Reconfigurable miniaturized computational spectrometer enabled by photoelastic effect

TL;DR

This paper introduces ElastoSpec, a reconfigurable, miniaturized computational spectrometer that uses the photoelastic effect with simple components, enabling accurate spectral measurement without complex fabrication.

Contribution

It presents a novel broadband spectrometer design leveraging the photoelastic effect, eliminating the need for micro-nano fabrication and enabling reconfigurable, cost-effective spectral sensing.

Findings

Achieves approximately 0.2 nm FWHM error for monochromatic inputs.

Maintains MSE around 10^-3 with only 10 spectral modulation units.

Demonstrates excellent reconstruction accuracy for simple and complex spectra.

Abstract

Miniatured computational spectrometers, distinguished by their compact size and lightweight, have shown great promise for on-chip and portable applications in the fields of healthcare, environmental monitoring, food safety, and industrial process monitoring. However, the common miniaturization strategies predominantly rely on advanced micro-nano fabrication and complex material engineering, limiting their scalability and affordability. Here, we present a broadband miniaturized computational spectrometer (ElastoSpec) by leveraging the photoelastic effect for easy-to-prepare and reconfigurable implementations. A single computational photoelastic spectral filter, with only two polarizers and a plastic sheet, is designed to be integrated onto the top of a CMOS sensor for snapshot spectral acquisition. The different spectral modulation units are directly generated from different spatial locations of the filter, due to the photoelastic-induced chromatic polarization effect of the plastic sheet. We experimentally demonstrate that ElastoSpec offers excellent reconstruction accuracy for the measurement of both simple narrowband and complex spectra. It achieves a full width at half maximum (FWHM) error of approximately 0.2 nm for monochromatic inputs, and maintains a mean squared error (MSE) value on the order of 10^-3 with only 10 spectral modulation units. Furthermore, we develop a reconfigurable strategy for enhanced spectra sensing performance through the flexibility in optimizing the modulation effectiveness and the number of spectral modulation units. This work avoids the need for complex micro-nano fabrication and specialized materials for the design of computational spectrometers, thus paving the way for the development of simple, cost-effective, and scalable solutions for on-chip and portable spectral sensing devices.