Start with the number, because the number is the whole story: after 1,500 charge–discharge cycles, a Paul Scherrer Institute button cell still held about 75 percent of its capacity — among the best published for a lithium-metal all-solid-state cell, reported in Advanced Science this year. Everything else — safety, range, the grid-storage dream — hangs off whether that curve stays flat. It nearly does.

←TODAY: A PSI lab cell hits 75% retention at 1,500 cycles using an ~80°C process. →3012: Buildings store their own noon sun in wall-thin solid cells, no fire margin, no liquid to leak. Fulcrum: The advance is not more energy — it is a densification you can run at coffee temperature, which is what makes it pourable into a supply chain.

Why 80 degrees is the frontier, not the range

The material is argyrodite Li₆PS₅Cl, a sulphide solid electrolyte that conducts lithium ions fast but fails in one stubborn way: densify it wrong and you leave voids. Into those voids grow dendrites — needle-like lithium filaments that spear the electrolyte and short the cell. Both classic fixes are bad bargains: cold-press at room temperature and the structure stays porous; sinter above 400°C and you compact it but risk breaking the electrolyte down chemically.

So Mario El Kazzi’s group split the difference — moderate pressure at roughly 80°C, which closed the cavities without touching the chemistry — then added a 65-nanometre lithium-fluoride film, vacuum-evaporated onto the lithium as a passivation layer. That LiF skin blocks the interfacial decomposition and stands as a physical wall against dendrites. Two problems, one coating and one temperature.

Read the error bar before the headline

Awe held under arithmetic: 75% at 1,500 cycles is excellent, but it is a button cell in a lab, not a stack. The honest caveat sits in adjacent work — researchers at the Institut Laue-Langevin used operando neutron diffraction to watch lithium move inside a working solid cell, and found the flow chaotic, not the smooth front the textbooks draw. Uneven lithium is exactly where a demo and a product part ways.

Where it lands on your desk

This is not a phone-battery story for AEC readers — it is a stationary storage one. A cell with no flammable electrolyte changes the fire compartmentation, the setback, and the insurance line for a building-scale battery room; the SIA and the local Brandschutz officer care more about “no liquid, no thermal-runaway path” than about ten extra Wh/kg. When cells like this reach the wall, the battery stops being a hazardous appliance you quarantine and becomes closer to a structural layer.

Atelier: Offices specifying on-site storage today are still sizing rooms around liquid-electrolyte fire ratings that solid-state will eventually relax — but not yet. The Monday move: add one column to your energy-storage spec sheet, “cycle-life at stated depth-of-discharge, independently replicated,” and refuse to let a supplier answer it with a single lab figure.

Hack: Convert the press-release headline into the number that actually decides a spec — the mean capacity fade per cycle, and the cycle where you cross an 80% floor. A linear model is crude (real fade accelerates), but it tells you fast whether “1,500 cycles” is generous or thin for a building that charges daily for twenty years.

cycles, cap = 1500, 0.75
fade = (1 - cap) / cycles          # mean loss per cycle
to_80 = (1 - 0.80) / fade           # cycles to the 80% floor
print(round(fade*1e6, 1), "ppm/cycle", int(to_80), "cycles to 80%")

Run it: ~167 ppm per cycle, and the 80% floor arrives near cycle 1,200 — under three and a half years of once-daily cycling. That is the arithmetic a datasheet hides. Pull the Advanced Science paper (Zhang et al., 2026), find the depth-of-discharge behind that 1,500, and bring the real fade number — not the headline — to your next storage-sizing meeting.

Source: psi.ch