(617j) Computational Quantification of Variation in CO2 Adsorption on Aluminum Substituted Zeolite Frameworks

Carbon-capture is an essential research domain due to the currently high emissions and consequences of global warming that we face. For carbon-capture, adsorption of CO₂ is critical in chemical processes to reduce the environmental impact of currently high emissions. Materials such as zeolites that are microporous aluminosilicate structures are used for gas separations, adsorption and ion-exchange [1]. These frameworks are built by connecting TO₄ tetrahedra (‘T’ denotes tetrahedrally coordinated Si, Al, or P, etc.). With various possible ways of connecting tetrahedra, there exist over 5 million hypothesized frameworks to date. Even amongst the synthesized zeolites, Al substitution in a pure Si-framework leads to significant changes in physical properties. Researchers have demonstrated the impact of Si/Al ratio on the observed physical properties of gas adsorption [2], [3]. Additionally, recent studies have shown that different configurations of Al substitutions are synthesizable for the same Si/Al ratio [4]. However, even for the same Si/Al ratio, there are multiple stable positions for Al and there is no effective way to overcome this uncertainty. The uncertainty in the Al positions leads to different properties and thus it is difficult to know exactly how much CO­₂ will be adsorbed.

Varying physical properties of materials is a challenge for process design since accurate estimation of physical properties is critical in process design. However, property estimation of these Al substituted zeolite frameworks is an extremely challenging computational task. The two major reasons that limit our ability to quantify these properties are the vast number of possible enumerations for substitutions and the time required for individual molecular simulations. In the case of CHA, having 6 substitutions in the unit cell (Si/Al = 5.00) leads to 1.9 million possible enumerations (following the Lowenstein’s rule) and molecular simulation at a single temperature pressure takes ~60 CPU hours. Thus, the traditional unit cell representation has limitations for enumeration of Al substituted framework generation. To that end, our proposed graph-theoretic representation of zeolite frameworks, the single repeating unit (SRU) can overcome these limitations [5]. For the CHA framework with Si/Al = 5.00, the SRU representation of 12 T-atoms has only 66 enumerated structures. The SRU representation is a novel contribution in the material representation space which has further applications in machine learning and property prediction [6]. Using the SRU representation, we enumerate the Al substitutions and perform molecular simulations to quantify the range of variation in physical properties based on the configuration of the Al substitution location.

Our results indicate that for the CHA framework, different configuration for the same Si/Al ratio leads to different adsorption capacities. The adsorption range varies over 12% for a Si/Al ratio of 5.00 at 298 K, 100 kPa. Though the computationally generated X-ray diffraction pattern for all these frameworks is the same, we observe key differences in the radial-distribution function of Al atoms. Overall, the insight that different positions of Al substitution leads to different properties opens up research for precise location of Al-substitution. Using the graph theoretic SRU representation and molecular simulations, the proposed method can be used to generate the optimal locations for Al substitutions in a zeolite framework.

References

[1] M. M. F. Hasan, E. L. First, and C. A. Floudas, “Cost-effective CO 2 capture based on in silico screening of zeolites and process optimization,” Phys. Chem. Chem. Phys., vol. 15, no. 40, pp. 17601–17618, 2013.

[2] F. N. Ridha and P. A. Webley, “Anomalous Henry’s law behavior of nitrogen and carbon dioxide adsorption on alkali-exchanged chabazite zeolites,” Sep. Purif. Technol., vol. 67, no. 3, pp. 336–343, 2009.

[3] F. N. Ridha and P. A. Webley, “Entropic effects and isosteric heats of nitrogen and carbon dioxide adsorption on chabazite zeolites,” Microporous mesoporous Mater., vol. 132, no. 1–2, pp. 22–30, 2010.

[4] S. Lee et al., “Evolution of Framework Al Arrangements in CHA Zeolites during Crystallization in the Presence of Organic and Inorganic Structure-Directing Agents,” Cryst. Growth Des., 2022.

[5] A. Gandhi and M. M. F. Hasan, “A graph theoretic representation and analysis of zeolite frameworks,” Comput. Chem. Eng., vol. 155, p. 107548, 2021.

[6] A. Gandhi and M. M. F. Hasan, “Machine learning for the design and discovery of zeolites and porous crystalline materials,” Curr. Opin. Chem. Eng., vol. 35, p. 100739, 2022.