Constrained Diffusion for Accelerated Structure Relaxation of Inorganic Solids with Point Defects
Jingyi Cui, Jacob K. Christopher, Ankita Biswas, Prasanna V. Balachandran, Ferdinando Fioretto
TL;DR
This work tackles the challenge of rapidly generating physically plausible point-defect structures in inorganic solids, where first-principles simulations are computationally expensive. It introduces a constrained diffusion framework that builds on score-based diffusion with a primal–dual augmented Lagrangian projection to enforce geometric, distributional, and force-based constraints, applied to final-state samples via $p_0$ instead of enforcing feasibility at every step. The key contributions include extending constrained diffusion to handle complex material constraints with neural surrogates, pioneering the use of constrained diffusion for Bi2Te3 defect generation, and delivering state-of-the-art performance across six defect configurations. The proposed method enables reliable, high-throughput defect structure generation with improved stability and realism, accelerating materials discovery and optimization for thermoelectric applications and potentially other layered crystalline systems.
Abstract
Point defects affect material properties by altering electronic states and modifying local bonding environments. However, high-throughput first-principles simulations of point defects are costly due to large simulation cells and complex energy landscapes. To this end, we propose a generative framework for simulating point defects, overcoming the limits of costly first-principles simulators. By leveraging a primal-dual algorithm, we introduce a constraint-aware diffusion model which outperforms existing constrained diffusion approaches in this domain. Across six defect configuration settings for Bi2Te3, the proposed approach provides state-of-the-art performance generating physically grounded structures.