An assessment on how different collection methods impact thermal properties, surface functional groups, nanostructure and morphology of diesel particulate matter

• Published in: Combustion and flame 2021-03, Vol.225, p.74-85
• Main Authors: Ávila, Cristian D., Botero, María L., Agudelo, Andrés F., Agudelo, John R.
• Format: Article
• Language: English
• Subjects: Automotive engines; Chemical composition; Collection; Collection method; Diesel engines; Diesel particulate matter; Dilution; Fiberglass; Fluorescence; Fractal analysis; Functional groups; Lattice parameters; Morphology; Nanostructure; Oxidation; Oxidation reactivity; Particulate emissions; Polytetrafluoroethylene; Stainless steels; Thermodynamic properties; Thermogravimetric analysis; Thermogravimetry; Uncertainty analysis; Vacuum pumps
• ISSN: 0010-2180; 1556-2921
• DOI: 10.1016/j.combustflame.2020.10.042

This work assessed the effect of gathering particulate matter from the exhaust of an automotive diesel engine through: a vacuum pump using Teflon filters; a particle trap with stainless-steel wools; a thermophoretic probe with Lacey-carbon TEM grids, and a partial dilution mini tunnel with fiberglass filters, on PM thermal behavior (TGA), chemical functional groups on PM surface (FTIR), and nanostructure (Raman, XRD, HRTEM and TEM). Results showed that for TGA, particulate trap was the most suitable method. The degree of compactness of PM samples gathered with vacuum pump, affected thermogravimetric analysis. The significant uncertainties induced by PM sample preparation from mini tunnel make it unsuitable for TGA. FTIR results revealed that PM samples collected with vacuum pump were the most suitable method to analyze functional groups. PM from particulate trap was passive partially oxidized due to long collection times as evidenced by the 1640 cm−1 wavelength peak. Given that particulate trap method provides a large amount of PM sample, it would be recommendable to shorten enough collection time to avoid PM passive oxidation. Particulate trap and vacuum pump methods showed similar Raman and XRD results. Mini tunnel samples exhibited significantly higher AD3/AG and AD4/AG peaks due to filter fluorescence, which besides induced interference in XRD analysis. Particulate trap samples exhibited a pronounced peak in 2θ = 43° corresponding to stainless-steel contamination. Although it did not affect the lattice parameters from XRD, HRTEM was more suitable for nanostructural characterization. Particles gathered by thermophoretic probe do not require additional post-treatment, making this more suitable for nanostructural and fractal analysis. Results revealed that using a unique collection method may induce significant uncertainties depending on the PM specific property to be analyzed. It was found that a combination of particulate trap and thermophoretic probe methods would be suitable for PM thermal and nanostructural characterization.