Solar-GECO: Perovskite Solar Cell Property Prediction with Geometric-Aware Co-Attention

2025-11-24

AI Summary

Perovskite solar cell (PSC) performance critically depends on intricate interfacial interactions among multilayer materials; however, conventional experimental screening is inefficient and costly, while existing machine learning models often neglect crystal geometric structure or focus solely on isolated properties, limiting their accuracy in predicting power conversion efficiency (PCE). To address this, we propose a novel multimodal fusion model that— for the first time—integrates an atomic-level geometric graph neural network with device-level textual embeddings via a geometry-aware co-attention mechanism, explicitly modeling interlayer interactions and quantifying prediction uncertainty. The model incorporates a probabilistic regression head to enable joint learning from heterogeneous, multi-source data. Evaluated on PCE prediction, it achieves state-of-the-art performance with a mean absolute error of 2.936, substantially outperforming prevailing approaches.

Technology Category

• Machine Learning: Multimodal Learning
• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Reasoning under Uncertainty: Relational Probabilistic Models

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

Perovskite solar cells are promising candidates for next-generation photovoltaics. However, their performance as multi-scale devices is determined by complex interactions between their constituent layers. This creates a vast combinatorial space of possible materials and device architectures, making the conventional experimental-based screening process slow and expensive. Machine learning models try to address this problem, but they only focus on individual material properties or neglect the important geometric information of the perovskite crystal. To address this problem, we propose to predict perovskite solar cell power conversion efficiency with a geometric-aware co-attention (Solar-GECO) model. Solar-GECO combines a geometric graph neural network (GNN) - that directly encodes the atomic structure of the perovskite absorber - with language model embeddings that process the textual strings representing the chemical compounds of the transport layers and other device components. Solar-GECO also integrates a co-attention module to capture intra-layer dependencies and inter-layer interactions, while a probabilistic regression head predicts both power conversion efficiency (PCE) and its associated uncertainty. Solar-GECO achieves state-of-the-art performance, significantly outperforming several baselines, reducing the mean absolute error (MAE) for PCE prediction from 3.066 to 2.936 compared to semantic GNN (the previous state-of-the-art model). Solar-GECO demonstrates that integrating geometric and textual information provides a more powerful and accurate framework for PCE prediction.

Problem

• Predicting perovskite solar cell efficiency using geometric-aware co-attention models
• Addressing complex multi-scale interactions in perovskite solar cell components
• Overcoming limitations of traditional ML models that neglect crystal geometry

Innovation

• Geometric graph neural network encodes perovskite atomic structure
• Language model embeddings process transport layer chemical compounds
• Co-attention module captures inter-layer interactions and dependencies