Secondary inorganic aerosols (SIA), comprising of sulfate (SO₄^2-), nitrate (NO₃^-), and ammonium (NH₄⁺), are major constituents of fine particle (PM_2.5, particle matter with aerodynamic diameter smaller than 2.5 μm). They play important roles in PM-related adverse impacts on human health, acid deposition, visibility degradation and climate system. Past studies on SIA in Hong Kong (HK) or the Pearl River Delta (PRD) mostly rely on filter-based measurements which typically need 12 h or longer to collect enough material for off-line analysis. Work in this thesis focuses on half-hourly or hourly measurements of PM_2.5 SIA in two locations in HK using a continuous system, PILS (Particle-into-Liquid System) coupled to two ion chromatographs. The high-resolution data sets allow the examination o...[ Read more ]

Secondary inorganic aerosols (SIA), comprising of sulfate (SO₄^2-), nitrate (NO₃^-), and ammonium (NH₄⁺), are major constituents of fine particle (PM_2.5, particle matter with aerodynamic diameter smaller than 2.5 μm). They play important roles in PM-related adverse impacts on human health, acid deposition, visibility degradation and climate system. Past studies on SIA in Hong Kong (HK) or the Pearl River Delta (PRD) mostly rely on filter-based measurements which typically need 12 h or longer to collect enough material for off-line analysis. Work in this thesis focuses on half-hourly or hourly measurements of PM_2.5 SIA in two locations in HK using a continuous system, PILS (Particle-into-Liquid System) coupled to two ion chromatographs. The high-resolution data sets allow the examination of SIA temporal dynamics in the scale of hours that the filter-based approach is incapable of providing. In addition, the PILS measurements of semi-volatile SIA species, i.e., NO₃^- and NH₄⁺, are more reliable as the filter-based sampling is subjected to various sampling artifacts due to gas-particle interactions and/or evaporation of semi-volatile species. The high-time resolution PM_2.5 ionic chemical composition data have been explored and the main findings are summarized below:

(1) Impacts of local emissions, regional transports and their interactions on chemical composition and concentrations of PM_2.5 SIA and other ionic species were investigated at the Hong Kong University of Science and Technology (HKUST), a receptor site, under three synoptic conditions. It is found air masses that passed the heavy urban areas located on the HK Island most likely attributed a night time NO₃^- episode in fall. Low temperature and high relative humidity (RH) that was associated with an approaching cold front efficiently shifted gaseous phase HNO₃ to aerosol phase. Regional transport from north PRD was likely responsible for elevated NH₄⁺ and SO₄^2- levels in a SO₄^2- episode in fall. During a cold front event, the variation of surface PM_2.5 and its SIA components was closely related to wind regime. That is, at the beginning of the cold front passage, the strong wind efficiently diluted the boundary layer pollutants, while as the wind weakened (in ~2 days), elevated pollutants levels were observed.

(2) Chemical compositions and size characteristics of ionic species were investigated at Tung Chung, a new town area located in the Southwest part of HK. The sampling period was from 17 to 26 December 2009, covering both normal conditions and an aerosol episode. The three major secondary inorganic ions, SO₄², NH₄⁺ and NO₃^-, accounted for 47 ± 6% of PM_2.5 mass. Nitrate in PM_2.5 had an average concentration of 4.5 μg m^-3, ranging from 0.8 to 40.5 μg m^-3. During the episode hours (defined to be periods during which the air pollution index exceeding 100), nitrate explained 37% of PM_2.5 mass and its concentration exceeded that of SO₄^2- to become the most abundant PM_2.5 species. Variable size distributions of NO₃^- were observed, with three sets of samples showing a dominant droplet mode, two sets showing a dominant coarse mode, and the rest four sets bimodally distributed in both droplet and coarse modes. During the episode, 84% of NO₃^- was distributed in fine mode particles, significantly contributing to the elevated PM_2.5 level. Further examination of size characteristics of NO₃^- shows that fine mode NO₃^- is more likely to occur in environments when the fine particles are less acidic and the sea-salt aerosol contributions are low.

(3) The ionic chemical composition of PM_2.5 and meteorological parameters (e.g., temperature, RH) obtained at the HKUST site under all three different synoptic conditions are input into Aerosol Inorganic Model (AIM-III) for estimation of in situ pH through calculation of H⁺ amount and aerosol liquid water content (LWC). The particle pH_IS ranged from -1.87 to 3.12, with an average at -0.03, indicating the PM_2.5 particles in Hong Kong are highly acidic. Unlike particle strong acidity, which was dominated by SO₄^2- concentration, the amount of aerosol liquid water content could significantly influence in situ particle acidity. Principal factor analysis has identified the equivalent concentration ratio between cations and anions (i.e., R_+/-) and RH to be the two most important factors influencing the particle pH_IS. pH_IS under different synoptic conditions in this study could be well approximated by a single linear regression equation (slope: 0.95, R²: 0.93), i.e., pH_IS = 4.94 R_+/- + 3.11 RH - 5.70. Such an empirical equation provides a convenient mean in estimating particle in-situ acidity for assessing the role of acid-catalyzed aerosol reactions.

The second part of this thesis work is to improve an observation-based model (OBAMAP) for SIA, which was first developed by Dr. Zibing Yuan (2006) to evaluate the sensitivity of formation of nitrate ad sulfate to changes in the emissions of their precursors (i.e., NO_x, SO₂, and VOCs). The improvement work includes incorporating updated chemical mechanisms, thermodynamic equilibrium for gas-aerosol phase apportionment and size distribution of SIA. While NO_x and SO₂ are recognized precursor gases to nitrate and sulfate, levels of volatile organic compounds (VOCs) also have impact on the production of SIA through mediating the level of oxidants. In the model, a sequence of present time-frame observations of precursors and particle compositions and meteorological data are used to drive the simulation and to determine responses to perturbed emission rates. The use of observational data ensures that the calculations are carried out for the proper combinations of NO_x and VOCs conditions.

The new OBM for SIA is applied to hourly gaseous and particulate composition data measured during a wintertime pollution episode encountered in Tung Chung for probing effectiveness of different precursor control strategies. The OBM analysis suggests that local formation of PM_2.5 SIA was strongly enhanced and contributed 78% of NO₃^- increase and 14% of SO₄^2- increase in the episode hours. The NO₃^- level is found to be sensitive to the change in NO_x and anthropogenic VOCs (AVOCs) emissions. Each percent reduction (near 10% reduction rate) in NO_x and AVOCs results in a reduction of 0.53% and 0.66% in NO₃^- mass formation, respectively. However, the significance of NO_x and AVOC control would be expected to decrease as more precursors are reduced. Reduction of SO₄^2- formation, on the other hand, is most sensitive to reduction of SO₂. At the measured mbient level, each percent reducing of SO₂ results in a reduction of 1.20% SO₄^2- mass formation and this value increases to 1.70% if the reduction of SO₂ increases to 60%, suggesting controlling of SO₂ would become more effective as more SO₂ is reduced. The OBM is demonstrated to provide a relatively simple and cost-effective tool for analyzing the increasing database of high time resolution measurements of VOCs and major aerosol ionic species.

The PILS measurements and the modeling framework are also used to investigate different sulfate formation regimes in respective of relevant precursor abundance. Past studies have shown formation of SO₄^2- could be severely inhibited under high NO_x conditions because oxidative potential of the atmosphere is greatly diminished through the reaction of NO₂+OH. Such a theory fails to explain high SO₄^2- loadings under low solar actinic fluxes conditions observed in mega cities in China. In this thesis, we propose a new production regime of SO₄^2- in which oxidation of S_(IV) is dominated by NO₂ and O₃ in the aqueous phase. Simulated with a simplified version of OBAMAP, it is shown elevation of NO_x favors productions of SO₄^2- in this regime, especially under high-SO₂ conditions. We then study the importance of NO₂-derived and O₃-derived SO₄^2- during haze episodes in PRD and during winter at urban/suburban locations in PRD. Our findings reveal these two pathways account for >70% of SO₄^2- productions. Since production of NO₂-derived SO₄^2- is independent on solar actinic fluxes while production by other pathways is, NO₂-derived SO₄^2- plays a more important role under low solar actinic fluxes conditions, even during the night time. In addition, it is noted that high levels of NO₂-derived SO₄^2- can only be expected under high-SO₂ conditions (like in PRD) because level of atmospheric SO₂ is the limiting parameter.

Results in this thesis contribute to improving our understanding of the complexity of chemical processes that SIA and their precursors are involved in atmospheric environments typical of rapid-developing regions. In addition, the OBM analysis provides valuable information for formulation of effective control strategies of SIA precursors.