Discrimination of Explosive Residues by Standoff Sensing Using Anodic Aluminum Oxide Microcantilever Laser Absorption...

Area of Science:

• Materials Science
• Spectroscopy
• Machine Learning

Background:

• Standoff laser absorption spectroscopy (LAS) is crucial for environmental safety applications.
• Developing sensitive and selective methods for detecting explosive residues is a significant challenge.

Purpose of the Study:

• To propose and evaluate an anodic aluminum oxide (AAO) microcantilever LAS system integrated with machine learning (ML) for standoff detection of explosive residues.
• To compare the performance of AAO microcantilevers with traditional silicon microcantilevers.

Main Methods:

• Fabrication of a nanoporous AAO microcantilever using micromachining.
• Measurement of standoff infrared (IR) spectra of explosive residues using the AAO microcantilever LAS.
• Analysis of spectral data using various ML models, including kernel extreme learning machines (KELMs), support vector machines (SVMs), random forest (RF), and backpropagation neural networks (BPNNs).

Main Results:

• The AAO microcantilever exhibited a lower spring constant compared to silicon microcantilevers.
• Standoff IR spectra obtained with the AAO microcantilever closely matched those from standard Fourier transform infrared spectroscopy.
• Kernel-based ML models (KELM and SVM) demonstrated high prediction accuracy (up to 95.8%) and efficient hyperparameter optimization.

Conclusions:

• The AAO microcantilever LAS system, coupled with kernel-based ML models, provides a sensitive and selective method for standoff discrimination of explosive residues.
• This integrated approach shows promise as an efficient sensing technology for enhanced safety monitoring.