Prediction of infinite-dilution activity coefficients with neural collaborative filtering

Accurate prediction of infinite dilution activity coefficient (gamma(infinity)) for phase equilibria and process design is crucial. In this work, an experimental gamma(infinity) dataset containing 295 solutes and 407 solvents (21,048 points) is obtained through data integrating, cleaning, and filtering. The dataset is arranged as a sparse matrix with solutes and solvents as columns and rows, respectively. Neural collaborative filtering (NCF), a modern matrix completion technique based on deep learning, is proposed to fully fill in the gamma(infinity) matrix. Ten-fold cross-validation is performed on the collected dataset to test the effectiveness of the proposed NCF, proving that NCF outperforms the state-of-the-art physical model and previous machine learning model. The completed gamma(infinity) matrix makes solvent screening and extension of UNIFAC parameters possible. Taking two typical hard-to-separate systems (benzene/cyclohexane and methyl cyclopentane/n-hexane mixtures) as examples, the NCF-developed database provides high-throughput screening for separation systems in terms of solvent selectivity and capacity.

Automated summary

Accurate prediction of infinite dilution activity coefficient (gamma(infinity)) is crucial for phase equilibria and process design. This study proposes a new method based on neural collaborative filtering (NCF) to fill in the gamma(infinity) matrix, and the experimental results show that it outperforms traditional models and previous machine learning models. The completed matrix can also be used for solvent screening and parameter extension.