Discussion Forum

I think it is the same reason as in one of your earlier posts. The material properties are constant and do not depend on the wavelength.

I don't have the ray optics module, but if I understand your model correctly, you have a 350-micron thick layer of silver. Maybe I'm just naive, but doesn't that represent a nice optical mirror? Or are you trying to model something subtle or unusual, like in this article: "Seeing through Silver," Phys. Rev. Focus 18, 4, August 7, 2006 ?

I don't have the ray optics module, but if I understand your model correctly, you have a 350-micron thick layer of silver. Maybe I'm just naive, but doesn't that represent a nice optical mirror? Or are you trying to model something subtle or unusual, like in this article: "Seeing through Silver," Phys. Rev. Focus 18, 4, August 7, 2006 ?

I'm aiming to characterize the reflectance versus wavelength for a glass + silver configuration, with glass as the substrate, similar to the pattern shown in the attached graph (a random Google image). My ultimate objective is to simulate a Passive Daytime Radiative Cooling device (PDRC) comprising layers of Glass, Silver, and PDMS, in that order. I initially used a 350-micron thick silver layer due to uncertainty about the precise thickness of the silver layer. However, the actual thickness of the Silver layer is intended to be 120 nm.

Well, no surprise you aren't getting expected results. 120 nm is by a factor 3000 smaller than the thick layer and it is only a fraction of the wavelength at the longer wavelengths of the spectrum you model. I am not sure if ray optics is appropriate for that. You may need to consider wave optics for that.

There is a presentation called "Visible Spectral Reflectance Analysis in a Metal-Insulator-Metal (MIM) Multilayer with COMSOL Multiphysics" available to you in the Technical papers section of the Comsol website that I think may be of interest to you. See https://www.comsol.com/paper/download/159335/oshikane_presentation.pdf

Hello again,

I've encountered a minor yet intriguing update with the glass(+silver) model.

To provide some background, in a previous discussion post titled, Reflectance Vs Wavelength (Just Glass), I sought advice on the unchanging reflectance of a sole glass substrate across various wavelengths in my plot. @Edgar J. Kaiser kindly suggested the use of an interpolation function for the refractive index, n(λ).

After exploring, I discovered certain materials in COMSOL already equipped with such functions. I decided to switch the glass in my model to a Corning glass variant with a built-in interpolation function for refractive indices, leading to the expected wavelength-dependent reflectance changes.

Encouraged, I applied this method to my Glass+Silver setup, given that Silver (specifically, Ag (Silver) (Choi et al. 2020: n,k 1.23-6.99 um)) also has its refractive index interpolation function. However, despite numerous adjustments to the glass and silver layers' thickness, the model persistently shows a 100% reflectance across all wavelengths, defying my expectations.

I had anticipated that using materials with predefined interpolation functions would rectify the issue, yet it hasn't. Any insights into why this might be happening?

I must stress that I am new to COMSOL :) Here's a newer version of the COMSOL file: https://drive.google.com/file/d/1T_YSvZhXtXUMZSuFbkdmWOjdfr9o7Tsa/view?usp=sharing Let me know if I can provide anything else.