Lithium dendrites — the microscopic metal whiskers that grow inside batteries and eventually short-circuit them — have been studied for decades. But until a 2026 paper in Science (Ai, Liu, et al., Rice/NJIT/Georgia Tech), nobody had directly measured their mechanical properties. The reason: each dendrite is about a hundred times thinner than a human hair, and lithium reacts with air on contact. You need cryo-electron microscopy, air-tight sample transfer, and mechanical testing apparatus that can stress individual filaments.
What they found inverts a common assumption. Uncoated lithium is soft — one of the softest metals. But inside a working battery, dendrites acquire a thin layer of solid electrolyte interphase (SEI), the chemical film that forms where lithium meets electrolyte. The SEI coating makes the dendrite rigid and brittle. Strong enough to pierce the separator. Prone to snapping under stress, leaving behind fragments of “dead lithium” that reduce capacity.
The inversion matters. The SEI layer is essential — it protects the lithium surface from continuous electrolyte decomposition. Without it, the battery would consume its electrolyte in cycles. The same film that enables the battery to function is what converts soft lithium filaments into rigid needles capable of destroying it. The armor is the weapon.
This is a specific instance of a general pattern: protective layers that change the mechanical behavior of what they coat. Oxide films harden aluminum surfaces. Biofilms stiffen medical implants. The protection is never passive — it transforms the substrate's interaction with its environment. In the lithium case, the transformation is from ductile (bends, accommodates stress) to brittle (resists, then fractures catastrophically). The SEI doesn't just sit on the dendrite. It makes the dendrite into something the battery cannot survive.
The engineering implication is that dendrite suppression isn't just about preventing growth — it's about controlling the mechanical consequences of the coating that inevitably forms on whatever does grow. A softer SEI might yield dendrites that bend rather than pierce. The problem was never the lithium. It was the jacket.