Toward Controllable Catalyst Inverse Design via Large-Scale Autoregressive Pretraining

2026-06-15

AI Summary

This work addresses the challenge of inverse design for heterogeneous catalysts, which is hindered by structural complexity and the difficulty of efficiently navigating vast chemical spaces. The authors propose a conditional generative model based on a generative pre-trained Transformer that, for the first time, integrates large-scale autoregressive pre-training (on 133 million structures) with explicit property conditioning—such as adsorption energy and catalyst class—and incorporates a numerical embedding layer. This enables high-fidelity, reaction-oriented catalyst generation without requiring additional fine-tuning. After fine-tuning on 460,000 optimized structures, the model achieves 98% structural validity, 95% optimization validity, and a 93% joint match rate between catalyst class and composition. Notably, the accuracy of adsorption energy conditioning improves fourfold, and screening efficiency increases by 1.5–4× compared to baseline approaches.

Abstract

Inverse design of heterogeneous catalysts remains challenging because catalyst surfaces exhibit substantial structural complexity with coupled surface-adsorbate interactions across a vast chemical space that is difficult to explore efficiently through conventional screening alone. Although machine learning-based high-throughput screening has accelerated catalyst discovery, its efficiency inevitably declines as the search space grows, motivating the development of generative models that can directly construct catalysts with target properties. Here, we present a conditional catalyst generative model based on the Generative Pretrained Transformer architecture with a numerical embedding layer that enables the generation of catalyst structures conditioned on both categorical and continuous properties within a single autoregressive framework. The model was pretrained on 133 million catalyst structures and subsequently fine-tuned on approximately 460,000 optimized structures with associated categorical properties and binding energies for conditional generation. The resulting model achieved 98% structural validity, 95% optimization validity, and high categorical condition fidelity, with a 93 % joint match rate for adsorbate type and composition. For binding energy conditioning, the match rate of approximately 20% represents a four-fold improvement over the baseline training distribution, and the generated distributions shift systematically toward the target values, enabling a 1.5 to 4-fold improvement in screening efficiency for reaction-targeted catalyst discovery without additional fine-tuning. These results show that large-scale autoregressive pre-training, combined with explicit property conditioning, provides a practical route toward controllable catalyst generation and accelerated catalysts discovery.

Problem

• inverse design
• heterogeneous catalysts
• structural complexity
• chemical space
• surface-adsorbate interactions

Innovation

• inverse design
• autoregressive pretraining
• conditional generative model
• heterogeneous catalysts
• property conditioning