Burner system for solid sample flame emission spectroscopy

We present a burner system to analyze solid inflammable samples utilizing flame emission spectroscopy without requiring any sample preparation procedures. The acetylene-nitrous oxide burner was designed to efficiently introduce solid particles into the flame through active injection, enabling real-time elemental analysis. Computational Fluid Dynamics (CFD) simulations were employed to study particle transport dynamics within the burner system. The emission was characterized through spectral analysis of the flame emission from metal powder mixtures, demonstrating its ability to determine elemental compositions without prior sample treatment. An artificial neural network (ANN) was implemented to analyze spectral data obtained from binary metal mixtures, enabling rapid and reliable identification of constituent elements with an uncertainty of = 2.7 % (mol/mol)