(698b) Process-Informed Optimization of Metal–Organic Frameworks for Pressure Swing Adsorption

The use of fossil fuel generates plentiful CO₂ that causes global climate change[1]. Among the various sources, the considerable CO₂ emissions from coal-fired power plants, gas-fired power plants, petrochemical and other industry, could be tackled with post-combustion carbon capture, which calls for efficient processes to separate N₂ and CO₂ [2,3].

Pressure swing adsorption (PSA) is regarded as the primary candidate process to conduct post-combustion carbon capture. In PSA, the adsorption and desorption of gas is controlled by adjusting the pressure in a series of adsorption columns [1]. As for the adsorbent material, metal-organic frameworks (MOFs) are considered as promising candidates [2]. MOFs are an emerging class of porous materials that are formed by repeatedly connected metal nodes and organic linkers [4]. These secondary building units, i.e., metal nodes and organic linkers, are highly tunable, with different combinations being able to make MOFs that possess different structures and properties [5]. This highly tunable nature of MOFs gives rise to a great number of possible structures, leading to a vast design space when one aims to select a MOF that is to exhibit high performance in a given application.

To this purpose, we aim to conduct co-optimization of the geometric descriptors of MOFs and the process parameters of PSA in order to identify the best material as well as process design that can enhance PSA performance in separating nitrogen from carbon dioxide. In our previous work [6], we established a machine-learning based workflow to obtain surrogate models that could predict the adsorption isotherms given specific MOF descriptors without having to resort to highly time-consuming GCMC simulations that are traditionally used to acquire the adsorption isotherms of given MOFs. The adsorption isotherms are essential for estimating the carbon capture performance of MOFs, and for encoding their performance in the context of a PSA process model. In this work, we integrate such surrogate models with an established PSA process simulator [7], towards a combined material-process model. Employing the latter as a black-box, we utilize the well-known derivative-free optimization software NOMAD (Nonlinear Optimization by Mesh Adaptive Direct Search) [8] to identify MOF structures (geometric descriptors) that lead to optimal values of process related metrics.

Compared to a baseline MOF, UTSA-16, we show how a carefully designed MOF can exhibit up to 80% increase in productivity for standard PSA process conditions. Additionally, we found the MOF design could increase the purity and the recovery of the carbon capture process by approximately 130% and 80%, respectively, while decrease the energy requirement by roughly 30%, compared to the baseline MOF. Moreover, we found that the enhancement of the productivity and that of recovery, purity, and energy requirement often go hand-in-hand through material optimization, with certain MOF structures showing improvements in all metrics at the same time. While our optimal sets of MOF geometric descriptors can serve as targets for experimental synthesis, this work more broadly demonstrates that there is ample potential to further improve our current knowledge of what constitutes a good MOF adsorbent for post-combustion carbon capture.

References:

[1] Farmahini, A.H., et al., Performance-Based Screening of Porous Materials for Carbon Capture. Chem Rev, 2021. 121(17)

[2] Taddei, M. and C. Petit, Engineering metal–organic frameworks for adsorption-based gas separations: from process to atomic scale. Molecular Systems Design & Engineering, 2021. 6(11)

[3] Bui, M., et al., Carbon capture and storage (CCS): the way forward. Energy & Environmental Science, 2018. 11(5)

[4] Ghanbari, T., F. Abnisa, and W.M.A. Wan Daud, A review on production of metal organic frameworks (MOF) for CO(2) adsorption. Sci Total Environ, 2020. 707

[5] Gutiérrez-Serpa, A., et al., Metal–Organic Frameworks as Key Materials for Solid-Phase Microextraction Devices—A Review. Separations, 2019. 6(4).

[6] Yin, X. et al., Inverse Design of Metal-Organic Frameworks for Adsorption Processes: Learning Surrogate Models of Isotherm Parameters, Forthcoming, 2024

[7] Yancy-Caballero, D., et al., Process-level modelling and optimization to evaluate metal–organic frameworks for post-combustion capture of CO2. Molecular Systems Design & Engineering, 2020. 5(7)

[8] Audet, C., et al., Nonlinear Optimization with the MADS Algorithm. ACM Trans. Math. Softw., 2022. 48(3)