Nonlinear Photonic Crystals: IV. Nonlinear Schrodinger Equation Regime

TL;DR

This paper rigorously derives the nonlinear Schrödinger equation as an approximation for electromagnetic wave propagation in nonlinear photonic crystals, introducing extended NLS equations that improve accuracy over classical models.

Contribution

It develops a systematic approximation framework for nonlinear Maxwell equations, deriving the NLS regime and extended NLS equations with quantitative error estimates.

Findings

NLS accurately models wave evolution under specific excitation conditions.

Extended NLS equations provide significantly better approximation accuracy.

Rigorous error bounds are established for the approximations.

Abstract

We study here the nonlinear Schrodinger Equation (NLS) as the first term in a sequence of approximations for an electromagnetic (EM) wave propagating according to the nonlinear Maxwell equations (NLM). The dielectric medium is assumed to be periodic, with a cubic nonlinearity, and with its linear background possessing inversion symmetric dispersion relations. The medium is excited by a current $\mathbf{J}$ producing an EM wave. The wave nonlinear evolution is analyzed based on the modal decomposition and an expansion of the exact solution to the NLM into an asymptotic series with respect to some three small parameters $\alpha $, $\beta $ and $\varrho $. These parameters are introduced through the excitation current $\mathbf{J}$ to scale respectively (i) its amplitude and consequently the magnitude of the nonlinearity; (ii) the range of wavevectors involved in its modal composition, with $\beta ^{-1}$ scaling its spatial extension; (iii) its frequency bandwidth, with $\varrho ^{-1}$ scaling its time extension. We develop a consistent theory of approximations of increasing accuracy for the NLM with its first term governed by the NLS. We show that such NLS regime is the medium response to an almost monochromatic excitation current $\mathbf{J} $. The developed approach not only provides rigorous estimates of the approximation accuracy of the NLM with the NLS in terms of powers of $\alpha $, $\beta $ and $\varrho $, but it also produces new extended NLS (ENLS) equations providing better approximations. Remarkably, quantitative estimates show that properly tailored ENLS can significantly improve the approximation accuracy of the NLM compare with the classical NLS.