Crystal-level accuracy.
Weeks, not months.
MaterSynq simulates crystal structures and surface energies at DFT accuracy in a fraction of the compute — shrinking cathode-material screening from eighteen months of trial-and-error to weeks of targeted experiments.
The full simulation stack, integrated.
From prototype enumeration to DFT-validated stability ranking — one workflow, no hand-off between disconnected tools.
Built for the full materials screening workflow
Cathode Material Screening
Rank hundreds of NMC, LFP, and novel oxide compositions by stability, voltage plateau, and thermal safety in a single compute campaign.
Learn moreSurface Energy Prediction
Compute surface termination energies for all relevant Miller planes. Predict SEI formation tendency and ionic conductivity at the cathode-electrolyte interface.
Learn moreCrystal Stability Ranking
Map formation energy on the convex hull. Filter dynamically unstable phases via phonon dispersion before committing to expensive synthesis.
Learn moreFrom hypothesis to ranked candidates
Define your composition space
Specify target chemistry (e.g., Ni-rich NMC with Co substitution range), oxidation states, and target properties. The platform enumerates symmetry-distinct prototype structures.
MLIP-accelerated relaxation
Physics-informed neural potentials pre-relax all candidates at a fraction of DFT cost. Structures are pre-filtered by geometric stability criteria.
DFT single-point validation
The top candidate pool receives full density functional theory calculations for formation energy, electronic structure, and surface termination energies.
Ranked results with experimental guidance
Output: a ranked stability table with synthesis conditions, predicted voltage plateaus, and flagged risk structures — ready for your experimental team.
Rigorous by design.
MaterSynq's accuracy benchmark: formation energy predictions within 50 meV/atom of DFT reference on held-out oxide test sets. Every MLIP deployed in the platform is validated against PBE+U calculations before production use.
Read the Methodology
Built in Pittsburgh, at the intersection of CMU materials science and applied computation.
Founded by Andrei Volkov. MaterSynq is an independent, bootstrapped laboratory tool — built for research teams who need computational accuracy without HPC infrastructure overhead.
About MaterSynqReady to compress your screening timeline?
Request early access. We onboard R&D labs selectively to ensure quality of support.