The separator is the bottleneck, not the catalyst — and DAC's batch nature means the separator may be unnecessary
- Quantitative voltage budget decomposition reveals the Zirfon diaphragm consumes 100–800 mV at ambient temperature — far more than HOR kinetics (20–60 mV on Pt).
- Replacing or eliminating the separator is worth 2–40× more voltage savings than any catalyst improvement.
- Furthermore, the assumption that acid and base must be generated simultaneously in separate compartments is inherited from continuous industrial electrolysis.
- DAC is inherently batch: CO₂ absorption and release happen at different times.
- A single electrolyte stream could be alternately acidified and basified, eliminating the separator entirely.
- The voltage budget analysis reveals the separator consumes 200–800 mV at ambient temperature — far more than previously appreciated.
- Commercially available AEMs can recover most of this, and multiple complementary approaches (pulsing, catalyst selectivity, feedback control) provide layered insurance.
If you prioritize speed and low risk, start with the AEM separator swap (4–8 weeks to proof-of-concept). If you're building a 5-year R&D roadmap and want to own the paradigm shift, fund the PBA intercalation work in parallel — even partial success would be a landmark result.
AEM Separator with Zero-Gap Architecture
Replace Zirfon with a thin AEM and adopt zero-gap cell design to cut separator resistance 10–100×, dropping cell voltage to 0.6–0.9V — but AEM behavior in K₂CO₃ electrolyte is uncharacterized.
Prussian Blue Analog Intercalation pH-Swing Cell
Replace the entire Faradaic H₂-loop with PBA intercalation electrodes operating at 0.0–0.9V, eliminating OER thermodynamically — but 200 mA/cm² hasn't been demonstrated yet.
- If this were my project, I'd split the work into two parallel tracks running simultaneously. **Track A (this month, <$30K):** Install the dissolved O₂ sensor on the acid stream outlet — this takes a day and gives you real-time visibility into OER losses you've never had before.
- While waiting for AEM samples, swap the DC power supply for a GaN pulse generator and sweep pulse parameters.
- These two changes together cost less than a single conference trip and could improve throughput by 10–20%.
- They also generate diagnostic data that informs every subsequent decision.
- In parallel, email Ionomr and Versogen requesting AEM samples for carbonate electrolyte testing.
- And — this is important — run the Ga-In permeation test.
- It's $5–10K and 2 weeks.
- Either you discover a potentially transformative separator technology, or you definitively eliminate it.
- The information value per dollar is the highest in the portfolio. **Track B (months 2–6, $100–200K):** Once AEM samples arrive, measure ASR in your actual electrolyte.
- If the numbers look good (ASR < 0.2 Ω·cm²), start the zero-gap cell redesign — this is your primary development path.
- Simultaneously, begin the ultra-low-loading anode optimization (concept-2) because you need to reduce PGM cost regardless of which separator you use.
- If you have the bandwidth, start the Bi/Pt RDE experiment to characterize the dual-selectivity approach — it's a 4–6 week experiment that either opens a powerful new design degree of freedom or closes that door.
- The PBA intercalation work (Track C) is a 6–18 month R&D investment that I'd fund through a research partnership rather than in-house.
- Contact Presser at INM or Pasta at Oxford.
- Even if 200 mA/cm² isn't achieved, demonstrating PBA pH-swing for DAC at any current density would be a landmark publication and a strong IP position.
- This is the bet on the future — the AEM separator is the bet on the present.
- One thing I'd watch carefully: before committing to the AEM path, have a 10-minute conversation with your team about whether they've already tried AEM separators.
- If they have and found carbonate poisoning to be severe, the entire strategy shifts toward catalyst selectivity (DSA/Bi approach) or paradigm shift (PBA intercalation).
- Don't assume the most obvious solution hasn't been tried.