This is not a material selection problem — it is a functional architecture problem
- The industry has been searching for THE material that simultaneously provides compliance, H₂ barrier, thermal cycling resilience, and steam stability at 300°C.
- No such material exists because the physics requirements are contradictory within a single phase.
- But if you separate compliance and barrier into different layers — as automotive MLS gaskets separate spring recovery from surface conformity, and as biological nacre separates structural rigidity from crack arrest — each layer's material requirements become dramatically easier to meet.
- The vermiculite doesn't need to block hydrogen.
- The copper foil doesn't need to be compliant.
- The spring element doesn't need to resist steam.
- Each material is proven in its source domain; only the combination is novel.
- The functional separation paradigm (compliance layer + barrier layer) has been demonstrated in adjacent applications; the integration for 300°C H₂ service requires validation, not invention.
If you prioritize speed and confidence, start with the MLS + Cu foil architecture (concept sol-primary) — every element is proven, and the first prototype test is 4–8 weeks away. If you want to explore whether a simpler solution exists first, run the Thermiculite 866 H₂ test and NiP permeation measurement in parallel for $20–30k in 2–4 weeks — either could eliminate the need for MLS tooling entirely.
MLS-Embossed Steel + Cu Foil Barrier + Vermiculite Body
Three-layer architecture using automotive stress concentration geometry to achieve 50+ MPa local contact from 5 MPa average, with copper foil providing a 200× margin H₂ barrier — requires prototype tooling and integration testing
Iron-Phosphate Glass Self-Healing Viscous Barrier
A phosphate glass with Tg ~270°C that transitions from rigid gasket to viscous self-healing barrier at operating temperature — eliminates all time-dependent degradation but requires composition discovery
If this were my project, I'd start Monday morning with three parallel actions that cost a total of $20–30k and take 2–4 weeks. First, I'd call Flexitallic and order Thermiculite 866 with copper foil facing in my cell geometry. This is a $200–500 purchase that could solve the entire problem. Even if it doesn't meet spec at 5 MPa (which I expect), the measured H₂ leak rate tells me exactly how much improvement I need from MLS stress concentration or NiP smoothing. It's the cheapest data point in the portfolio and it calibrates everything else. Second, I'd send 10 polished 316L coupons to a commercial NiP plating shop (MacDermid Enthone or Atotech) and ask for 20 µm high-phosphorus (10–12 wt% P) electroless nickel. Total cost: $500–1000. Then I'd measure H₂ permeation through the coated coupons at 300°C. This single measurement — which nobody in the world has made — determines whether the barrier can live on the interconnect surface rather than in the gasket. If NiP permeability is favorable, the entire gasket design simplifies to 'buy a commodity vermiculite sheet.' Third, I'd build a spreadsheet model of the stack mass balance: N cells, O₂ flow rate, per-cell H₂ crossover, recombiner conversion. This takes a day and costs nothing. It tells me how much the per-cell spec can be relaxed with a downstream recombiner — even 2–3× relaxation makes the Thermiculite test more likely to succeed.
- If Thermiculite meets spec: problem solved, ship it.
- If NiP permeability is favorable: NiP coat the interconnects, use bare vermiculite, done.
- If both fall short but are within 5–10× of target: commission the MLS die ($10–20k, 4 weeks) and build the three-layer prototype.
- If everything is >50× off: we have a harder problem than expected and the self-healing glass becomes the primary path.
I would NOT start with the MLS die — it's the most expensive near-term action and it might be unnecessary. I would NOT start with the phosphate glass — it's a 6–12 month program that should run in parallel, not as the primary path. And I would absolutely file a provisional patent on the MLS + Cu foil + vermiculite architecture as soon as I have positive prototype data — the 200-400°C electrolyzer seal patent landscape is nearly empty, and the first mover with validated data has a real competitive advantage.