Adjoint sensitivity analysis for nanophotonic structures

Placeholder Show Content


We present a local optimization method for design of dispersive and nondispersive dielectric structures for applications in nanophotonics, based on adjoint solutions of Maxwell's equations. The sensitivity of a merit function (e.g. total absorption) is sought with respect to design parameters (e.g. shape and size of scattering structures). We use adjoint design sensitivity analysis to obtain this sensitivity in a computationally efficient manner. By discretizing Maxwell's equations as a linear FDTD system with matrix elements that vary smoothly with design parameters, the entire numerical system is made differentiable. The derivative of the merit function with respect to all design parameters may be derived using the chain rule, and calculated efficiently using a solution to the adjoint FDTD system. Next we formulate the adjoint problem as a partial differential equation and solve it with the finite element method. This more accurate method is applied to metals. An ``optimization force'' may be calculated for all design parameters, or visualized at points along material interfaces to lend insight into design rules. We apply this method to the design of waveguide mode converters and resonant metallic nano-apertures.


Type of resource text
Form electronic; electronic resource; remote
Extent 1 online resource.
Publication date 2014
Issuance monographic
Language English


Associated with Hansen, Paul
Associated with Stanford University, Department of Applied Physics.
Primary advisor Hesselink, Lambertus
Thesis advisor Hesselink, Lambertus
Thesis advisor Byer, R. L. (Robert L.), 1942-
Thesis advisor Fan, Shanhui, 1972-
Advisor Byer, R. L. (Robert L.), 1942-
Advisor Fan, Shanhui, 1972-


Genre Theses

Bibliographic information

Statement of responsibility Paul Hansen.
Note Submitted to the Department of Applied Physics.
Thesis Thesis (Ph.D.)--Stanford University, 2014.
Location electronic resource

Access conditions

© 2014 by Paul Christopher Hansen
This work is licensed under a Creative Commons Attribution Non Commercial 3.0 Unported license (CC BY-NC).

Also listed in

Loading usage metrics...