Assessing the Mobility of Organic Chemicals Across the Sothern African Development Community

Background.
The number and diversity of organic chemicals produced global is consistently increasing, as
are the number of substances that are being detected in freshwater resources. 1 This is a
cause of significant concern for water quality, and associated impacts on human health. The
current scale of chemical contaminants in drinking water is only partly known and some
parts of the world remain unsurveyed, primarily due to lack of local analytical
instrumentation which is required to accurately analyse water samples. The Southern
African Development Community (SADC) is one such area, despite the fact that more than
75% of its population are dependent on groundwater for various activities. 2,3
Over the last three years, we have initiated a collaborative project with research groups in
Botswana to begin sampling aquatic systems across the region to better understand the
prevalence of organic contaminants. 4,5
Project aim.
In this project, we will build on our initial work and perform an extended sampling campaign
and in-depth analysis to obtain a broad picture of the nature, prevalence and mobility of
organic chemicals across the wider SADC.
Key objectives.
Sampling campaigns will be undertaken across a selected areas of the SADC (we expect this
to be the Okavango delta region) to obtain an extended set of water samples, with respect
to location, and time of sampling (e.g. wet versus dry season). Advanced analytical
techniques will be applied to analyse the samples within the Centre of Excellence for Mass
Spectrometry at York. Machine Learning (ML) and Artificial Intelligence (AI) tools will be
applied to expand information obtained via non-targeted and semi-quantitative analysis. A
database of observations of organic contaminants will be compiled, that will be analysed
using AI tools to identify correlations between chemical observations and mobility.
Novelty.
By conducting suspect and non-targeted screening, this project will make no assumptions as
to the molecular components in the SADC water sampled, and will generate a raw data
library that can be data-mined for years to come. Data obtained will deliver the widest
possible information on the number and nature of organic chemicals present. New ML
approaches will be applied to help elucidate unknown molecules detected. Second-stage
measurements will then use targeted analysis (the current gold standard) to quantify
selected key contaminants of high concern (e.g. antibiotics, agricultural chemicals). The
current gold standard for analysis of complex mixtures (such as aquatic samples) involves

liquid chromatography coupled high resolution mass spectrometry. The data bank acquired
will be analysed using routine artificial intelligence methods to obtain correlations in
contaminants versus geography (water courses, population centres, etc). 6-9
References
1 Wang, et al. (2020). Environmental Science and
Technology, 54, 2575-2584. https://doi.org/10.1021/acs.est.9b06379.
2 Selwe, K.P. et al., Environmental Toxicology and Chemistry, 41, 382, (2022).
3 https://link.springer.com/article/10.1007/s10661-024-12613-2
4 Selwe, K. P. et al., Environmental Toxicology and Chemistry, 43, 1, (52-61), (2023).
5 Kgato P. Selwe, et al. Emerging Contaminants, 100337, (2024).
6 https://doi.org/10.1016/j.eehl.2022.06.001
7 https://link.springer.com/article/10.1007/s40572-022-00389-x
8 https://www.youtube.com/watch?v=M3AQTs5iN5g
9 https://www.sciencedirect.com/science/article/pii/S014765132301254X