Investigation of Exhaust Particles on Different TEM Grids: a Comparison between Graphene Oxide and Silicon Nitride Grids

Abstract

<div class="section abstract"><div class="htmlview paragraph">Two different TEM (Transmission Electron Microscopy) grids - graphene oxide (GO) and silicon nitride (SiN) - were used to capture the particulates emitted with the exhaust of a modern 1.0 L GDI (Gasoline Direct Injection) engine. One speed-load condition (1250 rpm – idle) was chosen to generate a nanometric particulate output in the sub-23 nm regime which has been traditionally difficult to analyse in terms of composition and morphology. The overall aim was to understand if additional benefits can be obtained by analysing the particles captured in the exhaust on a nanoporous silicon nitride grid compared to state-of- the-art graphene oxide grids. The behaviour of porous SiN support films was of interest since nanopores are present in the grid in the 20 nm regime and the material is thermally and dimensionally stable under high temperatures, allowing thermophoretic capture directly within the engine exhaust stream. In addition to nanostructural and morphological comparison, the elemental composition of the particles was also analysed by EDX (Energy Dispersive X-Ray). Particles were thermophoretically captured directly in the exhaust stream using a dedicated probe holding the grid. Because of their tiny 2 nm thickness, GO grids work well for studying particle nanostructure, however background noise from copper and carbon interferes with compositional analysis by EDX. In contrast, the silicon nitride grids enable particle observation without this background noise, providing an intriguing platform for the analysis of the suspended particles collected by the pores. Future research will concentrate on producing particles with graphitic areas to assess imaging advantages in terms of morphological and nanostructural examination. The two grids were similar in their particle capture within the engine, with close mean primary particle diameters using both: 13.5 ± 3.1 nm standard deviation (SD) on GO and 14.1 ± 2.6 nm SD on SiN. EDX analysis suggests SiN grids, as C-free substrates, are preferable to GO for determining the carbon load in captured particles. This investigation is part of a larger project focusing on decarbonised fuels, so a carbon-free support film is pivotal in understanding the nature and composition of the fine particles linked to the lubricant oil.</div></div>