Determination of abundance, composition, and sources of carbonaceous components within particulate matter from urban environments within the UK

Nikkhah-Eshghi, Melissa (2023) Determination of abundance, composition, and sources of carbonaceous components within particulate matter from urban environments within the UK. PhD thesis, University of Glasgow.

Abstract

Airborne ambient particulate matter (PM) has detrimental effects on human health and the environment and its concentrations are usually higher in urban areas. The carbonaceous component is a major constituent of particulate matter with an aerodynamic diameter of ≤2.5 μM (PM₂.₅). It comprises of black carbon (BC) and many individual species of organic compounds such as polycyclic aromatic hydrocarbons (PAHs) and n-alkanes. Despite this importance, BC remains a poorly quantified and poorly characterised component of PM, relative to other PM fractions. The aims of this work were therefore to:

1) Determine the abundance and spatiotemporal trends of airborne BC in Glasgow, Scotland’s largest city.
2) Obtain samples of airborne PM from Glasgow and elsewhere in the UK, quantify the concentrations of PM-bound PAHs and estimate the associated health risks of exposure to these PM-bound PAHs.
3) Develop hydropyrolysis (hypy) methodology to isolate BChypy from airborne ambient PM, obtaining BC/Total Carbon % (BC/TC%) measurements.
4) Undertake source apportionment of the carbonaceous components (PAHs, n-alkanes, TC and BC) of PM from Glasgow and elsewhere. Radiocarbon (¹⁴C) analysis and delta carbon-13 (δ¹³C) were to be used for source apportionment of TC and BC. Principal component analysis (PCA) were to be used to determine sources of PAHs and n-alkanes.

To determine the abundance and spatiotemporal trends in Glasgow, two portable microaethalometer instruments were used. These were calibrated through co-location against the UK BC network aethalometer at the Glasgow Townhead urban background site. In four Glasgow wards, one microaethalometer was placed at a fixed location whilst the other was used in mobile measurements to determine the relative and absolute spatiotemporal variations in BC. BC increment concentrations calculated through subtracting the background BC concentrations from Glasgow Townhead site and absolute concentrations exhibited spatial variation across the wards. The ‘mobile’ measurements showed the city centre to be a hotspot whilst Baillieston, a residential ward on the outskirts of the city, had the lowest BC increments. Population, population densities and socio-economic data could be used as valuable BC indicators to determine other BC hotspots in Glasgow and beyond. Temporally, median BC increments were significantly higher in mid-morning (10:30 – 11:30) than earlier morning (08:00 – 09:00). However, absolute BC concentrations had the opposite trend showing that background BC contributes highly to morning BC concentrations. This suggests that use of BC increments rather than BC concentrations can help improve understanding of local pollution events by removing background influences. BC increments were correlated strongly with NOx concentrations and bus/heavy goods vehicles average daily traffic counts at the sampling sites. This showed that Glasgow City Council could focus local air quality measures towards reducing emissions from buses and HGVs.

Samples of PM2.5 were collected from two sites in Glasgow using a low volume air sampler: GLA-HH (University Avenue, Hillhead), GLA-CC (George Street, City). Samples of PM2.5, and of particulate matter with an aerodynamic diameter of ≤10 μM (PM10), from GLA-KS (Hope Street, City) were provided by Ricardo, an environmental consulting services company. Samples were also collected from a comparison site, an urban background site located in Manchester Fallowfield Campus (MAN-FF). The GLA-KS and MAN-FF samples were collected by high volume air samplers. For n-alkane and PAH analysis, samples were extracted by accelerated solvent extraction (ASE) using hexane and acetone, and then compound groups were separated using column chromatography.

GLA-KS benzo(a)pyrene (BaP) concentrations exceeded the UK National Air Quality Objective (NAQO) of 0.25 ng m⁻³ and were close to the EU Fourth Air Quality Daughter Directive (FAQDD) limit value of 1 ng m⁻³. This implied that inhalation of PM2.5 from GLA-KS could be associated with potential health effects such as lung cancer. BaP concentrations in the MAN-FF samples did not exceed the NAQO or FAQDD values, suggesting that suburban areas had lower associated health risks than inner-city areas. The increased lung cancer risk through inhalation (ILCRᵢₙₕ) was calculated for PM-bound PAHs. The ILCRᵢₙₕ using the World Health Organization IUR₈ₐₚ value resulted in ILCRᵢₙₕ of 2.09 x 10⁻⁴ for GLA-KS(PM₂.₅) and 4.60 x 10⁻⁵ for GLA-KS(PM₁₀). This equated to 209 excess lung cancer cases per 1,000,000 people for PM₂.₅ and 46 excess lung cancer cases per 1,000,000 people for PM₁₀. When applied to the Anderston, City and Yorkhill ward population this translated to 6 excess annual cases of lung cancer, assuming continuous and homogenous exposure of inhabitants. This shows that health risks posed to those living in Anderston, City and Yorkhill ward is not negligible and should be improved through PM₂.₅ reductions.

The hypy methodology was tested on various different materials. Purchased aerosol and dust standards (NIST 1649b Urban Dust, NIST 1648a Urban PM, BCR-723 Road dust and ERMCZ100 Fine Dust PM₁₀-like) were used in the study. Two of these materials represented urban aerosols (NIST 1649b and 1648a), and the other two materials (BCR-723 road dust and ERMCZ100 Fine Dust) represented aerosols with dominant vehicular emission sources as they were collected from traffic tunnels. An in-house road dust sample from Glasgow Hillhead was collected for analysis. Materials that represented end members of BC were also used. NIST 2975 diesel PM was used to represent high BC concentrations, whilst in-house charred biomass materials were used to represent lower BC concentrations. A purchased lignite material was used to represent a material with low BC content.

Hypy successfully determined BC/TC% in materials that contain low, moderate and high BC contents. As expected the BC/TC% of NIST 2975 Diesel PM was around 100 %BC/TC, for the lignite material (BZ) it was around 2 %BC/TC. The biomass materials showed variability with BT/TC% of 16 % (DB) and 63 % (EB). The urban aerosols had similar BC/TC% (25 – 28 %). The road dusts had different BC/TC% due to different sampling site characteristics (14 – 41 %). The variability between replicates was extremely low with a maximum standard deviation of 5.3 %BC/TC and standard error of 1.3 %BC/TC across all materials. The BC/TC% for NIST 1649b (28 %BC/TC) was in agreement with other hypy-derived BC/TC% reported in the literature. A method was then developed for the use of hypy on airborne ambient PM for subsequent 14C analysis for source apportionment.

Airborne ambient PM from two sites in Glasgow (Glasgow City Centre GLA-CC, and Glasgow Hillhead GLA-HH) and Fallowfield Manchester (MAN-FF) underwent 14C analysis for both TC and BC components. The results for the percentage modern carbon (%MC) for the TC were compared to other UK sites. The MC% to percentage fossil carbon (%FC), (%MC:%FC), proportions in TC sample for GLA-CC and GLA-HH were 46%:54% showing dominant fossil sources. Whilst for MAN-FF the %MC:%FC proportion was 63%:37%, hence exhibiting dominant modern carbon sources. Due to COVID-19 comparisons with other UK sites did not represent source apportionments for business-as-usual scenarios. The %MC (TC) followed this order:

Auchencorth Moss > London Honor Oak Park > Chilbolton Observatory > Manchester Fallowfield > Glasgow Hillhead > Glasgow City Centre.

Auchencorth Moss, a rural background site, had the highest proportion of modern carbon sources and Glasgow City Centre had the highest proportion of fossil carbon sources.

The PAH and n-alkane and PAH sources were in agreement with those %MC (TC) for GLA-CC and GLA-HH as fossil fuel combustion sources were dominant sources according to source categorisations from PAH diagnostic ratios (PAH DRs) and carbon preference index (CPI) for n-alkanes. Principal component analysis showed that the variance in PAH and n-alkane concentrations could be explained by two principal components (PC), PC 1 (79.6 % of the variance) and PC 2 (20.1 % of the variance). These components represented diesel and wood combustion emissions (PC 1) and vehicular emissions (direct and indirect) (PC 2). The n-alkane distributions showed the presence of epicuticular wax n-alkanes (C29, C31 and C33) suggesting biogenic vegetative detritus sources in GLA-KS samples, with the possibility of regionally transported biomass combustion PM-bound PAHs. The n-alkanes showing odd/even carbon preference in MAN-FF differed from GLA-KS, suggesting different sources of epicuticular waxes.

BCₕᵧₚᵧ was successfully isolated for urban aerosols and dust materials. The NIST 1649b %MC (BCₕᵧₚᵧ) of 6 %MC was in agreement with previous studies. BCₕᵧₚᵧ was then isolated for MAN-FF and GLA-HH samples. The reported %MC (BCₕᵧₚᵧ) was 24 % for GLA-HH and 20 – 25 % for MAN-FF. The two replicates of MAN-FF showed a good level of precision both yielding 25 %MC (BCₕᵧₚᵧ). The sources of carbonaceous components from GLA-HH and MAN-FF therefore dominantly fossil carbon. However, the ¹⁴C was higher than urban aerosol materials (6 – 12 %MC) and reported values in the literature for a UK site, Birmingham (11 % average %MC). This shows that the modern nature of BC has increased due to decreasing fossil fuel combustion emissions and an increase in biomass combustion sources.

• Item Type: Thesis (PhD)
• Qualification Level: Doctoral
• Subjects: G Geography. Anthropology. Recreation > GE Environmental Sciences
• Colleges/Schools: College of Science and Engineering > Scottish Universities Environmental Research Centre
• Supervisor's Name: Ascough, Dr. Phillipa, Heal, Professor Mathew and Toney, Professor Jaime
• Date of Award: 2023
• Depositing User: Theses Team
• Unique ID: glathesis:2023-84003
• Copyright: Copyright of this thesis is held by the author.
• Date Deposited: 20 Dec 2023 11:32
• Last Modified: 20 Dec 2023 11:38
• Thesis DOI: 10.5525/gla.thesis.84003
• URI: https://theses.gla.ac.uk/id/eprint/84003