On-Line Compositional Measurements of Auag Aerosol Nanoparticles Using Optical Emission from Spark Ablation

Abstract

Spark ablation is an established technique for generating aerosol nanoparticles. Recent demonstrations of compositional tuning of bimetallic aerosols have led to a demand for on-line stoichiometry measurements. In this work, we present a simple, non-intrusive method to determine the composition of a binary AuAg nanoparticle aerosol on-line using the optical emission from the electrical discharges. Machine learning models based on the least absolute shrinkage selection operator (LASSO) were trained on optical spectra datasets collected during aerosol generation and labelled with X-ray fluorescence spectroscopy (XRF) compositional measurements. Models trained for varying discharge energies demonstrated good predictability of nanoparticle stoichiometry with mean absolute errors < 10 at. %.  While the models utilized the emission spectra at different wavelengths in the predictions, a combined model using spectra from all discharge energies made accurate predictions of the AuAg nanoparticle composition, showing the method’s robustness under variable synthesis conditions.

Keywords: Spark ablation, bimetallic nanoparticles, plasma spectroscopy, optical diagnostics, machine learning