Machine learning for prediction of CO₂/N₂/H₂O selective adsorption and separation in metal-zeolites

Abstract

Carbon dioxide (CO₂) capture, utilization, and storage technologies are crucial in reducing global warming and producing various high-value chemicals and fuels. It is challenging to effectively separate the molecules of CO₂, N₂, and H₂O, whose kinetic diameters are close to each other. Although zeolites have garnered considerable attention in gas separation, the huge chemical space of metal-doped zeolites (metal-zeolites) coming from the combination of different metal active sites, topology, and Si/Al ratios poses a difficulty in finding an optimal material for selectively trapping CO₂. In this study, we build machine learning (ML) models to predict the selective adsorption of CO₂/N₂/H₂O on metal-zeolites through the regulation of electrostatic polarization interaction. The stability of 208 metal-zeolites encompassing five distinct topological structures is estimated through the formation energy (E_f), indicating the potential accessibility in the experiment for most of the studied systems, especially for Sc-, Y-, and Zr-zeolites. Adsorption of CO₂ on metal sites has two possible configurations: linear vs. bent CO₂, depending on different embedded metals. The concerted binding of CO₂ with both carbon and oxygen atoms on the metal center leads to the bent geometry and larger binding energies on metal-zeolites (Zr-, Nb-, Mo-zeolites). Accessible descriptors associated with the zeolites, adsorbates, and metals are selected to train the adsorption strength index (I), showing good performance [mean absolute error (MAE) = 0.04, R² = 0.88]. The predicted adsorption selectivity is in agreement with the experimental systems (Co-, Zn-, Cu-, Fe-SSZ-13). It is found that medium-pore-sized zeolites [pore limiting diameter (PLD) = ~7 Å] anchored with Zr, Nb, or Mo are promising materials for the CO₂ adsorption and separation. The proposed ML scheme may also be applicable to give a fast prediction of CO₂ adsorption and separation ability in other porous metal-organic frameworks or amorphous materials.

Keywords

Metal-zeolites / machine learning / separation / selective adsorption

References

[1] Velty A.Advanced zeolite and ordered mesoporous silica-based catalysts for the conversion of CO₂ to chemicals and fuels.Chem Soc Rev2023;52:1773-946

[2] Sifat NS.A critical review of CO₂ capture technologies and prospects for clean power generation.Energies2019;12:4143

[3] Usman M,Noor T.Advanced strategies in metal-organic frameworks for CO₂ capture and separation.Chem Rec2022;22:e202100230

[4] Madejski P,Subramanian N.Methods and techniques for CO₂ capture: review of potential solutions and applications in modern energy technologies.Energies2022;15:887

[5] Tomaszewski R.Citations to chemical resources in scholarly articles: CRC handbook of chemistry and physics and the merck index.Scientometrics2017;112:1865-79

[6] Wang H,Bai J.Multi-factor study of the effects of a trace amount of water vapor on low concentration CO₂ capture by 5A zeolite particles†.RSC Adv2020;10:6503-11 PMCID:PMC9049640

[7] Jia T,Li F.Progress and potential of metal-organic frameworks (MOFs) for gas storage and separation: a review.J Environ Chem Eng2022;10:108300

[8] Zhao M,Gu XS.MOF based CO₂ capture: adsorption and membrane separation.Inorg Chem Commun2023;152:110722

[9] Osman AI,Abdel Maksoud MIA,Rooney DW.Recent advances in carbon capture storage and utilisation technologies: a review.Environ Chem Lett2021;19:797-849

[10] Cheung O,Bacsik Z.Silicoaluminophosphates as CO₂ sorbents.Microporous Mesoporous Mater2012;156:90-6

[11] Li Y,Wang C.Achieving highly selective CO₂ adsorption on SAPO-35 zeolites by template-modulating the framework silicon content†.Chem Sci2022;13:5687-92 PMCID:PMC9116366

[12] Sun M,Hanif A,Shang J.Transition metal cation-exchanged SSZ-13 zeolites for CO₂ capture and separation from N₂.J Chem Eng2019;370:1450-8

[13] Gu Y,Yu C.Zeolite adsorption isotherms predicted by pore channel and local environment descriptors: feature learning on DFT binding strength.J Phys Chem C2020;124:9314-28

[14] Fu D,Davis ME.Zinc containing small-pore zeolites for capture of low concentration carbon dioxide.Angew Chem Int Ed Engl2022;61:e202112916

[15] Gu Y,Liu Z.Nitrogen reduction reaction energy and pathways in metal-zeolites: deep learning and explainable machine learning with local acidity and hydrogen bonding features.J Mater Chem A2022;10:14976-88

[16] Zhu Q,Liang X,Ma J.A machine learning model to predict CO₂ reduction reactivity and products transferred from metal-zeolites.ACS Catal2022;12:12336-48

[17] Choi HJ,Ahn SH.Framework flexibility-driven CO₂ adsorption on a zeolite†.Mater Horiz2020;7:1528-32

[18] Moliner M,Corma A.Machine learning applied to zeolite synthesis: the missing link for realizing high-throughput discovery.Acc Chem Res2019;52:2971-80

[19] Wu LM,Xie FM,Englezos P.Minireview of hydrate-based CO₂ separation from a CO₂/CH₄ gas mixture: progress and outlook.Energy Fuels2022;36:10478-88

[20] Yuan Y,Ricardez-sandoval L.Recent advances on first-principles modeling for the design of materials in CO₂ capture technologies.Chin J Cheml Eng2019;27:1554-65

[21] Schwalbe-Koda D,Paris C.A priori control of zeolite phase competition and intergrowth with high-throughput simulations.Science2021;374:308-15

[22] Zhang H,Rahimi M.Improving predictions of shale wettability using advanced machine learning techniques and nature-inspired methods: implications for carbon capture utilization and storage.Sci Total Environ2023;877:162944

[23] Wang S,Dai S.Prediction by convolutional neural networks of CO₂/N₂ selectivity in porous carbons from N₂ adsorption isotherm at 77 K.Angew Chem Int Ed Engl2020;59:19645-8

[24] Srinivasu K,Ghanty TK.Confinement directed adsorption of noble gases (Xe/Kr) in MFM-300(M)-based metal-organic framework materials.J Phys Chem C2019;123:27531-41

[25] Baerlocher C. Database of zeolite structures. Available from: http://www.iza-structure.org/databases/. [Last accessed on 4 Sep 2023]

[26] Cai X,Bingel LW,Walton KS.A collection of more than 900 gas mixture adsorption experiments in porous materials from literature meta-analysis.Ind Eng Chem Res2021;60:639-51

[27] Vandevondele J,Mohamed F,Chassaing T.Quickstep: fast and accurate density functional calculations using a mixed Gaussian and plane waves approach.Comput Phys Commun2005;167:103-28

[28] Grimme S,Ehrlich S.A consistent and accurate ab initio parametrization of density functional dispersion correction (DFT-D) for the 94 elements H-Pu.J Chem Phys2010;132:154104

[29] Goedecker S,Hutter J.Separable dual-space Gaussian pseudopotentials.Phys Rev B Condens Matter1996;54:1703-10

[30] Hartwigsen C,Hutter J.Relativistic separable dual-space Gaussian pseudopotentials from H to Rn.Phys Rev B1998;58:3641-62

[31] VandeVondele J.Gaussian basis sets for accurate calculations on molecular systems in gas and condensed phases.J Chem Phys2007;127:114105

[32] Broyden CG. The convergence of single-rank quasi-Newton methods. Math Comput 1970;24:365-82. Available from: https://www.ams.org/journals/mcom/1970-24-110/S0025-5718-1970-0279993-0/S0025-5718-1970-0279993-0.pdf. [Last accessed on 4 Sep 2023]

[33] Shanno DF.Conditioning of quasi-Newton methods for function minimization.Math Comput1970;24:647-56Available from: https://www.ams.org/journals/mcom/1970-24-111/S0025-5718-1970-0274029-X/S0025-5718-1970-0274029-X.pdf. [Last accessed on 4 Sep 2023]

[34] Kresse G.Efficiency of ab-initio total energy calculations for metals and semiconductors using a plane-wave basis set.Comput Mater Sci1996;6:15-50

[35] Blöchl PE.Projector augmented-wave method.Phys Rev B Condens Matter1994;50:17953-79

[36] Perdew JP,Ernzerhof M.Generalized gradient approximation made simple.Phys Rev Lett1996;77:3865-8

[37] Grimme S,Goerigk L.Effect of the damping function in dispersion corrected density functional theory.J Comput Chem2011;32:1456-65

[38] Pedregosa F,Gramfort A,Thirion B.Scikit-learn: machine learning in python.J Mach Learn Res2011;12:2825-30Available from: https://www.jmlr.org/papers/volume12/pedregosa11a/pedregosa11a.pdf. [Last accessed on 4 Sep 2023]

[39] Dietterich TG.Approximate statistical tests for comparing supervised classification learning algorithms.Neural Comput1998;10:1895-923

[40] Guo Y,Wei D.Robust SAR automatic target recognition via adversarial learning.IEEE J Sel Top Appl Earth Observations Remote Sensing2021;14:716-29

[41] Friedman JH.Stochastic gradient boosting.Comput Stat Data Anal2002;38:367-78

[42] Rodríguez JJ,Bustillo A.A decision-making tool based on decision trees for roughness prediction in face milling.Int J Comput Integr Manuf2017;30:943-57

[43] Patrick EA.A generalized k-nearest neighbor rule.Inf Control1970;16:128-52

[44] D’Angelo GM,Gu CC.Combining least absolute shrinkage and selection operator (LASSO) and principal-components analysis for detection of gene-gene interactions in genome-wide association studies.BMC Proc2009;3 Suppl 7:S62 PMCID:PMC2795963

[45] Hoerl RW.Ridge regression: a historical context.Technometrics2020;62:420-5

[46] Schumann TEW.XXVI. The principles of a mechanical method for calculating regression equations and multiple correlation coefficients and for the solution of simultaneous linear equations.Lond, Edinb, Dublin Philos Mag J Sci1940;29:258-73

[47] Abiodun OI,Omolara AE.Comprehensive review of artificial neural network applications to pattern recognition.IEEE Access2019;7:158820-46

[48] Seko A,Tsuda K.Machine learning with systematic density-functional theory calculations: application to melting temperatures of single- and binary-component solids.Phys Rev B2014;89:054303

[49] Li Y.New stories of zeolite structures: their descriptions, determinations, predictions, and evaluations.Chem Rev2014;114:7268-316

[50] Gu Y,Yu X.Zeolite adsorption database.J Phys Chem C2020;124:9314-28Available from: http://106.15.196.160:5656/index/en. [Last accessed on 4 Sep 2023]

[51] Jeffroy M,Boutin A.New molecular simulation method to determine both aluminum and cation location in cationic zeolites.Chem Mater2017;29:513-23

[52] Li Y,Fan W.A DFT study on the distributions of Al and Brönsted acid sites in zeolite MCM-22.J Mol Catals A Chem2011;338:24-32

[53] Hudson MR,Mason JA,Lobo RF.Unconventional, highly selective CO₂ adsorption in zeolite SSZ-13.J Am Chem Soc2012;134:1970-3

[54] Anderson R,Gómez-Gualdrón DA.Adsorption isotherm predictions for multiple molecules in MOFs using the same deep learning model.J Chem Theory Comput2020;16:1271-83

[55] Caro-Ortiz S,Rigutto M,Vlugt TJH.Competitive adsorption of Xylenes at chemical equilibrium in zeolites.J Phys Chem C Nanomater Interfaces2021;125:4155-74 PMCID:PMC8025683

[56] Wilson SMW.CO₂ and CO adsorption equilibrium on ZSM-5 for different SiO₂/Al₂O₃ ratios.Sep Sci Technol2019;54:722-30

[57] Prodinger S.Recent progress to understand and improve zeolite stability in the aqueous medium.Pet Chem2020;60:420-36

[58] Heard CJ,Rice CM.Fast room temperature lability of aluminosilicate zeolites.Nat Commun2019;10:4690 PMCID:PMC6795794

[59] Feng J,Bao Q,Xu Y.Carbon monoxide and carbon dioxide adsorption on alkali metal cation-exchanged SSZ-13 zeolites.Micro Nano Letters2020;15:529-34

[60] Chen S,Tang Y,Li W.Molecular simulation and experimental investigation of CO₂ capture in a polymetallic cation-exchanged 13X zeolite.J Mater Chem A2018;6:19570-83

[61] Crawford P.Importance of electronegativity differences and surface structure in molecular dissociation reactions at transition metal surfaces.J Phys Chem B2006;110:24929-35

[62] Wang L,Zhang D.Surface strategies for catalytic CO₂ reduction: from two-dimensional materials to nanoclusters to single atoms†.Chem Soc Rev2019;48:5310-49

[63] Yue B,Chai Y,Guan N.Zeolites for separation: fundamental and application.J Energy Chem2022;71:288-303

[64] Fu D.Carbon dioxide capture with zeotype materials.Chem Soc Rev2022;51:9340-70

[65] Joos L,Smit B.Molecular simulation study of the competitive adsorption of H₂O and CO₂ in zeolite 13X.Langmuir2013;29:15936-42

[66] Lee KM,Jo YM.Evaluation of moisture effect on low-level CO₂ adsorption by ion-exchanged zeolite.Environ Technol2012;33:77-84

[67] Mason JA. Metal-organic frameworks for gas storage and separation. Available from: https://escholarship.org/uc/item/07w7s2d2. [Last accessed on 4 Sep 2023]

[68] Brandani F.The effect of water on the adsorption of CO₂ and C₃H₈ on type X zeolites.Ind Eng Chem Res2004;43:8339-44

[69] Shafeeyan MS,Shamiri A.Adsorption equilibrium of carbon dioxide on ammonia-modified activated carbon.Chem Eng Res Des2015;104:42-52

[70] Huong PT.Improvement of selective separation of CO₂ over N₂ by transition metal-exchanged nano-zeolite.Microporous Mesoporous Mater2017;241:155-64