> 75 %of a planar stack's weight is
interconnects and seals
Escaping the flat stack
Planar and tubular stacks give effective electrode areas below 50 mm2 g−1 and 0.12 mm2 mm−3, and specific power below 0.2 W g−1. The metal also oxidizes and expands differently from the ceramic. Earlier 3D-printed designs, corrugated or multi-tubular, still needed thick 270 µm electrolytes and stayed close to the 2D layout.
≈ 140 µmprinted 8YSZ wall,
half earlier printed electrolytes
One wall, two labyrinths
The gyroid is a triply periodic minimal surface, sin x cos y + sin y cos z + sin z cos x = 0, with zero mean curvature everywhere. Coating one side with the fuel electrode and the other with the oxygen electrode turns the whole wall into an active cell. A sealed outer shell closes the monolith, so no metallic interconnect or stack seal is needed. The sintered wall reaches ≈ 99 % of nominal density, with ionic conductivity matching commercial 8YSZ and no penetrating defects in µCT.
> 1 W g−1and > 3 W cm−3,
fuel cell mode
Power per gram and per volume
The gyroid gives six times the mass-specific and four times the volume-specific electrode area of conventional stacks. Compared with commercial planar SOFC stacks, specific power and volumetric power density rise by factors of about 7 and 9, past the 1 W g−1 threshold set for aerospace. The cells reach over 450 mW cm−2 at 900 °C, and the open-circuit voltage is 1.06 V at 800 °C.
≈ 7 × 10−4Nm3 H2 h−1 g−1,
electrolysis mode
Electrolysis, and Mars
In electrolysis the monolith produces about 7 × 10−4 Nm3 h−1 g−1 and 2 × 10−3 Nm3 h−1 cm−3, nearly ten times planar SOEC stacks. Planned oxygen production for human activity on Mars relies on planar SOEC stacks of 6,400 kg and 6,600 l. The paper estimates that integrated gyroid monoliths could deliver the same in the 800 kg and 700 l range.
1.7 % kh−1SOFC degradation after
conditioning, 850 °C
Reversible and thermally robust
Monoliths ran for hundreds of hours in both modes; the SOEC degraded at 3.3 % kh−1 after its first 25 h at 800 °C. One cell switched between fuel cell and electrolysis every 10 h, another went through six 800–900 °C cycles at up to 5 °C min−1, both without failure. No delamination appeared between layers and no nickel coarsening was detected. An all-ceramic body has no metal-to-ceramic expansion mismatch.
≈ 5×the critical stress, withstood
with a 440 µm outer shell
Room to go thinner
Compression tests and finite element models show that the outer shell, not the active wall, can carry the mechanical load, and that stress spreads evenly over the gyroid with no single critical region. Thicker shells let the electrolyte drop below 100 µm, lowering ohmic loss and operating temperature, which would also open the way to cheaper silver or nickel current collectors.