Controlling molten carbonate distribution in dual-phase molten salt-ceramic membranes to increase carbon dioxide permeation rates

Maria Kazakli, Greg A. Mutch, Georgios Triantafyllou, Ana Gouveia Gil, Tao Li, Bo Wang, Josh J. Bailey, Dan J.L. Brett, Paul R. Shearing, Kang Li, Ian Metcalfe*

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

2 Citations (Scopus)
22 Downloads (Pure)


Dual-phase molten salt-ceramic membranes show high permselectivity for CO2 when molten carbonate is supported in a porous oxygen-ion and/or electron conductor. In this arrangement, the support likely contributes to permeation. Thus, if one is to understand and ultimately design membranes, it is also important to perform experiments with an inert support where permeation relies upon the molten carbonate properties alone. Here, a nominally inert material (Al2O3) was used in order to restrict permeation to molten carbonate. Model Al2O3 dual-phase membranes were fabricated using laser drilling to provide an order of magnitude difference in molten salt-gas interfacial area between feed and permeate sides. Molten carbonate thickness in the model membranes was also varied, independent of the molten salt-gas interfacial area. For all thicknesses studied, CO2 permeation rates showed a significant temperature dependence from 500 to 750 °C, suggesting an activated process was rate-limiting, likely a permeate-side molten salt-gas interfacial process, i.e. desorption of CO2. We applied these findings in asymmetric hollow-fibre supports, a geometry with inherent modularity and scalability, by developing a new carbonate infiltration method to control molten carbonate distribution within the hollow fibre. Compared to a conventionally prepared dual-phase hollow-fibre membrane with an uncontrolled distribution of carbonates, permeation rates were increased by up to 4 times when the molten salt was confined to the packed-pore network, i.e. without infiltrating the hollow-fibre micro-channels. X-ray micro-CT investigations supported the idea that the resulting increase in interfacial area for desorption of CO2 was the key structural difference contributing to increased permeation rates. For CO2 separation, where large volumes of gas must be processed, such increases in permeation rates will reduce the demand for membrane materials, although one must note the higher permeation rates achievable with oxygen-ion and/or electron conducting supports.

Original languageEnglish
Article number118640
Number of pages9
JournalJournal of Membrane Science
Early online date21 Aug 2020
Publication statusPublished - 01 Jan 2021
Externally publishedYes

Bibliographical note

Funding Information:
To date, dual-phase membranes have made use of supports in pellet [2,10], hollow-fibre [ 14?16], and tubular geometries [17,18]. Molten carbonate thickness has been controlled in pellet and tubular geometries by employing asymmetric structures, where two solid phases with disparate molten carbonate wettabilities are employed. One solid is a highly-wettable thin layer for supporting molten carbonate and the other is a thick non-wettable layer for mechanical strength [6, 17?19]. In these examples, fluxes were higher than those of comparable thicker membranes made of the highly-wettable material. Similar effects have been observed in tubular dual-phase membranes [18]. Even though a reduction in molten carbonate thickness leads to improved CO2 fluxes, it has been noted that it is difficult to precisely define molten carbonate thickness within a porous support. Thickness is also of limited utility as it is the path length of the permeant that is important, which is difficult to determine in sinuous networks of pores (more common in membranes than e.g. linear pathways). Furthermore, such poorly-defined support geometries will likely lead to uncontrolled differences in interfacial areas or triple-phase boundary lengths, providing additional discrepancies when comparing fluxes. Often, precise interfacial areas are unknown and cannot be used directly. Therefore, it is typical to use a membrane surface area which approximates the dual-phase membrane structure as a smooth surface. However, this surface has different interfaces within it, and often characterisation is lacking. In the absence of precise measurements of interfacial areas, this leads to comparisons of membrane performance that are potentially misleading.Dense Al2O3 cylindrical crucibles (1.4 cm H, 2.55 cm OD, 2.2 cm ID) (Almath), with a base of 440 ?m thickness, were used as the supports for the preparation of model dual-phase crucible membranes. The crucibles were laser-drilled by Laser Micromachining Limited, from the external base of the crucible towards the internal surface forming truncated cone-shaped pores within a central 11 mm diameter. The larger diameter of the truncated cone was on the external surface of the crucibles. Optical microscopy was used to determine the geometrical properties of the drilled surface of the Al2O3 crucibles.The infiltrated Al2O3 hollow fibres used for gas-tightness measurements at room temperature were heated to 600 ?C at 1 ?C min?1 before dwelling for 1 h in order to infiltrate the carbonate mixture into the porous support. The samples were then cooled down to room temperature, using the same rate, and the infiltrated hollow fibres were immediately stored in a desiccator (the carbonate mixture is highly hygroscopic). The loading of carbonates was determined by comparing the weight of the hollow fibres before and after the infiltration step. The DPHFMs used for CO2 permeation measurements were infiltrated in-situ during each experiment (Section 2.5). The microstructure of the prepared DPHFMs were analysed by scanning electron microscopy (SEM), energy dispersive X-ray spectroscopy (EDX) and X-ray micro-CT (detailed in Supporting Data).The authors wish to thank Dr Oliver B. Camus at Bath University for conducting MIP measurements. The research leading to these results has received funding from the European Research Council under the European's Union Seventh Framework Programme (FP/2007-2013) / ERC Grant Agreement Number 320725 and from the Engineering & Physical Sciences Research Council (EPSRC) via grants EP/M01486X/1, EP/P007767/1 and EP/P009050/1. GAM was supported by the Royal Academy of Engineering under the Research Fellowship scheme and would like to thank the EPSRC for a Doctoral Prize Fellowship (EP/M50791X/1) and Newcastle University for support via a Newcastle University Academic Track (NUAcT) Fellowship. X-Ray access was supported by UCL and EPSRC under EP/N032888/1. PRS acknowledges The Royal Academy of Engineering (CiET1718/59). JJB, DJLB, PRS acknowledge The Faraday Institution Nextrode Programme (EP/S003053/1, FIRG015). Data supporting this publication is available under a Creative Commons Attribution 4.0 International license, see DOI: 10.25405/data.ncl.9609200.

Publisher Copyright:
© 2020 The Authors

Copyright 2020 Elsevier B.V., All rights reserved.


  • Carbon dioxide separation
  • Dual-phase molten salt-ceramic membrane
  • Hollow-fibre membrane
  • Molten carbonate

ASJC Scopus subject areas

  • Biochemistry
  • Materials Science(all)
  • Physical and Theoretical Chemistry
  • Filtration and Separation


Dive into the research topics of 'Controlling molten carbonate distribution in dual-phase molten salt-ceramic membranes to increase carbon dioxide permeation rates'. Together they form a unique fingerprint.

Cite this