arXiv subjects

Johannes Gedeon

Publications and source records attributed to Johannes Gedeon.

5 recordsLinked to original sources

arXiv2026

Spectrally smooth broadband response via autocorrelation-constrained inverse design

Time-domain inverse design in photonics is known to be suitable for maximizing the efficiency of optical devices over broad frequency ranges. Objectives commonly used in this context include time-integrated field quantities derived from Poynting's theorem, such as energy, flux, or dissipated power, which can be directly linked to integrated frequency-domain responses via Parseval's theorem. While computationally efficient, these objectives measure only the total response over the targeted bandwidth and, as we demonstrate, are insufficient to capture undesired in-band ripple, narrow spectral features, or sidelobes. We overcome this limitation by introducing a time-domain metric quantifying such spectral variations based on the weighted long-lag autocorrelation energy of the optical response. We incorporate this metric into an FDTD-based topology-optimization framework and demonstrate its beneficial effect on the example of inverse designing one-dimensional dielectric Bragg mirrors via the time-domain adjoint method.

physics.optics↗

arXiv2025

Topology optimization of a superabsorbing thin-film semiconductor metasurface

We demonstrate a computational inverse design method for optimizing broadband-absorbing metasurfaces made of arbitrary dispersive media. Our figure of merit is the time-averaged instantaneous power dissipation in a single unit cell within a periodic array. Its time-domain formulation allows capturing the response of arbitrary dispersive media over any desired spectral range. Employing the time-domain adjoint method within a topology optimization framework enables the design of complex metasurface structures exhibiting unprecedented broadband absorption. We applied the method to a thin-film Silicon-on-insulator configuration and explored the impact of structural and (time-domain inherent) excitation parameters on performance over the visible-ultraviolet. Since our incorporated material model can represent any linear material, the method can also be applied to other all-dielectric, plasmonic, or hybrid configurations.

physics.optics↗

arXiv2024

Time-domain topology optimization of power dissipation in dispersive dielectric and plasmonic nanostructures

We present a density-based topology optimization scheme for locally optimizing the electric power dissipation in nanostructures made of lossy dispersive materials. By using the complex-conjugate pole-residue (CCPR) model, we can accurately model any linear materials' dispersion without limiting to specific material classes. We incorporate the CCPR model via auxiliary differential equations (ADE) into Maxwell's equations in the time domain, and formulate a gradient-based topology optimization problem to optimize the dissipation over a broad spectrum of frequencies. To estimate the objective function gradient, we use the adjoint field method, and explain the discretization and integration of the adjoint system into the finite-difference time-domain (FDTD) framework. Our method is demonstrated using the example of topology optimized spherical nanoparticles made of Gold and Silicon with an enhanced absorption efficiency in the visible-ultraviolet spectral range. In this context, a detailed analysis of the challenges of topology optimization of plasmonic materials associated with a density-based approach is given.

physics.optics↗

arXiv2023

Free-form inverse design of arbitrary dispersive materials in nanophotonics

In the last decades nanostructures have unlocked myriads of functionalities in nanophotonics by engineering light-matter interaction beyond what is possible with conventional bulk optics. The space of parameters available for design is practically unlimited due to the large variety of optical materials and geometries that can be realized by nanofabrication techniques. Thus, computational approaches are necessary to efficiently search for optimal solutions. In this paper, we enable the free-form inverse design in 3D of linear optical materials with arbitrary dispersion and anisotropy. This is achieved by implementing the adjoint method based on the complex-conjugate pole-residue pair model within a parallel finite-difference time-domain solver, suitable for high-performance computing systems. Our method is tested on the canonical nanophotonic problem of field enhancement in a gap region. The obtained free-form designs of metallic and dielectric materials satisfy the fundamental curiosity of how optimized nanostructures look like in 3D. Unconventional free-form designs revealed by our method, although may be challenging or unfeasible with current technology, bring new insight into how light interacts with nanostructures, and could provide new ideas to inspire forward design.

physics.optics↗

arXiv2021

Machine learning the derivative discontinuity of density-functional theory

Machine learning is a powerful tool to design accurate, highly non-local, exchange-correlation functionals for density functional theory. So far, most of those machine learned functionals are trained for systems with an integer number of particles. As such, they are unable to reproduce some crucial and fundamental aspects, such as the explicit dependency of the functionals on the particle number or the infamous derivative discontinuity at integer particle numbers. Here we propose a solution to these problems by training a neural network as the universal functional of density-functional theory that (i) depends explicitly on the number of particles with a piece-wise linearity between the integer numbers and (ii) reproduces the derivative discontinuity of the exchange-correlation energy. This is achieved by using an ensemble formalism, a training set containing fractional densities, and an explicitly discontinuous formulation.

physics.chem-ph↗