How to evaluate solid electrolyte candidates
Solid electrolyte selection is one of the highest-stakes decisions in battery development. This guide walks through the evaluation framework used by Lattice Graph teams to narrow thousands of candidates to a defensible shortlist.
Why solid electrolyte selection is hard
Solid-state electrolytes must simultaneously satisfy ionic conductivity, electrochemical stability, mechanical compatibility, and manufacturability constraints. Public databases disagree on key properties for many candidates, making it difficult to know which values to trust. Teams that rely on a single source risk optimizing against measurement artifacts rather than material reality.
Step 1: Define your operating envelope
Start by specifying the non-negotiable constraints for your cell architecture: minimum ionic conductivity at operating temperature, voltage window, target areal resistance, and compatibility with your chosen anode and cathode chemistries. These constraints eliminate the majority of candidates before any detailed analysis begins.
Step 2: Cross-source confidence screening
Use Lattice Graph cross-source confidence to identify candidates where multiple independent DFT databases agree on stability and electronic properties. Materials with high cross-source agreement on formation energy and band gap are more likely to behave as predicted in practice. Flag candidates where sources disagree by more than 0.1 eV/atom for manual review.
Step 3: Rank by composite fitness
Combine ionic conductivity estimates, stability scores, provenance depth, and any available experimental validation into a weighted ranking. The Lattice Graph battery workflow automates this ranking and produces an exportable evidence package for each candidate, including source attribution and confidence intervals.
Step 4: Validate and iterate
Select the top 5-10 candidates for simulation or experimental validation. Use PyBaMM digital twin workflows to model cycling behavior before committing to synthesis. Feed experimental results back into your confidence model to improve future screening rounds.
Ready to try it yourself?
Start exploring materials data with cross-source confidence, simulation workflows, and the tools described in this resource.