The heterogeneous interaction of trace gases on mineral dust and soot

The present thesis work deals with the investigation of the heterogeneous reactions involving nitrate radical (NO3), dinitrogen pentoxide (N2O5) and ozone (O3) on surrogates of atmospheric mineral dust particles characteristic of the troposphere. An additional investigation of heterogeneous reaction of NO3 on flame soot was carried out. The goal is to characterize the kinetics (the uptake coefficient γ) as well as the reaction products. The obtained results are intended to provide reliable data for numerical modelling studies. The experiments were performed in a very low pressure flow reactor (Knudsen cell reactor), coupled to mass spectrometry (MS) and optical probing (Resonance Enhanced Multiphoton Ionization (REMPI)). The used mineral dust powder samples were: Kaolinite, CaCO3, natural limestone, Saharan Dust and Arizona Test Dust. Two different types of soot were produced: soot originating from a rich decane flame at a high fuel/oxygen ratio ("grey" soot) and soot generated from a lean flame at a low fuel/oxygen ratio ("black" soot). Uptake experiments of NO3 on mineral dust were carried out under continuous molecular flow conditions (steady state) at 298 ± 2 K using the thermal decomposition of N2O5 as a NO3 source. In situ laser detection (REMPI) was employed in addition to beam-sampling electron-impact mass spectrometry in order to specifically detect NO2 and NO in the presence of N2O5, NO3 and HNO3. We found a steady state uptake coefficient γss ranging from (3.4 ± 1.6) × 10-2 for natural limestone to 0.12 ± 0.08 for Saharan Dust with γss decreasing as [NO3] increased. NO3 adsorbed on mineral dust led to uptake of NO2 in an Eley-Rideal mechanism where usually no uptake is observed in the absence of NO3. The disappearance of NO3 was in part accompanied by the formation of N2O5 and HNO3 in the presence of NO2. NO3 uptake performed on small amounts of Kaolinite and CaCO3 led to formation of some N2O5 according to NO3(ads) + NO2(g) –> N2O5(ads) –> N2O5(g). Slow formation of gas phase HNO3 on Kaolinite, CaCO3, Arizona Test Dust and natural limestone has also been observed and is clearly related to the presence of adsorbed water involved in the heterogeneous hydrolysis of N2O5(ads). Uptake of N2O5 on mineral dust samples led to γss values ranging from (3.5 ± 1.1) × 10-2 for CaCO3 to 0.20 ± 0.05 for Saharan Dust with γss decreasing as [N2O5]0 increased. We have observed delayed production of HNO3 upon uptake of N2O5 for every investigated sample owing to hydrolysis of N2O5 with surface-adsorbed H2O. At high and low [N2O5] Arizona Test Dust and Kaolinite turned out to be the samples to produce the largest amount of gas phase HNO3 with respect to N2O5 taken up. In contrast, the yield of HNO3 for Saharan Dust and CaCO3 is lower. On CaCO3 the disappearance of N2O5 was also accompanied by the formation of CO2. For CaCO3 sample masses ranging from 0.33 to 2.0 g, the yield of CO2 was approximately 42 – 50% with respect to the total number of N2O5 molecules taken up. The reaction of N2O5 with mineral dust and the subsequent production of gas phase HNO3 leads to a decrease in [NOx] which may have a significant effect on global ozone decrease. The rate of uptake of ozone on various mineral dust substrates was very similar for all the examined substrates. Both initial and steady state uptake coefficients γ0 and γss were found to be similar for all examined substrates. Uptake experiments on cut marble samples have shown that γ0 and γss based on the geometric and total internal (BET) surface area may be over and underestimated between a factor of 50 to 100, respectively. Based on these considerations we proposed initial and steady state uptake coefficients of the order of 10-4 and 10-5, respectively. For all uptake experiments on mineral dust surrogates, the disappearance of O3 was accompanied by formation of O2. The different mineral dust surrogates may be more accurately distinguished by their time-dependent relative O2 yield rather than the magnitude of γ. The heterogeneous reaction of O3 on mineral dust has been found to be non-catalytic and of limited importance in the atmosphere. Uptake experiments of NO3 on decane flame soot led to a large steady state uptake coefficient γss of 0.2 ± 0.02 for grey and black soot with γss decreasing as [NO3] increased. Adsorbed NO3 led to an uptake of NO2 admitted from the hot NO3 source. On large quantities of grey soot we observed production of HONO (nitrous acid) corresponding to almost 100% of NO2 taken up, whereas no HONO was formed on black soot. The disappearance of NO3 was in part accompanied by the formation of N2O5 according to reaction: NO3(ads) + NO2(ads) –> N2O5(ads) –> N2O5(g) probably due to the presence of adsorbed NO3 on the substrate. Subsequently, hydrolysis of N2O5(ads) with adsorbed H2O led to production of gas phase HNO3. For both grey and black soot we observed production of NO which did not depend of the amount of soot and [NO3]. Decomposition of NO3 and HONO on the soot substrates has been proposed to be responsible of gas phase NO formation.

A laboratory study of heterogeneous reactions relevant to the atmospheric boundary layer

Dominik Stadler

The present work deals with two subjects. The interaction of NO2 and HONO with different types of soot are examined in the first part whereas in the second part an experimental set-up is presented which has been built in order to measure the kinetics of the degradation of organic compounds by OH radicals. Both soot particles as well as NO2 are mainly produced by fossil fuel and biomass burning. The two species are therefore ubiquitous in the atmospheric boundary layer where they may react with each other. It has been shown that one possible product of this heterogeneous interaction is nitrous acid (HONO). HONO is an important trace gas in atmospheric chemistry because it is easily photolysed resulting in OH radical and NO. Thus, the photolysis of HONO significantly enhances photooxidation processes early in the morning. In the present study two different types of decane soot have been produced. It has been shown that the air/fuel ratio of the diffusion flame is a key parameter influencing the reactivity of soot towards NO2. Whereas soot from a rich flame ("grey" decane soot) leads to HONO with yields up to 100% upon interaction with NO2, only small amounts of HONO, but significant amounts of NO are formed in the presence of soot originating from a lean flame ("black" decane soot). A reaction mechanism has been developed showing that NO2 is initially converted to HONO by a redox reaction. HONO produced in this way is subsequently either quantitatively desorbed into the gas phase on "grey" decane soot or it decomposes to a large extent in a disproportionation reaction on "black" decane soot. The products of disproportionation are NO, which is instantaneously desorbed into the gas phase and NO2, which undergoes secondary reactions. In addition to the HONO producing pathway there is a reaction channel that leads to adsorption of NO2 for both types of soot which is irreversible on the time scale of our standard experiments. Saturation effects that are occurring on the time scale of our experiments are primarily caused by saturation of adsorption sites and not by the depletion of the reducing agent. It seems that the fraction of NO2 that is irreversibly adsorbed is blocking part of the surface sites where NO2 is initially adsorbed. Uptake coefficients γ take initial values of up to 0.1 for both types of soot. However, with increasing uptake of NO2 γ is decreasing. After a NO2 consumption of approximately 8×1013 molecule cm-2, which corresponds to 13% of a formal monolayer, γ drops to values of 3×10-7 for "grey" decane soot and 6×10-7 for "black" decane soot, respectively. Furthermore, our results show that the rate of initial uptake is the rate limiting step of the NO2/soot interaction mechanism. Experiments where "black" decane soot has been exposed to a flow of HONO revealed that it readily decomposes into NO and NO2. For HONO concentrations above 3.3×1011 molecule cm-3 a total product yield of 50%, whereof 40% NO and 10% NO2, has been observed. The initial uptake coefficient ( γ0) for HONO on "black" decane soot did not show clear saturation effects but varied randomly in the concentration range from 1.2×1012 to 5.9×1012 molecule cm-3. The average value measured at these concentrations was equal to 0.027. The strong interaction of HONO with "black" decane soot shows that soot is not only a potential source, but also a potential sink of HONO. On the other hand, no significant uptake could be observed when "grey" decane soot was exposed to HONO. This is consistent with the observation that NO2 is converted to HONO at yields of up to 100%. Acetylene soot is an additional type of soot that has been examined. The results for the initial uptake coefficients were basically the same as for "grey" and "black" decane soot, that is maximum values of γ0 of 0.1 for NO2 uptake decreasing with increasing concentrations of NO2. This similarity holds only for the first few seconds of interaction because acetylene soot saturates much faster than "grey" and "black" decane soot with continuous exposure. The acetylene samples are generally almost completely saturated after only three minutes of exposure to NO2. Soxhlet extractions of "grey" decane soot which have been performed using tetrahydrofuran as a solvent also produced HONO upon interaction with NO2. This shows that the active reactants are not part of the graphitic soot backbone but are extractable, rather polar compounds of the organic fraction of soot. This assumption is supported by the observation that the corresponding benzene extraction is not reactive in relation to NO2 uptake and HONO formation. In the second part of the present work an experimental set-up is described which has been built in order to measure the rate constants of the reaction of organic compounds with the OH radical in the gas phase using a relative rate technique at atmospheric pressure. The system may be called a "mini smog chamber" as the reaction cell has a volume of only 480 cm3. The system was validated by measuring the relative rate constants of the reactant pair cyclohexane/toluene. These measurements have been performed for temperatures ranging from 300 to 360 K and led to results which are in excellent agreement with literature data. However, the experimental set-up also had some important limitations. Biphenyl which has a boiling point of 529 K was the compound with the lowest volatility that could be examined in our reaction cell. Compounds with lower vapor pressures or compounds with polar groups adsorbed to a large extent onto the glass walls of the reaction cell even at temperatures of up to 373 K. Additional problems occurred in the closed ion source of the mass spectrometer leading to non linear and/or unstable MS signals. These phenomena were most probably caused by secondary ion-molecule reactions in the closed ion source.

EPFL2000

A laboratory study of heterogeneous reactions relevant to the atmospheric boundary layer

Dominik Stadler

EPFL2000