentor Has Been in XSEDE Community	4-5 years
Project Title	DFT Simulation of Novel Lithium Rich Double Anti-Perovskites for Solid Electrolyte in Lithium Ion Battery
Summary	In this project, we will use the density functional theory (DFT) approach to formulate and thoroughly investigate a series of novel inorganic solid electrolytes for the fast diffusion of Li+ ions. Specifically, we will perform systematic modeling on the structural stability and ion transport characteristics by fine-tuning the lattice chemistry of lithium-rich double anti-perovskite structures. This computational study will also lead to identifying effective ways to reduce the diffusion barriers for Li+ ions in double anti-perovskite phases via the suitable designation of lattice occupancy, which will be done by alloying on the chalcogen lattice site (A) and alternative occupancy of the halogen site (X). The high-performance computing facility will be used to run the simulations. Subsequently, the electrolytes for lithium-ion batteries will be synthesized at the Center for Advanced Photovoltaics at South Dakota State University to match with the theoretical results.
Job Description	The student will perform a series of optimized lattice structure modeling, convergence testing, ion/cell relaxation, band structure, and atom projected density of states calculations using the DFT approach as implemented in Quantum Espresso. The student will also import the previous codes in this project and cross-validate their results using CP2K. Additionally, the student will perform MD simulations using the large-scale atomic/molecular massively parallel simulator (LAMMPS) engine, optimized for GPU-computing. Finally, the student will write a paper/report on research findings for publication in a scientific manuscript and/or oral/poster presentations. This position requires a commitment of a minimum of 10 weeks at 30 hours per week.
Computational Resources	The student will work with the PI, Dr. Yue Zhou, and along with a group of graduate students (experimentalists) to validate the simulation results with lab experiment data. The simulations will be performed on the NSF-funded Roaring Thunder Cluster facility housed at the South Dakota State University.
Contribution to Community	This position will add new knowledge and possible innovation associated with the subject matter.
Position Type	Intern
Training Plan	The student was previously trained in parallel computing with OpenMP. This summer, the main focus of training would be GPU-based DFT simulation in Quantum Espresso. The student will be also trained in CP2K to cross-validate the previous DFT results. Additional training will be provided on LAMMPS and DeepMD Kit for implementing machine learning in molecular dynamics (MD) for predicting a new class of inorganic electrolytes for future study. The student will focus on a scientific literature review, running regular DFT simulations on the cluster, and meeting with the PI and mentor regularly to get advice/directions regarding the plan and progress of the research.
Student Prerequisites/Conditions/Qualifications	South Dakota State University will work with the student on a regular basis to ensure effective training. The student has also completed HPC Summer Boot Camp last year, Computational Chemistry, other training workshops, and webinars by XSEDE.
Duration	Summer
Start Date	05/17/2021
End Date	08/06/2021