The Battery Genome Project: AI for Energy Storage

About This Course

This three-week, hands-on course equips learners with practical skills to run atomic-scale simulations and perform data-driven screening for battery materials discovery. You’ll compute electronic properties using DFT, build an ML pipeline to rank candidate materials at scale using Materials Project data, and simulate Li-ion diffusion using molecular dynamics to connect transport behavior with charging-rate performance.

Each module includes a clear, tangible outcome—so you finish the course with real outputs you can showcase:

• Module 1: DOS plot

• Module 2: Top-5 candidate shortlist

• Module 3: Arrhenius diffusion plot

Aim

To train participants to predict, screen, and validate next-generation battery materials through an end-to-end workflow:
DFT (Quantum ESPRESSO) → ML ranking (Materials Project + Python) → Ion-transport insights (LAMMPS MD).

Course Objectives

By the end of this course, participants will be able to:

• Understand how DFT, ML, and MD contribute to battery materials discovery

• Run and interpret DFT outputs (band/DOS) relevant to conductivity and stability proxies

• Build a Materials Project → Python dataset pipeline for materials informatics

• Engineer descriptors using pymatgen/matminer, train and evaluate ML models, and rank materials

• Simulate ion diffusion in LAMMPS and compute diffusion coefficients across conditions

• Convert computational results into synthesis prioritization and performance-relevant insights

Course Structure

Module 1: Atomic-Scale Modeling with DFT (Quantum ESPRESSO)

• Battery materials discovery workflow: where DFT fits and what it predicts

• DFT essentials for energy materials: electronic structure, band basics, DOS interpretation

• Practical setup: pseudopotentials, k-points, convergence, input file structure

• Hands-on: Configure and run a Quantum ESPRESSO calculation for a candidate anode crystal and extract electronic outputs

Deliverable: A Density of States (DOS) plot with a short interpretation of conductivity relevance

Module 2: Machine Learning Screening at Scale (Materials Project + Python)

• Materials informatics pipeline: descriptors → model → ranking → validation strategy

• Data sourcing through the Materials Project API and dataset preparation

• Feature engineering (pymatgen/matminer), model selection, and evaluation basics

• Hands-on: Mine a large dataset, build features, train a Random Forest model to predict a target proxy (e.g., voltage/capacity-related)

Deliverable: A Top 5 predicted candidate materials shortlist for synthesis prioritization

Module 3: Ion Transport and Charging-Speed Insights (LAMMPS Molecular Dynamics)

• Why diffusion matters for rate performance: diffusion coefficient and transport basics

• MD setup fundamentals: force fields/potentials, temperature control, workflow

• Turning trajectories into insight: diffusion estimation and temperature dependence

• Hands-on: Run a Li-ion diffusion simulation in LAMMPS and compute diffusion coefficients across conditions

Deliverable: An Arrhenius plot projecting diffusion-driven performance across temperatures

Who Should Enrol?

This course is ideal for:

• UG/PG/PhD students, researchers, and professionals in Materials Science, Chemistry, Physics, Chemical Engineering, Energy, or related fields

• Learners interested in batteries, energy storage, computational materials, and materials informatics

Basic Python familiarity is helpful, but no prior DFT/MD experience is required (guided notebooks + templates provided).