Path 6: Membrane Separation (Selective CO&sb2; Filters)

Idea: Membranes – thin films or hollow fibers with selective permeability – can act as filters that let CO&sb2; pass through faster than other gases. By feeding a gas stream (air or flue gas) under pressure on one side, one can obtain an enriched CO&sb2; stream on the other side as CO&sb2; preferentially diffuses through the membrane material. Membranes can be polymer-based (special plastics) or inorganic (ceramic, facilitated transport membranes, etc.), and often work on either solution-diffusion principles (CO&sb2; dissolves in the membrane and diffuses faster than N&sb2; or O&sb2;) or molecular sieve principles (CO&sb2; is smaller and more condensable, so it permeates through size-selective pores). The concept for DAC would be to use membranes perhaps in multiple stages to gradually strip CO&sb2; from air, or to use a combination of membrane and sorption (hybrid processes).

Rationale: Membranes are widely used in gas separations (e.g., for enriching oxygen or removing CO&sb2; from natural gas) – they have no moving parts and can be energy-efficient when the selectivity is high. For point sources, membrane CO&sb2; capture is already commercial in some cases (natural gas processing often uses polymer membranes to remove CO&sb2;). The rationale for air capture is more challenging because the driving force (partial pressure difference of CO&sb2;) is very low; however, if one can create a strong vacuum or couple the membrane with a vacuum pump, it could continuously extract CO&sb2;. Evidence that it might work: researchers have made ultrathin membrane materials with very high CO&sb2;/N&sb2; selectivity (for instance, certain polyethers or facilitated transport membranes that have carriers binding CO&sb2;). Additionally, membranes excel in modularity – you can scale by adding more area. There have been proposals to use membrane contactors in conjunction with solvents (membrane absorption) to enhance contact efficiency. Partial successes include experimental membranes that achieved >1000 GPU (gas permeation unit) permeability for CO&sb2; and high selectivities over nitrogen, which would significantly reduce the membrane area needed. Another promising sign: ionic liquid membranes or hybrid membranes with embedded CO&sb2;-philic carriers have shown the ability to “facilitate” CO&sb2; transport by reversible reaction, effectively shuttling CO&sb2; across faster than pure diffusion. Failures or difficulties: standard polymer membranes suffer from a trade-off (the Robeson upper bound) between permeability and selectivity – extremely selective membranes tend to be slow. But novel materials like polymers of intrinsic microporosity (PIMs) or carbon molecular sieve membranes are pushing that boundary. For direct air, membranes alone might require large surface areas and vacuum energy, but integrating them smartly (e.g., wind-driven air flow or using waste pressure) might make them viable. Membranes also avoid phase changes (no heating/cooling of sorbents), potentially offering a continuous, energy-light process if a suitable membrane is found.

Prerequisite Themes: Polymer science (knowing how gases permeate through polymers, solubility and diffusivity), basic fluid mechanics (to design modules like spiral-wound or hollow fiber modules and to handle flow and pressure drop), and knowledge of permeation models (Fick’s law, etc.). Also, familiarity with vacuum systems and compression (since membrane capture might use vacuum pumps or compressors to create driving force). If tackling facilitated membranes: some chemistry of carrier agents (like amine carriers or carbonate forming carriers in the membrane) is needed.

Dependencies: This path can complement others – for example, a membrane could enrich CO&sb2; from 0.04% to a few percent, and then another method (like Path 1 or 4) could do the final capture; thus it might work in tandem rather than alone. It draws on advanced materials from Path 5 if MOFs or porous organic cages are used within membranes (mixed-matrix membranes). Also, any improvements in Path 4 regarding how materials discriminate CO&sb2; could inspire membrane fixed-carrier designs. Path 7 (electrochemical) even overlaps: “electrochemical membranes” or ion pumps that separate CO&sb2; (like using a bipolar membrane to create a pH swing that draws in CO&sb2;) could be seen as a hybrid of membrane and electrochemistry.

Signs of Progress: An eye-catching milestone would be a CO&sb2;-selective membrane module tested on air that achieves a noticeable CO&sb2; enrichment (e.g. output stream with >1% CO&sb2;) without exorbitant energy cost. Metrics to watch: CO&sb2; permeability (how easily CO&sb2; goes through) and CO&sb2;/N&sb2; selectivity – if new materials exceed the Robeson limit significantly, that’s important. For example, a membrane with a CO&sb2;/N&sb2; selectivity of 200 and CO&sb2; permeability of 5000 Barrer would be far beyond conventional polymers and could make air capture more feasible. Also, demonstration of long-term stability: membranes often suffer plasticization or fouling (especially if exposed to contaminants or humidity). If a membrane can run with real air for months without losing performance, that’s a great sign. Integrating renewable energy: say a solar-powered vacuum pump running a membrane DAC unit – if someone shows the energy per ton CO&sb2; captured is in a competitive range using such a setup, it validates the concept. Another sign: successful fabrication of membranes at scale (meters squared) with consistent performance, because often lab membranes are tiny films. If companies report scaling up membranes specifically for CO&sb2; (some startups are exploring this), that indicates the path is moving forward.

Base Camp 6.1: Polymer Science for Gas Separation Membranes

Scope: Understand the types of polymer membranes and the solution-diffusion mechanism. Key terms: permeability (P) measured in Barrer, selectivity (α) = ratio of permeabilities of gases (CO&sb2; vs N&sb2;). Learn that permeability = solubility × diffusivity (P = S × D). Polymers can be glassy (rigid chains, like polyimides) offering higher selectivity but lower permeability, or rubbery (flexible chains, like PDMS silicone) offering high permeability but low selectivity. The trade-off is captured by the Robeson upper bound curve. Students should get why small, condensable molecules (CO&sb2;) dissolve more and often diffuse not too slowly, giving some selectivity over larger like N&sb2; or O&sb2;. Cover examples: cellulose acetate (an early membrane for CO&sb2; removal, moderate P, α_CO2/N2 ~ 30), polyimide Matrimid, polysulfone, PEBAX copolymers, etc. Introduce PIMs (Polymers of Intrinsic Microporosity) – high free volume, high permeability polymers due to contorted structures. Also mention aging: glassy polymers’ permeability drops over time as polymer chains relax (important for stability). The aim is to know common membrane materials and their performance envelope. (Stepping stones: define units and do example calc: if CO&sb2; permeability is 100 Barrer and membrane thickness 1 micron, how much CO&sb2; flux at a given pressure difference; interpret a Robeson plot – see how new materials (like PIM-1) push boundary; basic notion of polymer chain packing and how bulky substituents create gaps aiding gas diffusion; mention crosslinking or mixed matrix as ways to improve properties).

Stepping Stones: Define units and do example calc: if CO&sb2; permeability is 100 Barrer and membrane thickness 1 micron, how much CO&sb2; flux at a given pressure difference; interpret a Robeson plot – see how new materials (like PIM-1) push boundary; basic notion of polymer chain packing and how bulky substituents create gaps aiding gas diffusion; mention crosslinking or mixed matrix as ways to improve properties.

• B. D. Freeman, “Basis of Permeability/Selectivity Tradeoff Relations in Polymeric Gas Separation Membranes,” Macromolecules, 1999. – Why: This paper outlines the trade-off and gives fundamental reasons. It’s technical but has explanatory parts about polymer structure and gas transport. It basically established the upper bound concept. Good for understanding the core limitations.
• Robeson, “The Upper Bound Revisited,” Journal of Membrane Science, 2008. – Why: A follow-up by Robeson updating the upper bound. It lists many polymer data points; reading it is challenging but one can glean how new classes of polymers improved performance. For an advanced student, it’s a treasure trove linking performance to polymer classes.
• B. T. Low et al., “Natural Gas Treating with Polymeric Membranes: Problems and Solutions,” Polymer, 2009 (sections on polymer membranes). – Why: Though about natural gas, it clearly discusses polymer membrane limitations like plasticization by CO&sb2;, aging, etc. It’s approachable and pragmatic, giving a real-world angle useful for CO&sb2;/air too.

Base Camp 6.2: Advanced Membranes (Mixed-Matrix & Facilitated Transport)

Scope: Explore enhancements to polymer membranes. Mixed-Matrix Membranes (MMMs): embedding porous fillers (zeolites, MOFs, carbon molecular sieves) in polymers to improve selectivity or permeability. For example, adding a zeolite 4A could increase CO&sb2; affinity, or a MOF like ZIF-8 to increase free volume and CO&sb2; diffusion. But issues include filler/polymer compatibility (no voids or blocked pores). Facilitated transport membranes: incorporate carriers (e.g. amines, or fixed carrier like a quaternary ammonium that can reversibly react with CO&sb2;) to facilitate CO&sb2; transport. This can yield very high selectivity by a reaction-diffusion mechanism (CO&sb2; binds, diffuses as complex, releases). These often require water for carrier mobility (e.g. membranes with fixed amines that work in humid conditions, transporting CO&sb2; as bicarbonate). Also mention emerging things like Ionic Liquid membranes (polymers swollen with ionic liquid that has CO&sb2; affinity) or graphene oxide membranes (selective laminar structures). Students should appreciate these strategies aim to break the upper bound by adding new transport mechanisms beyond simple solution-diffusion. Also cover stability: e.g., facilitated transport membranes can be poisoned by contaminants or the carrier can degrade; MMMs can have defects causing losses in selectivity. But some have shown impressive performance in lab (like a facilitated membrane with selectivity 200+ for CO&sb2;/N&sb2;). (Stepping stones: example MMM result: e.g. 20% UiO-66 in Matrimid improved CO&sb2;/N&sb2; selectivity by X; reasoning: MOF pore gates help discriminate; or an amino acid salt in a membrane achieving facilitated CO&sb2; hopping; consider effect of humidity – often needed but then membrane must resist water swelling; mention performance of Pebax polymer with ionic liquid additive as an example in literature).

Stepping Stones: Example MMM result: e.g. 20% UiO-66 in Matrimid improved CO&sb2;/N&sb2; selectivity by X; reasoning: MOF pore gates help discriminate; or an amino acid salt in a membrane achieving facilitated CO&sb2; hopping; consider effect of humidity – often needed but then membrane must resist water swelling; mention performance of Pebax polymer with ionic liquid additive as an example in literature.

• Yongsoo Shin et al., “Mixed Matrix Membranes Incorporating Metal–Organic Framework Nanocrystals for Carbon Dioxide Capture,” Nature, 2017. – Why: A high-profile example showing MMMs with MOF fillers achieving good CO&sb2; separation. It’s well-explained and demonstrates real improvements. It highlights the compatibility issue and how they solved it (surface functionalization of MOF). Good for illustrating MMM potential.
• K. Nagai et al., “Facilitated Transport Membranes for CO&sb2; Separation,” Progress in Polymer Science, 2010. – Why: A thorough review of facilitated transport. It covers carriers like amines, carbonates, tasks to keep carriers mobile, etc. It might be dense, but one can get key principles from the introduction and conclusions. It shows how CO&sb2; flux can surpass usual limits when facilitated by reaction.
• R. Merkel et al., “Thin Film Composite Membranes for Post-Combustion CO&sb2; Capture,” Journal of Membrane Science, 2013. – Why: This paper (from Membrane Technology Research) discusses making thin, high-permeance membranes for flue gas CO&sb2;. It covers materials and also module design, and techno-economics a bit. It’s valuable to connect advanced membrane concept to actual application needs (like needing high permeance to reduce membrane area).

Base Camp 6.3: Membrane Modules and System Design

Scope: Transition from material to membrane units. Cover hollow fiber vs flat sheet modules. Most large-scale gas membranes use hollow fibers (like in natural gas CO&sb2; removal) because you can pack a lot of area. Understand how a membrane unit is arranged: feed on one side, permeate on the other, often multi-stage setups because a single stage only partially removes CO&sb2;. For DAC specifically, membranes might need multiple passes or a sweep gas. Also discuss pressure ratio: performance depends on feed/permeate pressure difference. For CO&sb2; from air, likely we do vacuum on permeate side since feed is ambient. That means large vacuum pumps and careful design to avoid leaks. Understand stage-cut concept: fraction of feed that permeates. And for high purity, multi-stage or recycle configurations often needed. Possibly touch on pressure drop and how designing fiber dimensions is a trade-off between area and pressure drop. Provide a feel for how large membrane area is needed for capturing, say, 1 ton CO&sb2;/day given current membrane permeances (like if a membrane has 1000 GPU for CO&sb2; in a composite ~0.1 micron selective layer, how many m² to get needed flux). Also consider contamination: particulates can foul membranes (so need filters), and long-term durability of polymers under constant airflow with possibly ozone or other pollutants in air (for DAC scenario). And module maintenance: membranes in field may need periodic cleaning or replacement after years. For flue gas, discuss controlling water and plasticization (CO&sb2; can swell polymers if high pressure). The student should be able to outline a membrane capture plant with approximate flows and energy inputs (mostly compression/vacuum costs). Possibly quantify: capturing from 0.04% to maybe ~20% CO&sb2; in permeate might be a realistic goal per stage, then compress that etc. Also mention hybrid processes (membrane + sorbent or cryo) if relevant. (Stepping stones: illustrate a simple 2-stage membrane design: first stage takes air to ~5% CO&sb2; in permeate, second takes that to ~50%, leaving mostly N&sb2; on residue; compute needed vacuum level to get decent transmembrane driving force; quick calc: a 1000 m² membrane with 1000 GPU might capture around order of magnitude 10 kg CO&sb2; per day from air – showing needed scale; mention current industrial uses of membranes for CO&sb2; to emphasize viability (like biogas upgrading, natural gas sweetening)).

Stepping Stones: Illustrate a simple 2-stage membrane design: first stage takes air to ~5% CO&sb2; in permeate, second takes that to ~50%, leaving mostly N&sb2; on residue; compute needed vacuum level to get decent transmembrane driving force; quick calc: a 1000 m² membrane with 1000 GPU might capture around order of magnitude 10 kg CO&sb2; per day from air – showing needed scale; mention current industrial uses of membranes for CO&sb2; to emphasize viability (like biogas upgrading, natural gas sweetening).

• B. Belaissaoui et al., “Air Capture of CO&sb2; by Vacuum Swing Membrane Separation,” Industrial & Engineering Chemistry Research, 2012. – Why: This study specifically analyzes DAC using membranes and vacuum. It gives an idea of process design and energy requirements. Good because it directly addresses using membranes for air capture and thus highlights the enormous flow rates needed and the multi-stage approach.
• Matteo C. Ferrari et al., “Efficient Operation of Membrane Systems for CO&sb2; Capture: Key Factors and Process Integration,” Energy Procedia, 2011. – Why: Although for flue gas, it deals with multi-stage membrane systems and optimization. It’s useful to understand how one optimizes number of stages, inter-stage compression, etc., to minimize energy. The language is moderate, and it ties materials performance to system outcome.
• Baker & Low, “Gas Separation Membrane Materials: A Perspective,” Macromolecules, 2014. – Why: A perspective on where membrane materials and designs are heading for gas sep. It’s conceptual but addresses what is needed for, say, post-combustion capture (higher permeance, etc.). It also indirectly applies to DAC. It’s inspiring and consolidates the notion that materials alone won’t solve if not combined with clever process design (like hollow fiber modules, staging).

Base Camp 6.4: Economics and Integration of Membranes

Scope: Evaluate where membranes fit in CO&sb2; capture economically and as part of an overall system. For point sources, membranes might not achieve high purity in one step, so often combined with another operation (e.g., membrane to bulk-remove CO&sb2; then polish by amine, etc.). For DAC, maybe membranes alone are tough, but perhaps combined with sorbent? Look at cost factors: membrane capital cost largely from membrane area (which relates to permeance) and module packaging; operational cost mainly energy (compression/vacuum). If renewable electricity costs drop, membrane might become more attractive. Provide any known study results: e.g. capturing CO&sb2; from 12% to 90% purity via membranes was more energy intensive than amines beyond certain selectivity limit. Or for DAC, membrane route might be e.g. 200 kWh/ton electrical vs 50 kWh/ton mechanical for fans for sorbent – just hypothetical numbers. Also consider modular advantages: membranes are compact, no hazardous chemicals, can be easily scaled by numbering-up. And they start working immediately (no need to heat/cool cycles). But possibly lower thermodynamic efficiency capturing from low concentration (blowing a lot of air through). A student should be able to list pros/cons: e.g., Membranes are continuous, potentially simpler (no sorbent handling), but likely require substantial energy and high-performance polymers that push limits. Also note that membranes struggle to reach high purity, which is fine if we want to release CO&sb2; concentrated for storage but might need further compression. Summarize current status: membranes widely used for moderate CO&sb2; removal (natgas, biogas), being tested for flue gas (some pilot units exist), not yet for DAC in any serious way. But future improvements could open that door. (Stepping stones: recall energy per mole separation fundamental (not too detail but concept of more energy when separating from dilute source – which penalizes membranes too as it does sorbents, basically DAC is energy heavy no matter what); think of synergy: maybe use waste pressure or integrate membrane with cryogenic or electrochemical steps to improve overall efficiency).

Stepping Stones: Recall energy per mole separation fundamental (not too detail but concept of more energy when separating from dilute source – which penalizes membranes too as it does sorbents, basically DAC is energy heavy no matter what); think of synergy: maybe use waste pressure or integrate membrane with cryogenic or electrochemical steps to improve overall efficiency.

• Merkel et al., “Futuristic Membranes for CO&sb2; Capture,” Journal of Membrane Science, 2010. – Why: Lays out performance targets needed and potential costs for membrane capture of CO&sb2;. Good forward-looking piece with some economics.
• IEA Greenhouse Gas Programme, “Evaluation of Membrane Technology for CO&sb2; Capture from Flue Gas,” 2012 report. – Why: This technical report goes through an economic assessment of membranes in a power plant scenario. Even skimming summary and conclusions informs about cost and feasibility, and improvements needed.
• Article: “Controlling Climate Change by Direct Air Capture: Importance of Membrane Performance,” Membranes, 2021 (if available; otherwise use Belaissaoui from BC6.3 which addresses DAC economics as well).

Full Bibliography

• Baker, R. W. Membrane Technology and Applications, 3rd ed. John Wiley & Sons, 2012.
• Robeson, L. M. “The Upper Bound Revisited.” Journal of Membrane Science, vol. 320, 2008, pp. 390–400.
• Merkel, T. C., et al. “Power Plant Post-Combustion Carbon Dioxide Capture: An Opportunity for Membranes.” Journal of Membrane Science, vol. 359, 2010, pp. 126–139.
• Bernardo, P., E. Drioli, and G. Golemme. “Membrane Gas Separation: A Review/State of the Art.” Industrial & Engineering Chemistry Research, vol. 48, 2009, pp. 4638–4663.
• Sundarraj, P., et al. “An Overview of Mixed Matrix Membranes for CO&sb2; Separation: Recent Progress and Future Directions.” Carbon Capture Science & Technology, vol. 1, 2022, 100014.