Desalination membranes face a tradeoff: permeability versus selectivity. Membranes that let water through quickly also let salt through. Membranes that reject salt effectively are slow. The tradeoff has a theoretical upper bound — the permeability-selectivity frontier — and decades of membrane development have pushed performance along this frontier without breaking through it.
A membrane with 6-nanometer channels achieves 55 times higher selectivity than the frontier predicts at its permeability level. The channels are sub-nanometer constrictions within a metal-organic framework, precisely sized to pass water molecules (0.28 nm) while geometrically excluding hydrated sodium ions (0.72 nm). The selectivity is steric — it comes from the pore being the right size, not from chemical interactions or charge exclusion.
The existing frontier was defined by polymeric membranes where selectivity depends on solution-diffusion: both water and salt dissolve into the membrane material and diffuse through it, with selectivity arising from differential solubility and diffusivity. Increasing permeability means making the membrane thinner or more permeable, which also increases salt transport. The tradeoff is intrinsic to the mechanism.
The new membrane operates by a different mechanism — molecular sieving rather than solution-diffusion. The water doesn't dissolve and diffuse; it flows through rigid pores that are geometrically sized to exclude salt. This decouples permeability from selectivity because speed (determined by pore density and length) and rejection (determined by pore diameter) are controlled by independent structural parameters.
The tradeoff wasn't a law of physics. It was a law of one particular mechanism. Changing the mechanism changes the tradeoff. The frontier was the boundary of the solution-diffusion approach, not the boundary of desalination.