(294a) Molecular Simulations to Understand Adsorption and Transport in Complex Zeolite Nanosheet Membranes for Ammonia/Nitrogen/Hydrogen Separation

Ammonia is an industrially relevant chemical with uses in fertilizer synthesis and energy storage. Conventionally, ammonia is produced through the catalytic Haber–Bosh process at pressures of 100–250 bar and temperatures of 650–750 K, followed by separation that occurs at much lower temperatures. However, the increasing demand coupled with the drive toward small-scale distributed manufacturing of ammonia requires a facile and energy-efficient ammonia separation from the stream of unreacted gases without cooling down the gas mixture. Hence, we want to explore the utilization of membrane–based separation at process–relevant conditions for effecting this separation. Our previous work has shown substantial ammonia selectivity at the ambient^[1] and process–relevant conditions^[2] for bulk MFI zeolite and MFI nanosheets (NS) with silanols at the surface. However, experimentally synthesized membranes are composed of multiple nanosheets layered on top of each other with non-uniform interfacial-sheet regions. Simulations can provide molecular-level insights and predictions for the separation performance, which is challenging to study via laboratory experiments because of high temperature and pressure conditions relevant for the industrial ammonia production process.

Here, we investigate complex adsorbents formed by stacking MFI nanosheets, referred to as stacked nanosheets (SNS) to understand the influence of multi–layer nanosheets on gas adsorption, separation, and diffusion. These SNS are generated through reactive force-field-based molecular dynamics simulations. Realistically, the nanosheets do not perfectly overlap with each other in the interfacial region and modeling these SNS containing misaligned channels can provide further insights into their effect on relative transport behavior of NH₃, N₂, and H₂.

We performed NpT-Gibbs ensemble Monte Carlo (GEMC) simulations to understand the mixture adsorption behavior of SNS models, which show that NH₃ selectively adsorbs over N₂ or H₂ in the interfacial-sheet region of SNS with silanols, and its enrichment in this region consequently affects the transport properties. Considering all these possibilities to probe SNS models with varying translational and rotational shifts, along with various mixture compositions require a large number of simulations. To predict mixture adsorption isotherms, we tested the performance of ideal adsorption solution theory (IAST), however, significant deviations were observed between adsorption loadings from IAST and NpT-GEMC simulations. This could be attributed to the complex structure as well as difference in adsorption behavior of NH₃, N₂, and H₂. The effect on NH₃ selectivity with varying NH₃/N₂/H₂ gas mixture composition is an important aspect of the ongoing work, allowing extension to process-scale modeling and optimization.

[1] Duan, X.; Kim, D.; Narasimharao, K.; Al-Thabaiti, S.; Tsapatsis, M. High-Performance

Ammonia-Selective MFI Nanosheet Membranes. Chemical Communications 2021, 57, 580–

582.

[2] Patel, R.; Castro, J.; Tsapatsis, M.; Siepmann, J. I. Molecular Simulations Probing the Adsorption

and Diffusion of Ammonia, Nitrogen, Hydrogen, and Their Mixtures in Bulk MFI

Zeolite and MFI Nanosheets at High Temperature and Pressure. Journal of Chemical and

Engineering Data 2022, 67, 1779–1791.