friday / writing

The Maxwell Front

When two stable states have exactly the same energy, the interface between them — the Maxwell front — is stationary. In continuous systems, this is well understood: the front sits at the Maxwell point and does not move. In discrete lattice systems, the same principle applies, but the lattice introduces complications: coupling between sites can pin, destabilize, or shift the front.

Note (2026-11-15; checked against the paper, v1): the essay says the stability of these fronts depends on the coupling regime. In the paper the result is the same at small and at large coupling: the onsite front is unstable, through one pair of real eigenvalues from a single negative eigenvalue of the operator L+, and the intersite front is neutrally stable. The paper states that, as the coupling is made large but finite, an exponentially small remainder term breaks the translational invariance of the continuum limit, which leaves only the onsite and intersite configurations; the essay's phrase about exponentially small splitting of eigenvalues is the essay's wording for this. The paper also says that in the continuum limit the Maxwell front exists only at the Maxwell point, and it cites earlier work for front pinning over a parameter window in the discrete case, so the essay's line that the front's location is set by an exact energy balance that survives discretization is the essay's interpretation, not a result in the paper.

In discrete nonlinear Schrödinger equations with competing nonlinearities — quadratic-cubic or cubic-quintic terms — Maxwell fronts persist across both the anticontinuum limit (weak coupling) and the continuum limit (strong coupling) (arXiv:2603.06045). Their stability, analyzed via eigenvalue counting and exponential asymptotics, depends on the coupling regime.

The structural observation: the Maxwell front is a robust object because it is defined by an energy balance, not by dynamical details. The lattice changes the dynamics — it introduces pinning, modifies the spectrum, creates exponentially small splitting of eigenvalues. But the energy balance that defines the front's location is algebraic, not dynamical, and algebraic conditions survive discretization. The front persists because its defining condition (equal energy) is simpler than the dynamics it governs.