PM10 samples were collected simultaneously at three representative areas (urban, industrial, and rural areas). Their morphology and elemental composition were determined by scanning electron microscopy coupled with energy-dispersive analysis (SEM-EDS). Twenty-four chemical parameters (C, O, Na, Mg, Al, Si, P, Cd, Cl, K, Ca, S, Sn, Ti, V, Cr, Mn, Fe, Co, Ni, Cu, Zn, W, and Pb) were determined and three morphological parameters (area, roundness, and fractal dimension) were measured by Image Pro Analyzer 6.3. The particles were classified into ten groups based on morphology and elemental composition: Ca-rich and metal particles, soot aggregates, cenosphere, alumosilicates, sea salt, calcium sulfate, spherical particles of iron, biological carbonaceous particles, and various. Particles of natural origin were predominantly found in the coarse size fraction and particles of anthropogenic origin in the fine size fraction. The greatest contribution to particulate matter belonged to aluminum-silicates and calcium-rich particles. The cenosphere were recognized only in the coastal urban site, while all the other particles were present in each site. The coastal industrial site was characterized by the prevalence of alumosilicates and Ca-rich particles, due to construction activity in this site during the sampling period (movement of vehicles, transport of terrigenous materials, and use of construction products). The coastal urban site was characterized by a higher amount of soot and by the presence of cenosphere, due to the presence of vehicular traffic.
Keywords: Individual particle analysis; Particle shape; Particulate matter; Scanning electron microscopy.