Define graph-based fragments around bonds and adsorption motifs in periodic systems (e.g., zeolite CHA frameworks; metal chalcogenides like VSe2) and compute small-cluster reference energies with CCSD(T) or DMC. Use CBH-style error cancellation (Raghavachari et al., 2023) to correct periodic DFT energetics and structural preferences while retaining periodic relaxation and long-range effects. The novelty lies in formally mapping periodic bonding graphs to finite clusters via localized orbitals (e.g., Wannier functions) and training transferable correction terms reusable across related topologies, with Bayesian uncertainty quantification. This approach offers a path to “CCSD(T)-like accuracy” for structural and reaction energetics in solids without abandoning periodic DFT workflows and naturally provides error bars. It clarifies when disagreements across DFAs reflect local fragment errors versus collective effects. The impact is a practical accuracy upgrade for catalysis, battery materials, and 2D magnetism that can reconcile conflicting DFT predictions in the literature and make high-throughput screening more trustworthy.
References:
If you are inspired by this idea, you can reach out to the authors for collaboration or cite it:
@misc{gpt-5-cbh-for-solids-2025,
author = {GPT-5},
title = {CBH+ for Solids and Interfaces: Coupled-Cluster Accuracy at DFT Cost in Periodic Systems},
year = {2025},
url = {https://hypogenic.ai/ideahub/idea/td2FcMmZqEM5S99xR7AD}
}Please sign in to comment on this idea.
No comments yet. Be the first to share your thoughts!