Sources of anthropogenic secondary organic aerosol in the atmosphere
Loading...
Date
2026-10-07
Authors
Παυλίδης, Δαμιανός
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Organic aerosol is a major component of fine particulate matter and a large fraction of it is secondary in origin, formed through the atmospheric oxidation of gas-phase precursors. Anthropogenic activities, and in particular motor vehicles, are an important source of these precursors, yet their contribution to secondary organic aerosol (SOA) formation remains poorly understood. This thesis addresses these knowledge gaps through a combination of laboratory studies using a smog chamber and an oxidation flow reactor, together with real-world field measurements. The findings contribute to improving the representation of vehicle emissions in emission inventories and SOA formation and chemical transport models. This work first addresses a methodological challenge that affects the interpretation of chamber SOA formation experiments. The high-resolution time-of-flight aerosol mass spectrometer (AMS) has become one of the most important instruments in the field of atmospheric sciences, providing real-time measurements of the mass concentration, size distribution, and chemical composition of submicrometer non-refractory aerosol, along with the scanning mobility particle sizer, which measures the particle number size distribution. However, AMS measurements require correction for the instrument’s collection efficiency, which is commonly assumed to be independent of particle size. This assumption often leads to significant discrepancies between the measurements of the two instruments in SOA formation experiments with new particle formation. To address this, a new methodology is proposed for the estimation of a size- and composition-dependent collection efficiency. Applied to two sets of chamber experiments, nonylcyclohexane oxidation and α-humulene ozonolysis, the new approach reduced the average deviation between the two instruments to below 10% for both total aerosol and SOA mass. SOA formation from the photooxidation of large anthropogenic volatile and intermediate-volatility organic compounds with hydroxyl radicals under high-NOx conditions, often encountered in urban areas, was investigated next. The precursors studied included cyclic alkanes (amylcyclohexane, hexylcyclohexane, nonylcyclohexane, and decylcyclohexane) and aromatic compounds of increasing size (1,3,5-trimethylbenzene, 1,3,5-triethylbenzene and 1,3,5-tritert-butylbenzene). A combination of measurements was performed for each precursor to develop SOA parameterizations for use in chemical transport models. Specifically, the volatility and effective vaporization enthalpy of the oxidation products were estimated using an algorithm that combined the observed SOA yields with results from two evaporation techniques, thermodenuding and isothermal dilution. Significant SOA formation was observed for all investigated compounds, with cycloalkanes exhibiting higher yields than aromatic compounds, while aromatic precursors produced more oxidized SOA. Additionally, the oxidation products of cycloalkanes were dominated by semivolatile organic compounds, whereas those of aromatic precursors were mainly distributed in the intermediate-volatility range. To enhance our understanding of cold start emissions from passenger cars, which dominate urban traffic, measurements of approximately 21,000 cold starts were carried out inside an underground parking garage from November 29 to December 20, 2024. The underground facility, located beneath a shopping mall in Patras, Greece and consisting of 246 parking spaces, provided a controlled environment representative of real-world cold start conditions. Measurements of traffic and pollutant concentrations were combined to derive average emissions of gas- and particulate-phase pollutants. The average parking time of vehicles was 1.5 h and vehicle emissions were driven almost entirely by cold starts. The emitted submicron particulate matter consisted primarily of organic aerosol (61%) and black carbon (39%), while volatile organic compounds were dominated by C6–C9 aromatics (65%). The estimated cold start emission factors were then compared with existing emission inventory estimates to assess the representation of cold start emissions in European emission inventories. In the final part of the thesis, individual on- and off-road vehicles were tested on a dynamometer, with their exhaust introduced into an atmospheric simulation chamber and an oxidation flow reactor to investigate SOA formation across a range of photochemical ages. In total, tailpipe emissions from 14 vehicles, including 4 gasoline cars, 4 diesel cars, 4 scooters, and 2 tractors, covering a range of engine types and emission control technologies, were examined. The SOA formation potential (SOAFP, SOA formed per kilogram of fuel) estimated using the two reactors showed similar trends across vehicle categories and was strongly correlated (R2=0.90). However, the oxidation flow reactor consistently produced more oxidized SOA than the smog chamber at comparable OH exposures, with oxygen-to-carbon ratios on average 47% higher. Scooters (pre-Euro–Euro 5) and the off-road vehicles had the highest SOAFP values, exceeding those of all on-road gasoline and diesel vehicles. For these vehicles, SOA also accounted for more than 80% of the total aged organic aerosol. The volatility characterization further showed that the aged emissions (with a saturation concentration below 106 μg m-3) of most vehicles were dominated by intermediate-volatility organic compounds (58 ± 19%). Overall, these findings improve the characterization of vehicle-related SOA, highlighting scooters and off-road vehicles as potentially important SOA sources, and support the oxidation flow reactor as a practical tool for SOAFP evaluation in the development of future vehicle emission control policies.
Description
Keywords
Atmospheric chemistry, Secondary organic aerosol (SOA), Anthropogenic sources