Pixelated Bayer Spectral Router Based on Sparse Meta-atom Array

TL;DR

This paper introduces a pixelated spectral router using sparse meta-atom arrays that significantly enhances light separation efficiency in Bayer sensors, offering polarization insensitivity and high incident angle tolerance for improved image sensing.

Contribution

It presents a novel CMOS-compatible spectral router based on Si3N4 nanostructures that improves light collection and spectral separation in image sensors, with robustness and scalability.

Findings

Over 56% signal enhancement compared to traditional filters

High incident angle tolerance over 30 degrees

Polarization insensitivity and compatibility with various pixel sizes

Abstract

It has long been a challenging task to improve the light collection efficiency of conventional image sensors built with color filters that inevitably cause the energy loss of out-of-band photons. Although various schemes have been proposed to address the issue, it is still very hard to make a reasonable tradeoff between device performance and practicability. In this work, we demonstrate a pixelated spectral router based on sparse meta-atom array, which can efficiently separate the incident R (600-700 nm), G (500-600 nm), and B (400-500 nm) band light to the corresponding pixels of a Bayer image sensor, providing over 56% signal enhancement above the traditional color filter scheme. The CMOS-compatible spectral router has superior characteristics of polarization insensitivity and high incident angle tolerance (over 30{\deg}), enabled by simple compound Si3N4 nanostructures which are very suitable for massive production. Imaging experiments are conducted to verify its potential for real applications. Our pixelated spectral router scheme is also found to be robust and could be freely adapted to image sensors of various pixel sizes, having great potential in building the new generation of high-performance image sensing components.