Abundance and sources of organic nitrogen in fine (PM2.5) and coarse (PM2.5-10) particulate matter in urban Hong Kong.

Organic nitrogen (ON) in atmospheric particles is much less monitored compared to inorganic nitrogen (IN), despite its significant contribution to atmospheric N deposition budget. In this study, we expanded a newly developed instrumental method for IN and ON in PM2.5 samples to PM10 samples. We determined the quantities of ON and IN for paired PM2.5 and PM10 samples collected at an urban coastal site in Hong Kong, southern China over a year. These measurements also allowed the determination of IN and ON abundance in the coarse PM (i.e., PM2.5-10) by taking the difference between PM10 and PM2.5. The measurement results show that ON accounted for 27.6 % and 21.1 % of total N in fine and coarse particles, respectively, and was mainly (87.7 %) distributed in the fine mode at the site. The seasonal variation of ON/total N was relatively small in PM2.5 (23.6-30.4 %) while considerably larger in coarse PM (4.3-42.1 %). Analysis aided by concurrently measured source indicators revealed that sea spray, biological particle emissions, and dust mixed with anthropogenic pollutants are potentially significant sources of ON in coarse particles. Positive matrix factorization (PMF) source apportionment further revealed that industrial emissions/coal combustion (43.6 %), soil dust emission (16.3 %), fresh sea salt emission (15.2 %), and aged sea salt (24.9 %) are major sources of PMcoarse-bound ON at the site. The contributions of industrial emissions/coal combustion and soil dust emission to ON were significantly higher in autumn and winter. Fresh sea salt emissions contributed greater proportions to ON in spring and summer, while ON associated with the aged sea salt source was higher in spring and autumn. These findings have advanced our quantitative understanding of the sources of PMcoarse-bound ON, which was scarcely determined in the past. Furthermore, the ON measurement data in fine and coarse particles helps estimate ON deposition, which has been previously under-evaluated.