Plasmonic Mid-Infrared Filter Array-Detector Array Chemical Classifier Based on Machine Learning

Abstract

Numerous applications exist for chemical detection, ranging from the industrial production of chemicals to pharmaceutical manufacturing, environmental monitoring and hazardous risk control. For many applications, infrared absorption spectroscopy is the favored technique, due to attributes that include short response time, high specificity, minimal drift, in-situ operation, negligible sample disruption, and reliability. The workhorse instrument for infrared absorption is the Fourier transform infrared (FTIR) spectrometer. While such systems are suitable for many purposes, new applications would be enabled by small, lightweight, low power and low cost infrared microspectrometers. Here we perform a detailed study on a microspectrometer chemical classifier comprising an array of plasmonic mid-infrared spectral filters used with a photodetector array, whose outputs are analyzed by a machine learning algorithm. We conduct simulations (including noise), demonstrating the identification of six gas-phase and six liquid-phase chemicals. We study the performance of our method at detecting the concentration of acetylene.

Infrared metamaterial for surface-enhanced infrared absorption spectroscopy: pushing the frontier of ultrasensitive on-chip sensing

“…[163] (e) Plasmonic mid-infrared filter array-detector array chemical classifier based on machine learning. [164] I: Schematic diagram of the proposed detector. II: Schematic illustration of the filter array.…”

“…2021. [164] In this work, support vector machine (SVM) classifiers were adopted to directly analyze the spectra and determine the chemical species. It is described that five analytes were successfully distinguished after the training of ML classifiers, and the minimum concentration was 10 ppm (Figure 7(e)).…”

Intelligent metaphotonics empowered by machine learning

“…First, metasurfaces can be employed to optimize sensors [130][131][132][133]. On the other hand, metasurfaces empowered by ML techniques provide a powerful tool for classification tasks, which can be exploited to analyze output data of a sensor [134][135][136][137].…”

“…Classical ML methods, such as k nearest neighbours or principal component analysis with latent Dirichlet allocation, may provide no less good results. For instance, gaseous and liquid chemicals may be recognised using plasmonic mid-infrared spectral filters with a photodetector array [137] or plasmonic metasurfaces integrated with microfluidic channel [135].…”

Ultrathin Optics-Free Spectrometer with Monolithically Integrated LED Excitation

“…For example, the locations of a spectrum’s peaks and the relative intensities between them are often sufficient information. Moreover, the reconstruction algorithms can be enhanced by adding constraints that can either be improved over time via learning or be constructed based on the specific characteristics of the application of interest [ 26 , 27 , 28 ]. This is highly desirable for many applications targeted by a miniaturized spectrometer.…”