(699c) NH3 Separation Via ZnCl2-Immobilized Molten Salt (IMS) Membrane – Experimental and Modeling

The utilization of membrane-based separation demonstrates significant potential to mitigate energy consumption and emissions in crucial industrial processes, such as the Haber-Bosch (H-B) process [1]. The industrial H-B process requires harsh operating conditions for its operation (150 – 250 bar and 300 – 450 °C) and emits CO₂ (1.5 – 1.6 ) [2]. Being constrained by kinetics and thermodynamics, the single-pass conversion revolves around 15% [3]. Conventionally, a chain of heat exchangers and a final refrigeration stage are usually adopted to recover NH₃ from unreacted N₂ and H₂. It has been reported that adopting alternative techniques for NH₃ separation could be one strategy to enable low-carbon NH₃ production [1]. A recent study on ZnCl₂ immobilized molten salt (IMS) membranes has revealed that they can be potentially used at high temperatures for NH₃ separation with exceptional separation characteristics [4].

This study systematically explores the performance of ZnCl₂ immobilized molten salt (IMS) membranes, both theoretically and experimentally, at high pressures for separating NH₃ from a mixture of N₂ and H₂. Firstly, the ZnCl₂ IMS membrane's separation characteristics were experimentally determined with a 1 µm pore-sized wire mesh support when exposed to pure and mixed gases at 300 °C and 100 - 350 kPa. The membranes were synthesized in situ via the direct deposition method and were characterized by a 50 µm thickness. When exposed to a single gas, the membranes exhibit NH₃ permeance of ~211 GPU, with both NH₃/N₂ and NH₃/H₂ ideal selectivities > 10⁷ at 100 kPa. In the case of binary mixtures, NH₃ permeance within the range of 6000 to 7000 GPU was attained at a feed NH₃ partial pressure of ~3 kPa. Secondly, the theoretical part of this study employed a mathematical model initially introduced by Xu et al. [5] to examine the facilitated transport mechanisms of NH₃ across ZnCl₂ IMS membranes. The mathematical model was fit to experimentally measured NH₃ fluxes as a function of NH₃ partial pressure (~10 to ~100 kPa) and membrane thicknesses (50 µm, 80 µm [4], and 200 µm [6]) to deduce the kinetic and thermodynamic parameters related to the permeation of NH₃ through the membrane. The mean absolute percentage error (MAPE) between model and experimental data was less than 5%.

In general, even though ZnCl₂ salt is corrosive to stainless steel, the membrane was stable for at least 640 h under different feed mixtures with no significant performance loss. In addition, the membrane exhibited sustained mechanical stability and selectivity up to 350 kPa, underscoring its potential suitability for high-pressure systems.

References

[1] A.I. Osman, Z. Chen, A.M. Elgarahy, M. Farghali, I.M.A. Mohamed, A.K. Priya, H.B. Hawash, P.-S. Yap, Advanced Energy and Sustainability Research (2024) 2400011.

[2] C. Smith, A.K. Hill, L. Torrente-Murciano, Energy Environ Sci 13 (2020) 331–344.

[3] H. Liu, Ammonia Synthesis Catalysts (2013).

[4] M. Adejumo, L. Oleksy, S. Liguori, Chemical Engineering Journal 479 (2024) 147434.

[5] H. Xu, S.G. Pate, C.P. O’Brien, Chemical Engineering Journal 460 (2023) 141728.

[6] Daniel V. Laciak, Guido P. Pez, Peter M. Burban, Journal of Membrane Science, (1992), Volume 65, Issues 1–2, Pages 31-38.