Assessment of the PM2.5 oxidative potential in a coastal industrial city in Northern France: Relationships with chemical composition, local emissions and long range sources

Abstract The objective of this work was to relate PM2.5 Oxidative Potential (OP) data to PM composition and PM local and distant source contributions. PM2.5 collected in Dunkerque, a coastal industrial city in North of France, was extensively characterized for major and minor chemical species. PM2.5 filters were extracted using a synthetic pulmonary fluid to achieve OP estimation based on Ascorbic Acid (AA) and dithiothreitol (DTT) depletion assays. In order to evidence relationships between OP values, chemical composition and local and distant source contributions, correlation coefficient, Principal Component Analysis (PCA), concentration roses and polar plots and concentration weighted trajectories were used. Heterogeneous conclusions were drawn using the three first methods as the bivariate polar plots lead to dismiss some of the correlation evidenced using correlation coefficient and PCA. Both AA and DTT tests appeared complementary as they were not sensitive to the same species/source contribution. The bivariate polar plot representation of OP values versus wind direction and wind speed revealed that PM2.5 concentration and combustion sources were linked to OP-AA, whereas emissions from integrated steelworks, electric steelworks, heavy fuel oil combustion and traffic non-exhaust significantly contribute to OP-DTT. Sea-salts, aged sea-salts, crustal, secondary sulfates and secondary nitrates sources were not found to contribute to OP values. Constant weighted trajectories evidenced several source regions responsible for high OP values with Belgium, Germany and Netherland at the leader position. Contribution of inland regions appeared possibly related to the biomass and traffic related combustion while heavy fuel oil combustion could be also involved in the contribution of marine and coastal areas.