Coal dust is inextricably linked to the development of dust diseases. To date, the role of mineral matter in coal has been investigated for its links to pulmonary damage; however, no consensus has been reached on which characteristics are relevant to pulmonary toxicity. This study hypothesises that the toxic potential of inhalable coal dust can be attributed to reactive mineralogy and the specific surface area for interaction between the particles and primary phagocytes such as macrophages. To test this hypothesis, the study developed an advanced understanding of the relationship between the physicochemical and mineralogical characteristics of coal particles and pulmonary toxicity. Three objectives were constructed to achieve this aim. Objective 1 developed a detailed particle characterisation dataset on coal particle samples utilising both routine (X-ray diffraction and X-ray fluorescence) and advanced methods of coal analysis (automated scanning electron microscopy systems). Objective 2 elucidated multivariant relationships between the particle characteristics and the immunological responses from exposed macrophage cells in vitro using advanced statistical methods. Lastly, objective 3 developed a protocol to empirically characterise the relative risk of coal dust-related damage on a cellular level. In developing a detailed characterisation dataset on the coal samples, both routine and automated analysis tools were used to define general, chemical, mineralogical, and mineral specific characteristics. An auto-SEM-EDS-XRD (Automated scanning electron microscope coupled with Energy Dispersive X-ray Spectroscopy and analyses generated by X-ray Diffraction) protocol was developed to obtain a broad spectrum of particle data by mineralogically mapping each particle. This protocol involved the rigorous analysis of uncertainty in the data using comparative datasets generated from XRD and XRF (X-ray Fluorescence) analyses. In summary, the study demonstrated that the combined use of both routine and advanced particle analysis tools allowed for the classification of chemical and mineralogical distributions as well as a discrimination between general and mineral specific particle characteristics. Generally, these results suggested that features relating to general particle characteristics (size, shape, roughness, and surface area) are more strongly a function of mechanical breakage and deformation than compositional variation. To assess the multivariant relationships between the numerous characteristics defined and response measures of cellular toxicity, a PLSR (partial least squares regression) was applied in a novel approach to attempt a single model comparison of such relationships. This model was chosen for its ability to relate response and explanatory variables based on a new set of variables which have undergone dimensionality reduction whilst maximising the covariance. The results from the relationship analysis showed that physical characteristics (particle shape in particular) displayed a greater influence on cytotoxicity and lipid peroxidation over mineral and chemical-based characteristics. Relating this observation to previous research it was suggested that the influence of shape and roughness on phagocytosis may have strong implications for magnitude of direct and indirect cellular harm and the predominance of either intracellular or extracellular damage. The results also showed that, apart from the influence of particle shape, radical-induced stress and cytotoxicity displayed a strong dependency on (1) the chemical and mineralogical reactivity Ca hosted in gypsum, (2) the release/inhibition of Fe from pyrite and Fe-sulfates, and (3) the surface activity of quartz based on its crystallite size. However, the relationships defined in the context of cytotoxicity displayed a more nuanced dependency with the silicate mineral content and their associated properties compared to lipid peroxidation. From this it was suggested that non-radical related pathways to cytotoxicity could also occur from coal dust exposure. Ultimately, the study demonstrates the first analysis which assesses relative impact and magnitude of multiple particle characteristics on cytotoxicity and cellular stress. Finally, to provide a more easily interpretable format for the analysis of the PLSR relationships, a protocol was developed to screen variables based on: (1) their level of importance to the defined relationship and (2) the rank of importance for each influential variable represented on a unified scale. Elements which explained the variability within the sample characteristics and the responses were clustered using the k-means algorithm to determine classes of samples which display similar characteristics or levels of toxicity. The comparison of the classes grouping samples with similar properties versus samples groups with similar toxicity levels showed that even though samples may share similar properties, their reported level of toxicity may differ. This confirms the observations from previous studies which have shown that the relative toxicity of coal dust cannot be explained on the basis of isolated properties. Rather the set of ‘influential variables' showed that a combination of general, chemical, mineralogical and mineral specific data are needed to determine the differences between levels of toxicity. Ultimately, the application of this protocol on 17 different dust-sized coal samples demonstrated the key differences between samples and their influence on levels of cytotoxicity and lipid peroxidation, which until this study have not been demonstrated by a single regression. As an outcome of such results, this study provides a robust analysis strategy for elucidating particle cell relations which can further advance the understanding of coal dust induced disease pathology. Additionally, the protocol has demonstrated the usefulness of disseminating the complex data structures to more easily interpretable data formats such that a generalisable analysis of risk factors related to coal dust-based cellular damage can be utilised by stakeholders in data-based decision making. Ultimately, the results of this study propose that the toxic potential of coal dust is primarily a function of the reactive mineralogical and chemical components within the particles, however, the magnitude of this intrinsic reactivity is subject to the mitigative factors which can either neutralise of supress the anticipated reactivity.
| Identifer | oai:union.ndltd.org:netd.ac.za/oai:union.ndltd.org:uct/oai:localhost:11427/38502 |
| Date | 11 September 2023 |
| Creators | Kamanzi, Kankindi Conchita, von Holdt, Johanna, Jacobs, Muazzam |
| Contributors | Broadhurst, Jennifer, Becker Megan |
| Publisher | Faculty of Engineering and the Built Environment, Department of Chemical Engineering |
| Source Sets | South African National ETD Portal |
| Language | English |
| Detected Language | English |
| Type | Doctoral Thesis, Doctoral, PhD |
| Format | application/pdf |
Page generated in 0.002 seconds