(32b) Immobilised Solvent Systems: Evaluating the Potential of High Surface Area Membrane/Solvent Hybrid Sorbent Materials

Authors: 
Moore, T. - Presenter, University of Melbourne
Webley, P. A., The University of Melbourne
Stevens, G. W., The University of Melbourne
Mumford, K. A., The University of Melbourne
Microencapsulation of liquid solvents (MECS) [1] and Solvent Impregnated Polymers (SIPs) [2] are two recently developed hybrid materials that aim to exploit the selectivity, versatility and thermodynamic properties of liquid solvents for gas separation, while simultaneously increasing the surface area available for mass transfer. In each technology, liquid solvents are immobilised inside high surface area solids. In MECS, the solvent is trapped inside a capsule a few hundred microns in diameter. The shell is composed of a thin, permeable polymer membrane, and the liquid solvent is held in the capsule core. In SIPs, the solvent is trapped inside a polymer matrix, forming a gel that may be shaped into thin sheets or particles. Immobilised solvents have 1-2 orders of magnitude greater surface area than the liquid in a traditional absorption column, and they may in principle enable thermodynamically favourable solvents with slow absorption kinetics or high viscosity, volatility or corrosivity to be used in a practical way.

In practice, however, the technology faces several obstacles. The most obvious is the need to upscale the manufacture of these materials. MECS particles are currently manufactured inside microfluidic devices at a rate on the order of 1 g/hr, and the production of tonnes of particles for industrial-scale gas separation will be a major challenge. SIPs, which were specifically developed to overcome this barrier, are manufactured via a scalable one-pot method which bypasses the need for any microfluidic chips. Instead, a fine solvent-in-polymer emulsion is first created, and after it is formed into the desired shape the polymer is crosslinked to create a solid gel. Particles may be manufactured by dispersing the single emulsion as a double emulsion prior to crosslinking, while thin sheets may be manufactured by spreading the emulsion on a supporting surface. Though SIPs may prove simpler to manufacture at scale than MECS particles, selection of appropriate polymers and surfactants remains a major challenge. Initial efforts focused on immobilising concentrated K2CO3 solutions inside PDMS for carbon capture applications. Several PDMS-based surfactants were identified which were able to stabilise 50/50wt% K2CO3(aq)-in-PDMS emulsions, and SIP particles with internal droplet size on the order of 20-50 microns were created. However, these emulsions were quite viscous, and when dispersed as a double emulsion the mean particle size was approximately 2mm, substantially larger than MECS particles. Furthermore, the relatively large internal droplet size means very fine particles or layers cannot be produced without significant solvent leakage. One possible means of overcoming these challenges is to use ethanol to reduce the internal surface energy. By adding 1-5wt% ethanol to the aqueous phase, a low-viscosity emulsion may be created with internal droplet size on the order of 1-5 microns. These emulsions are very stable, however when cross-linked the solid gels were sticky and not as rigid as other SIP materials. While SIPs may be a feasible pathway to large-scale production of immobilised solvent systems, further work is required to identify ideal polymers, emulsifers and manufacturing conditions.

A second question relates to the need to rigorously quantify the enhancement in absorption rate that solvent immobilisation can provide, as compared to traditional absorption operations. To date, estimates for this kinetic enhancement have been based upon the increase in surface area alone, and have not accounted for differences in the hydrodynamics and transport mechanisms, shell resistance, or absorber voidages. Moore et al [3] used surface renewal theories and well-established mass transfer correlations to rigorously compare absorption rates inside MECS with absorption rates inside traditional packed columns. General expressions for the increase in overall absorption rate were derived, and both chemical and physical solvents were considered. Two model systems - chemical absorption of CO2 into 30wt% K2CO3, and physical absorption of CO2 into Selexol - were investigated in depth.

It was found that, for most practical chemical solvents, the suppression of liquid mixing inside immobilised solvents will not significantly reduce the gas flux, relative to a traditional absorption column. For physical solvents, suppression of liquid mixing may reduce the flux, as mass transfer is more reliant upon convective mixing to the bulk, however this effect is counterbalanced with an increase in gas flux associated with the increase in concentration gradients that occurs as spatial scales are reduced. Different effects will dominate for different systems, however for CO2 absorption into Selexol, the gas flux into 100 micron MECS particles was roughly 4 times larger than in a traditional packed column, leading to more than a 200-fold increase in the overall gas absorption rate.

The increase in surface area that microencapsulation can provide was also calculated, using reliable correlations for surface area in absorption columns. When the differences in absorber voidage and the very high surface area of structured packings such as Mellapak 500Y were accounted for, the increase in surface area was 3-10 times smaller than previous publications had suggested, though it was still more than an order of magnitude for the finest particles. Overall, MECS may be expected to increase the gas absorption rate per unit volume of absorber by up to an order of magnitude for chemical solvent systems, and over two orders of magnitude for physical solvent systems.

This analysis must be modified for SIP particles, as the mean gas flux is more dependent upon the specific size of the particles. As the solvent is not free to move inside the material, a dead-zone forms near the surface, similar to that inside a shrinking-core model. In order to better understand this behaviour, a volume-averaged model for gas absorption inside SIPs materials was developed and experimentally verified. If SIPs are too large, the diffusion resistance through the outer core dominates, and mass transfer is poor. On the other hand, for very concentrated solvents (such as 40wt% K2CO3 for CO2 capture) the permeability of the gas through the polymer matrix may be 1-2 orders of magnitude larger than the permeability inside the liquid, and for appropriately sized particles this can lead to a 2-4 fold increase in the mean gas flux during absorption.

Immobilised solvent systems such as MECS and SIPs are a promising means of increasing gas uptake rates, and they may find application in a wide range of gas separations. Plausible pathways for upscaling their manufacture are available, and the mass transfer enhancement such materials may provide are substantial. However the technology is still in its infancy, and further work is required to identify the processes and solvents to which it is best suited, to refine manufacturing methods and material selection, and to quantify the thermodynamic enhancement that the use of otherwise impractical solvents may provide.

References

[1] Vericella, J. J. et al. (2015) Encapsulated liquid sorbents for carbon dioxide capture. Nat. Commun. 6:6124 doi: 10.1038/ncomms7124

[2] Moore, T. et al. (2017) Solvent Impregnated Polymers for Carbon Capture. Poster Session presented at Australia CCS Conference, Melbourne, Australia

[3] Moore, T. et al. (2018) Enhancement in Specific Absorption Rate by Solvent Microencapsulation. Submitted for publication.