# On the importance of light scattering for high performances nanostructured antireflective surfaces

## TL;DR

This paper introduces a novel simulation method to predict and optimize light scattering losses in nanostructured antireflective coatings, demonstrating that simple bilayer coatings can achieve near-perfect broadband transparency.

## Contribution

It presents an original simulation approach for scattering losses in nanostructured antireflective surfaces and compares discrete versus continuous gradient coatings for enhanced transparency.

## Key findings

- Discrete bilayer coatings achieve 98.97% transmittance across 400-1800 nm
- Scattering losses are minimized by small nanostructure dimensions and interference effects
- The method accurately predicts scattering behavior validated by electron tomography and FDTD simulations

## Abstract

An antireflective coating presenting a continuous refractive index gradient is theoretically better than its discrete counterpart because it can give rise to a perfect broadband transparency. This kind of surface treatment can be obtained with nanostructures like moth-eye. Despite the light scattering behavior must be accounted as it can lead to a significant transmittance drop, no methods are actually available to anticipate scattering losses in such nanostructured antireflective coatings. To overcome this current limitation, we present here an original way to simulate the scattering losses in these systems and routes to optimize the transparency. This method, which was validated by a comparative study of coatings presenting refractive indices with either discrete or continuous gradient, shows that a discrete antireflective coating bilayer made by oblique angle deposition allows reaching ultra-high mean transmittance of 98.97% over the broadband [400-1800] nm. Such simple surface treatment outperforms moth-eye architectures thanks to both interference effect and small dimensions nanostructures that minimize the scattering losses as confirmed by finite-difference time domain simulations performed on reconstructed volumes obtained from electron tomography experiments.

---
Source: https://tomesphere.com/paper/1906.06299