journal article·2025-10-09Open access

Machine Learning-Based Predictions of Henry Coefficients for Long-Chain Alkanes in One-Dimensional Zeolites: Application to Hydroisomerization

Shrinjay Sharma, Ping Yang, Yachan Liu, Kevin Rossi, Peng Bai, Marcello S. Rigutto, Erik Zuidema, Umang Agarwal, Richard Baur, Sofia Calero, David Dubbeldam, Thijs J.H. Vlugt

The Journal of Physical Chemistry C· Volume 129· Issue 40

Abstract

High Resolution Image Download MS PowerPoint Slide Shape-selective adsorption in zeolites plays a pivotal role in catalytic hydroisomerization of long-chain alkanes, a key process in producing sustainable aviation fuels from Fischer–Tropsch products. Accurately predicting adsorption behavior for the large number of alkane isomers in different zeolite frameworks is computationally intensive. To address this, we have developed a machine learning framework that rapidly and accurately predicts Henry coefficients of linear (C 1 –C 30 ) and branched (C 4 –C 20 ) alkanes in one-dimensional zeolites. Using descriptors based on chain length, branching patterns, and molecular graphs, we evaluate multiple ML models, including Random Forest, XGBoost, CatBoost, TabPFN, and D-MPNN in MTT-, MTW-, MRE-, and AFI-type zeolites. TabPFN and D-MPNN offer the highest predictive accuracy. Active learning further boosts model performance by efficiently selecting diverse and structurally informative isomers. We also uncover activity cliffs, where small changes in molecular structure lead to sharp variations in adsorption, and demonstrate that targeted oversampling of these cases improves model robustness. Finally, we combine the ML-predicted Henry coefficients with gas-phase thermodynamics to compute reaction equilibrium distributions for C 16 hydroisomerization. This integrated, data-driven approach enables efficient screening and design of shape-selective zeolite catalysts, thereby reducing the need for costly simulations.
Source: openalex_inverted_index · Quality: baseline

Research context

OpenAlex topics: Machine Learning in Materials Science · Phase Equilibria and Thermodynamics · Catalysis and Oxidation Reactions

SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.

Affiliations

Process and Energy Department, Faculty of Mechanical Engineering, Delft University of Technology, Leeghwaterstraat 39, 2628CB Delft, The Netherlands

Department of Applied Physics and Science Education

Eindhoven University of Technology

Department of Chemical Engineering

University of Massachusetts Amherst

Department of Materials Science and Engineering, Faculty of Mechanical Engineering, Delft University of Technology, Mekelweg 2, 2628CD Delft, The Netherlands

Climate Safety and Security Centre, TU Delft The Hague Campus

Delft University of Technology

Shell Global Solutions International B.V., Grasweg 39, 1031HW Amsterdam, The Netherlands

Shell Chemical LP

Van’t Hoff Institute of Molecular Sciences, University of Amsterdam, Science Park 904, 1098XH Amsterdam, The Netherlands

Funding

Exacte en Natuurwetenschappen · Defense Advanced Research Projects Agency · Office of Advanced Cyberinfrastructure · Advanced Research Center for Chemical Building Blocks

References (77)

Show source-provided references
1. 10.1021/acsenergylett.0c0141810.1021/acsenergylett.0c01418 ↗
2. Calis H. · 2021 · 16110.1002/9783527813490.ch7 ↗
3. 10.1038/nature0655210.1038/nature06552 ↗
4. 10.1063/5.020921010.1063/5.0209210 ↗
5. 10.1021/acs.jpcb.4c0535510.1021/acs.jpcb.4c05355 ↗
6. Frenkel D. · 2023
7. 10.1186/s13321-022-00677-610.1186/s13321-022-00677-6 ↗
8. 10.1021/ci00057a00510.1021/ci00057a005 ↗
9. Swain, M. PubChemPy: A Python wrapper for the PubChem PUG REST API, 2023. https://github.com/mcs07/PubChemPy (accessed 01/12/2024).
10. 10.1093/nar/gkv95110.1093/nar/gkv951 ↗
11. 10.1186/s13321-021-00529-910.1186/s13321-021-00529-9 ↗
12. 10.1186/s13321-022-00604-910.1186/s13321-022-00604-9 ↗
13. Baerlocher C. · 2007
14. Favre H. A. · 201310.1039/9781849733069 ↗
15. 10.1063/1.173411010.1063/1.1734110 ↗
16. 10.1080/0026897920010006110.1080/00268979200100061 ↗
17. 10.1080/0026897930010206110.1080/00268979300102061 ↗
18. 10.1021/jacsau.4c0042910.1021/jacsau.4c00429 ↗
19. 10.1063/5.025384710.1063/5.0253847 ↗
20. 10.1039/D5DD00221D10.1039/D5DD00221D ↗
21. 10.1039/D2SC03351H10.1039/D2SC03351H ↗
22. 10.1021/acs.iecr.4c0350010.1021/acs.iecr.4c03500 ↗
23. 10.1021/acs.chemrev.0c0000410.1021/acs.chemrev.0c00004 ↗
24. Petković, M.; Vicent-Luna, J. M.; Menkovski, V.; Calero, S. Zeolite adsorption property prediction using deep learning. arXiv e-prints arXiv–2403, 2024. https://arxiv.org/html/2403.12659v1.
25. 10.1021/acs.jpcc.1c0526610.1021/acs.jpcc.1c05266 ↗
26. 10.1021/acs.iecr.1c0387610.1021/acs.iecr.1c03876 ↗
27. 10.1021/acs.jpcc.4c0006610.1021/acs.jpcc.4c00066 ↗
28. 10.1039/D3TA01911J10.1039/D3TA01911J ↗
29. 10.1109/5.72679110.1109/5.726791 ↗
30. 10.1021/jp906984z10.1021/jp906984z ↗
31. 10.1002/anie.20035111010.1002/anie.200351110 ↗
32. 10.1021/jp035200m10.1021/jp035200m ↗
33. 10.1080/08927022.2018.142685510.1080/08927022.2018.1426855 ↗
34. 10.1021/acs.jcim.2c0090310.1021/acs.jcim.2c00903 ↗
35. 10.1021/jm400428510.1021/jm4004285 ↗
36. 10.1021/ci700409310.1021/ci7004093 ↗
37. Tanimoto T. T. · 1958
38. 10.1021/ci990316u10.1021/ci990316u ↗
39. Levenshtein V. I. · Sov. Phys. Dokl. · 1966 · 10 · 707
40. 10.1021/jm100933w10.1021/jm100933w ↗
41. 10.1021/jm200452d10.1021/jm200452d ↗
42. 10.1021/acscentsci.7b0057210.1021/acscentsci.7b00572 ↗
43. 10.1021/jm201706b10.1021/jm201706b ↗
44. 10.1093/bioinformatics/btn18610.1093/bioinformatics/btn186 ↗
45. 10.1038/s41589-022-01131-210.1038/s41589-022-01131-2 ↗
46. 10.1023/A:101093340432410.1023/A:1010933404324 ↗
47. Chen T. · 2016 · 78510.1145/2939672.2939785 ↗
48. Dorogush, A. V.; Ershov, V.; Gulin, A. CatBoost: gradient boosting with categorical features support. arXiv preprint, submitted on 24-10-2018. 10.48550/arXiv.1810.11363.10.48550/arXiv.1810.11363 ↗
49. Prokhorenkova, L.; Gusev, G.; Vorobev, A.; Dorogush, A. V.; Gulin, A. CatBoost: Unbiased Boosting with Categorical Features. In Advances in Neural Information Processing Systems 31 (NeurIPS 2018). arXiv preprint, submitted on 28-06-2017; 10.48550/arXiv.1706.09516.10.48550/arXiv.1706.09516 ↗
50. Hollmann, N.; Müller, S.; Eggensperger, K.; Hutter, F. TabPFN: A transformer that solves small tabular classification problems in a second. In The Eleventh International Conference on Learning Representations. arXiv preprint, submitted on 05-07-2022; 10.48550/arXiv.2207.01848.10.48550/arXiv.2207.01848 ↗
51. 10.1038/s41586-024-08328-610.1038/s41586-024-08328-6 ↗
52. 10.1021/acs.jcim.9b0023710.1021/acs.jcim.9b00237 ↗
53. 10.1021/acs.jcim.1c0097510.1021/acs.jcim.1c00975 ↗
54. Jin, W.; Barzilay, R.; Jaakkola, T. Multi-objective molecule generation using interpretable substructures. In 37th International Conference on Machine Learning. arXiv preprint, submitted on 08-02-2022; 10.48550/arXiv.2002.03244.10.48550/arXiv.2002.03244 ↗
55. 10.1021/acs.jcim.3c0125010.1021/acs.jcim.3c01250 ↗
56. 10.1080/00268979816788110.1080/002689798167881 ↗
57. 10.1080/0892702990802211210.1080/08927029908022112 ↗
58. 10.1080/08927022.2013.81910210.1080/08927022.2013.819102 ↗
59. 10.1080/08927022.2015.101008210.1080/08927022.2015.1010082 ↗
60. 10.1063/5.022624910.1063/5.0226249 ↗
61. 10.1039/dc978660009510.1039/dc9786600095 ↗
62. 10.1021/jp972543+10.1021/jp972543+ ↗
63. 10.1098/rspa.1924.008110.1098/rspa.1924.0081 ↗
64. 10.1021/jp037672710.1021/jp0376727 ↗
65. 10.1021/jp407422410.1021/jp4074224 ↗
66. 10.1021/jp026393110.1021/jp0263931 ↗
67. 10.1002/andp.1881248011010.1002/andp.18812480110 ↗
68. Berthelot D. · C. R. · 1898 · 126 · 15
69. Pedregosa F. · J. Mach. Learn. Res. · 2011 · 12 · 2825
70. 10.1186/s13321-020-00456-110.1186/s13321-020-00456-1 ↗
71. 10.1021/acs.jcim.3c0143010.1021/acs.jcim.3c01430 ↗
72. 10.1039/C7CP08017D10.1039/C7CP08017D ↗
73. Levenspiel O. · 1998
74. James G. · 201310.1007/978-1-4614-7138-7 ↗
75. Chen L. · 2024 · 496
76. Shen, W. X.; Cui, C.; Su, X.; Zhang, Z.; Velez-Arce, A.; Wang, J.; Shi, X.; Zhang, Y.; Wu, J.; Chen, Y. Z.; Zitnik, M. Activity Cliff-Informed Contrastive Learning for Molecular Property Prediction. ChemRxiv preprint, 2024. https://chemrxiv.org/engage/chemrxiv/article-details/6703d9c351558a15ef5b9e06.
77. Shirekar, O. K.; Singh, A.; Jamali-Rad, H. Self-attention message passing for contrastive few-shot learning. arXiv preprint, 2023. https://arxiv.org/abs/2210.06339.10.1109/WACV56688.2023.00539 ↗