Cinétique de réaction hétérogène d'importance troposphérique sur des substrats modèles et sur des aérosols

The first part of this study consists in the investigation of the evaporation and the condensation rate of H2O in the presence of different types of ice in a low pressure flow reactor. The second part of the present work reports the study of chlorine activation on different types of frozen salt substrates in a low pressure flow reactor. The third part concerns the construction and testing of a laminar flow reactor at atmospheric pressure in order to study bromine and chlorine activation reactions on different types of sea salt and model aerosols at different relative humidity. The kinetics of H2O condensation (kc) and the evaporation flux (Jev) of H2O on ice was studied in the range 130 to 210 K using pulsed-valve and steady state admission techniques in a low pressure flow reactor. The uptake coefficient γ was measured for different types of ice, namely, Condensed (C), Bulk (B), Single Crystal (SC), Snow (S) and Cubic ice (K). The negative temperature dependence of γ for C, B, SC and S ice reveals a precursor-mediated adsorption/desorption process. The non-Arrhenius behavior of the rate of condensation kc manifests itself in a discontinuity in the range 170 to 190 K depending on the type of ice and is consistent with a precursor model of H2O condensation/evaporation. The average of the energy of sublimation Δ0S is (12.0 ± 1.4) kcal/mol for C, B, S, SC ice and is identical within experimental uncertainty between 136 and 210 K. The same is true for the entropy of sublimation ΔSS. In contrast, both γ and the evaporative flux Jev are significantly different for different ices. In the range 130 to 210 K, Jev of H2O ice was significantly smaller than the maximum theoretically allowed value. This corroborates γ values significantly smaller than unity in the specified T range. Based on the present kinetic parameters the time to complete evaporation of a small ice particle of radius 1 µm is approximately a factor of 5 larger than previously thought. The HOCl heterogeneous reactions on frozen solutions of sea salt (SS), frozen KCl or NaCl were studied in a low pressure flow reactor in order to measure the uptake coefficient γ and the reaction products. The HOCl sample always contained up to 25% of Cl2O that was studied apart in order to relate the Cl2 produced to the HOCl taken up. The main product formed on NSS frozen solution is gas-phase Cl2 that is sustained for at least one hour in contrast to KCl or NaCl frozen solution that produced a transient Cl2 flow that decreases after 100s. We found that approximately 30% of the HOCl taken up from the HOCl/Cl2O mixture reacts to produce Cl2 on NSS, RSS and KCl frozen solution at 200 K at a HOCl concentration larger than 1011 molecule cm-3. In contrast, HOCl concentrations of a few 1010 molecule cm-3 at 200 K failed to produce additional Cl2 in contrast to pure Cl2O uptake. A single Br2 burst event was also monitored when Cl2O or a HOCl/Cl2O mixture is taken up on fresh frozen NSS solution during the first uptake run at 200 K. Further Cl2O or HOCl/Cl2O uptake on the same sample will not lead to a Br2 pulse. Annealing of a previously exposed frozen NSS solution to 240 K and subsequently cooling did not lead to a Br2 burst event either. The HOBr and HOCl uptake coefficient γ on acidified submicron salt aerosol was measured in an atmospheric flow tube reactor. The interaction time of the trace gas with the aerosol was in the range 15 to 90 s and led to γ values ranging from 10-4 to 10-2. The acidity of the aerosol is essential in order to enable heterogeneous reactions on NaCl, recrystallized sea salt (RSS) and natural sea salt (NSS) aerosols for both HOCl and HOBr. Specifically, HOCl exclusively reacts on acidified NSS aerosols with a γ ranging from 1.8 × 10-3 to 0.4 × 10-3 at a relative humidity between 40 and 85%, respectively. In addition, γ(HOBr) is approximately 10-2 for relative humidity in the range 77 to 90% on NaCl and RSS acidified aerosol. Uptake experiments of HOBr on aqueous H2SO4 as well as on acidified NaCl, RSS or NSS aerosols were performed for rh ranging from 40 to 93%. γ of HOBr on acidified NSS reaches a maximum at rh = 77% and decreases significantly for higher rh in contrast to acidified NaCl and RSS aerosols. This difference in γ is attributed to the presence of an organic phase in NSS aerosols that forms an organic coating at high rh. In addition, preliminary uptake experiment of HNO3 was performed on NaCl and NSS aerosols using the atmospheric flow tube reactor. We were able to perform experiments only from 2.8 to 25% relative humidity. For a relative humidity larger than 25%, all gas phase HNO3 is taken up on the reactor wall and no experiments could be performed. γ of HNO3 in the range (2-3) × 10-2 for NaCl and NSS aerosol, respectively, at 2.8% relative humidity. At rh = 25%, γ(HNO3) is approximately 10-1 for both NaCl and NSS aerosols.

Chemistry and climate interactions

Luca Pozzoli

Atmospheric trace gases and aerosols and climate interact in many ways. A quantitative assessment of the influence of trace gases and aerosols on climate can only be achieved if the interactions and feedbacks among these three major components are accounted for. The goal of this thesis was to develop, evaluate and apply the ECHAM5-HAMMOZ chemistry-aerosol-climate model. The model includes fully interactive simulations of the NOx-Ox-hydrocarbons chemistry and of the aerosol chemistry and microphysics, the two simulations being embedded into the well established ECHAM5 climate model. In particular, on-line calculation of the photolysis rates (that accounts for aerosols and clouds) and heterogeneous reactions of trace gases on aerosols are accounted for in the model. In addition, the aerosol simulation provides a prognostic representation of the mixing state of the aerosol components, a feature that is particularly important to examine the effect of the SO2 uptake and subsequent sulfate formation on aerosols. A thorough model evaluation of the model was performed using observations from the Transport and Chemical Evolution over the Pacific (TRACE-P) aircraft campaign. We also used ground based measurements from the European network EMEP (SO2 and sulfate), from the north American network IMPROVE (sulfate, black carbon and organic carbon), and from AERONET (aerosol optical depth). The coupled model is able to reproduce fairly well many of the observations over the TRACE-P region, even though some improvements are still needed. For example, we find a mean absolute bias of 20 ppbv (30%) and 40 ppbv (13%) for the simulated O3 and CO, respectively. Sulfate concentrations are represented fairly well, for all the regions considered, while SO2 is overestimated by a factor of 2 in general. Black carbon concentrations are underestimated over the TRACE-P region (mean absolute bias of 80%), most probably because of too low emissions, but well reproduced over north America. The aerosol optical depths compare well with observations at many sites in general, both in terms of annual means and seasonal variations. We show the results from a series of sensitivity simulations which goals are to assess the impact of heterogeneous reactions, photolysis reactions and sulfur chemistry on the regional and global trace gas distribution, and aerosol distribution, composition and optical properties. We found that heterogeneous reactions result in a reduction of 7% in the global O3 burden, and by up to 15% in surface O3 over regions rich in mineral dust. OH burden decreases by 10%, NO and NO2 by 20% and 29%, respectively while CO burden increases by 7%. Our numbers fall in general within the range of previous studies. We find that the effect of aerosols through the modifications of photolysis rates do not affect significantly the trace gas distributions and global burdens, while previous studies suggested larger effect. Heterogeneous reactions reduce the global mean SO2 surface concentration by 14%, while the global burden remains unchanged. This reflects the effect of competing processes. In particular, the depletion in SO2 by direct uptake on sea salt and dust is counterbalanced by the decrease in SO2 oxidation that is associated with the decrease in OH. The balance between the two competing effects differs depending on the regions (SO2 increases by more than +10% over north Africa and Indian Ocean and decreases by up to -20% at high latitudes). These processes, in turn, affect the sulfate formation. We find in general an increase (by up to 4%) in sulfate burden over the oceans because of sea salt uptake of SO2, and a decrease by up to 6% over Sahara and India, where SO2 increases. Our global burden of SO2 and sulfate amount to 0.77 and 0.87 (Tg(S)), respectively, which fall in the range of previous estimates. We find that SO2 uptake on sea salt contributes to the production of 3.69 Tg(S) yr-1 of sulfate (5% of total sulfate production) in good agreement with recent observation-based estimates. Uptake of SO2 on dust only contributes to the production of 0.55 Tg(S) yr-1 (< 1%) of sulfate, but it is an important process as it is responsible for the coating of mineral dust particles. A total of 300 Tg yr-1 of dust are transferred from the insoluble to the soluble mixed modes (56% of total emissions) because of this later process. The total burden of black carbon does not change significantly because of the heterogeneous reactions, but it is redistributed between insoluble/soluble modes. The changes in the mixing state of aerosols produce significant regional variations in aerosol optical depth and absorption. The aerosol optical depth and aerosol absorption increase by 10 to 30% and by 15%, respectively, during winter in the main polluted regions of the north hemisphere because of enhancing absorption efficiency of black carbon mixed with sulfate. On the contrary aerosol absorption decreases by up to 20% over the Atlantic ocean because of enhancing black carbon removal by wet scavenging. ECHAM5-HAMMOZ is a fully coupled model which includes many of the trace gas-aerosol and chemistry-climate interactions, and is thus a tool that allows an integrated assessment of the impact of the major short-life species that contribute to climate change and air quality.

EPFL2007

Chemistry and climate interactions

Luca Pozzoli

EPFL2007