Analysis of sub-micrometric particulate emitted by different types of internal combustion engines: a Raman microscopy and a Principal-Component Analysis study

Particulate matter emitted by internal combustion engines is a major contributor to air pollution and is associated with adverse health effects. While black carbon is the dominant component of engine-derived particulates, the detailed chemical nature of fine and ultrafine particles still needs to be accurately characterized. This study aims to develop an advanced analytical approach to investigate the physico-chemical composition of fine and sub-micrometric exhaust particulate matter from a wide range of vehicle types and fuel technologies. Raman micro-spectroscopy was employed to analyse particulate samples collected during chassis dynamometer tests from both light-duty and heavy-duty vehicles running on diesel, gasoline, liquefied petroleum gas, compressed natural gas, and hydrotreated vegetable oil. Samples were taken both before and after exhaust dilution using a constant volume sampling system. Spectra revealed that black carbon, composed of disordered graphitic structures, was the primary constituent across all samples. However, additional carbonaceous and non-carbonaceous species were also detected. Notably, Raman features characteristic of multi-walled carbon nanotubes were identified exclusively in samples collected before dilution, and further confirmed by transmission electron microscopy and principal component analysis. Iron oxides, sulphates, and nitrogen-containing compounds were also observed, particularly in diesel exhaust particulate, suggesting contributions from engine wear, lubricant additives, and exhaust after-treatment systems. This work demonstrates that Raman spectroscopy combined with multivariate statistical analysis offers a powerful tool to differentiate between various chemical species in exhaust particulate matter. Multi wall carbon nanotubes have been only identified in samples collected before dilution. Their absence in diluted samples raises concerns about the ability of standard sampling procedures to retain such fine, potentially volatile or reactive species. These findings highlight the need to revise sampling strategies and extend chemical characterization to ultrafine particulate below 100 nm. This integrated spectroscopic–statistical approach provides a methodology that can be extended to other combustion- and non-combustion-related particulate sources. Future research will focus on enhancing analytical resolution, expanding the range of tested vehicle technologies, and investigating non-exhaust emissions such as those from brakes and tyres.