Ferroelectric nematic liquid crystals have a macroscopic electric polarization — the molecular dipoles align, creating a bulk electric field. Confine this material between two flat electrodes, and the polarization must meet the boundary conditions imposed by the electrodes. If the electrode geometry is simple (parallel plates, uniform spacing), the boundary conditions are trivially satisfied by uniform polarization.
The paper shows that even in the simplest planar electrode cells, the boundary conditions create electrostatic frustration. The polarization cannot be simultaneously uniform (minimizing elastic energy), aligned with the electrodes (satisfying boundary conditions), and divergence-free (minimizing electrostatic energy). At least one of these conditions must be violated, and the violation produces domains — regions of different polarization orientation separated by domain walls.
The domains are not equilibrium structures in the usual sense. They're the minimum-energy configuration given that no configuration satisfies all constraints simultaneously. The domain structure is the solution to a frustration problem: which constraint to violate, where to violate it, and how much to violate it.
The result is “polarization plasticity” — the material deforms its polarization pattern under applied fields in ways that resemble plastic deformation of solids. The polarization doesn't smoothly rotate (elastic response); it switches between domain configurations (plastic response). The switching involves domain wall motion, nucleation of new domains, and hysteresis — all from a liquid crystal, which has no crystalline lattice to pin domain walls. The frustration comes from the electrostatics, not from the molecular structure. Even a perfect liquid produces domains when the geometry frustrates the polarization.