Burner system for solid sample flame emission spectroscopy

Hans-Peter Loock,Boyd Davis,Adam Bernicky
DOI: https://doi.org/10.26434/chemrxiv-2024-jjj4n
2024-06-18
Abstract: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)
Chemistry
What problem does this paper attempt to address?
This paper presents a combustion system for flame emission spectroscopy analysis of solid combustible sample, aiming to achieve real-time elemental analysis of solid particles without any sample pretreatment. In the study, a burner fueled by acetylene-nitrous oxide was designed to effectively introduce solid particles into the flame using the active injection technique. The transport dynamics of particles in the burner were studied through computational fluid dynamics (CFD) simulations, and the capability to determine the elemental composition was demonstrated by the flame emission spectroscopy characteristics of metal powder mixtures without the need for sample pretreatment. The paper also mentions that the process control in copper smelting is very limited due to the continuously changing composition of raw materials, requiring frequent adjustments of oxygen and fuel ratios. Existing monitoring methods, such as flame-based blackbody temperature and brightness measurements, have limitations and cannot measure element concentrations. Therefore, the research team developed an artificial neural network (ANN) algorithm for analyzing spectral data obtained from binary metal mixtures, which achieved fast and reliable element identification with an uncertainty of about 2.7% (sum of molar ratios). In the study, the authors designed a burner with an inclined surface to ensure that the powder samples are surrounded by the flame when leaving the feed pipe, enabling effective conversion and consumption of mineral sulfides. In addition, a Venturi ejector was used to improve powder conveying, and the issue of powder accumulation at the burner orifice was addressed through computational fluid dynamics analysis. With this new combustion system, solid samples can be directly analyzed spectroscopically without traditional dissolution steps, and the neural network algorithm can accurately analyze the copper and iron concentrations of the samples, providing a new approach for process control in the copper smelting industry.