Graph Neural Networks embedded into Margules model for vapor-liquid equilibria prediction

2025-02-26

AI Summary

This work addresses the insufficient accuracy of vapor–liquid equilibrium (VLE) prediction in early-stage process design. We propose a novel hybrid method that integrates a graph neural network (GNN) with an extended Margules model, trained exclusively on infinite-dilution activity coefficient data. To our knowledge, this is the first approach to embed a GNN within a classical thermodynamic activity coefficient framework—enabling high-accuracy VLE prediction even for systems where molecular fragmentation is infeasible or group-contribution parameters are unavailable. On diverse binary mixtures, the method consistently outperforms UNIFAC-Dortmund. Beyond offering a practical solution for data-scarce scenarios, this work establishes a lightweight “simple Gibbs free energy model + GNN” coupling paradigm as a new performance benchmark for VLE prediction, balancing theoretical innovation with engineering applicability.

Technology Category

• Machine Learning: Graph-based Machine Learning
• Reasoning under Uncertainty: Graphical Models
• Game Theory and Economic Paradigms: Adversarial Learning

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Predictive thermodynamic models are crucial for the early stages of product and process design. In this paper the performance of Graph Neural Networks (GNNs) embedded into a relatively simple excess Gibbs energy model, the extended Margules model, for predicting vapor-liquid equilibrium is analyzed. By comparing its performance against the established UNIFAC-Dortmund model it has been shown that GNNs embedded in Margules achieves an overall lower accuracy. However, higher accuracy is observed in the case of various types of binary mixtures. Moreover, since group contribution methods, like UNIFAC, are limited due to feasibility of molecular fragmentation or availability of parameters, the GNN in Margules model offers an alternative for VLE estimation. The findings establish a baseline for the predictive accuracy that simple excess Gibbs energy models combined with GNNs trained solely on infinite dilution data can achieve.

Problem

• Predict vapor-liquid equilibria using GNNs
• Compare GNN-Margules with UNIFAC-Dortmund model
• Evaluate GNNs for binary mixture accuracy

Innovation

• Graph Neural Networks
• Margules model
• vapor-liquid equilibrium prediction