Online detection of tobacco combustion components based on laser-induced breakdown spectroscopy

The impacts of tobacco combustion on the atmosphere and human health are very enormous. In order to study the tobacco smoke in the local air, an online observation experimental system based on LIBS (laser-induced breakdown spectroscopy) was designed. This paper conducts an online analysis of the smoke generated by tobacco combustion. Moreover, it compares the ash composition resulting from tobacco combustion with the composition of the smoke. By comparing the spectral line distributions of smoke and ash, it can be observed that the types of elements in the ash are more than those in the smoke. The study found that the composition of ash is more complex than that of smoke. Comparing the spectral intensities of smoke and ash reveals that some elements in the smoke are volatile. Moreover, the CN radicals detected in the experiment were simulated by LIFBASE, and the vibrational and rotational temperatures of the CN radicals were obtained. Additionally, comparisons were made to detect whether substances emitted from burning different parts of tobacco contain heavy metal components. PCA (Principal Component Analysis) was employed to perform dimensionality reduction and analysis of smoke generated from burning different parts of tobacco and cigarettes. By comparing different types of machine learning models, the machine learning with the highest accuracy is selected for classification and tracking. Combined with error back propagation training of artificial neural networks, the classification and traceability models of the smoke from tobacco are performed. This was undertaken to develop a system capable of detecting and categorizing these substances. The online analysis of tobacco combustion components based on LIBS has the advantages of being able to perform online in situ detection. At the same time, it also verifies the feasibility of LIBS application in the tobacco industry.