Heterogeneous reactions involving HO2 radicals and atmospheric aerosols

The hydroxyl radical (OH), along with the closely coupled species the hydroperoxyl radical (HO2), have a profound effect on the concentration and distribution of most trace atmospheric species associated with climate change and poor air quality as they are essential to the daytime oxidizing capacity of the atmosphere. Tropospheric and mesospheric models that only consider gas-phase chemistry often over predict concentrations of HO2 indicating that heterogeneous reactions with aerosols could be a possible sink. In order to investigate the kinetics of such reactions, the HO2 uptake coefficient (γ(HO2), i.e. the probability that HO2 will collide and react within or on the surface of an aerosol) has been measured onto a variety of aerosols applicable to the troposphere, stratosphere and mesosphere using an aerosol flow tube experiment coupled to a highly sensitive HO2 detector, known as Fluorescence Assay by Gas Expansion (FAGE), and a Scanning Mobility Particle Sizer (SMPS). Deliquesced inorganic aerosols containing transition metal ions (TMI) have been shown to react rapidly with HO2. Measurements of γ(HO2) onto deliquesced inorganic aerosols doped with different concentrations of Cu(II), Fe(II), Mn(II), mixtures of Cu(II) and Fe(II) and I- are presented within this work. HO2 uptake onto deliquesced inorganic aerosols doped with transition metal ions may not be as significant as previously thought. The Thornton expression, used in global modelling studies of HO2 uptake, can predict γ(HO2) at a relative humidity (RH) of 65%, however at a RH of 43%, near the efflorescence point ((NH4)2SO4 = 37%), good agreement was only observed at higher concentrations of Cu(II) and Fe(II) (> 0.1 M) possibly indicating that HO2 solubility decreases as HO2 diffuses further into the bulk of the aerosol. It was expected that as deliquesced NaCl aerosols have a higher pH (7) that most HO2 accommodated within the aerosols will dissociated to the more reactive species O2-. This should result in high values of γ(HO2), however γ(HO2) onto Cu(II)-doped NaCl aerosols was measured to be lower than γ(HO2) onto Cu(II)-doped (NH4)2SO4 aerosols with a lower pH, possibly due to the formation of [Cu(Cl)4]2- complexes which are repelled into the bulk of the aerosol by enhanced concentrations of Cl- ions within the interfacial layer. Measurements of γ(HO2) onto Fe(II)-doped NaCl aerosols were relatively high and agreed with predictions made by the Thornton expression. When irradiated with UVA light, γ(HO2) onto Cu(II)-doped (NH4)2SO4 was lowered, however γ(HO2) onto Cu(II)-doped NaCl remained the same. When the effect of irradiating Fe(II)-doped (NH4)2SO4 aerosols on γ(HO2) was investigated, results indicated possible production of OH. Measurements of γ(HO2) onto mixed Cu(II) and Fe(II)-doped (NH4)2SO4 aerosols could not verify the Mao hypothesis that an electron transfer reaction occurs between Cu(I) and Fe(III) resulting in the conversion of HO2 to H2O, rather than H2O2. However, values of γ(HO2) did not simply equal the sum of γ(HO2) onto Cu or Fe-doped aerosols individually, indicating that the presence of both TMI in the aerosol does alter the chemistry of the aerosol in some way. Irradiation of Cu(II) and Fe(II)-doped aerosols resulted in an enhancement of γ(HO2), possibly indicating an alternative mechanism than that proposed by Mao, where HO2 is converted to H2O via a photochemical mechanism. The presence of I- within NaCl aerosols does not result in a change of γ(HO2), however when converted to I2 by reaction with Cu(II) an enhancement of γ(HO2) greater than when doped with Cu(II) alone was observed Measurements of γ(HO2) onto TiO2, a possible candidate aerosol for use within solar-radiation management (SRM) schemes, showed a positive dependence on Relative Humidity (RH) which correlated with the number of monolayers of water adsorbed onto the TiO2 nanoparticle. This dependence suggests a mechanism by which HO2 adsorbs to the surface of the TiO2 particle by forming complexes with water molecules bound to bridging OH groups. The TOMCAT chemical transport model was used by Professor Chipperfield to predict the possible effects of HO2 uptake (using an upper limit of γ(HO2) = 1) onto the surface of TiO2 nanoparticles on the stratospheric concentrations of HO2 and O3. The amount of TiO2 used was chosen to achieve a similar cooling to that following the Mt. Pinatubo eruption, but the model predicted a very small loss of both stratospheric HO2 and O3. Upon illumination of airborne TiO2 nanoparticles with UV light, significant quantities of HO2 was formed within the gas-phase, thought to be the first direct observations of radicals emitted from the surface of airborne particles. The reaction is dependent on the presence of gas-phase O2 and H2O within the system. The production of HO2 was shown to slow down as a function of time irradiated pointing towards a photochemical aging process occurring on the surface of the TiO2 particles. The dependence of HO2 production on O2 and H2O concentrations was determined, which shows a typical Langmuir adsorption saturation curve for O2 suggesting it is the gas-phase reactant in this process. The addition of H2O into the system inhibits the reaction and reduces the adsorption equilibrium coefficient for both species. Reduction of O2 by photogenerated electrons is likely to be the initial step in this process followed by reaction with a proton. Hydrogen extraction from hydroxyl bridging groups (OHbr) groups by O2- could explain the slow down observed in the rate of HO2 production. Production of gas-phase OH radicals was investigated and showed OH was produced only when large concentrations of TiO2 aerosols entered the aerosol flow tube, probably associated with the decomposition of H2O2 formed from reactive uptake. Although the production of HO2 by TiO2 aerosols initially would not be advantageous for its use within SRM schemes, the reaction ceases upon prolonged photocatalytic aging of the aerosol surface. Meteoric smoke particles (MSP) provide the only significant surfaces within the mesosphere for heterogeneous reactions to occur. To investigate whether such reactions could, in some part, be responsible for the ‘HOx Dilemma’ measurements of γ(HO2) onto analogues of MSP, forsterite, olivine and fayalite, were conducted at a RH of 10%. These experiments showed forsterite to have the lowest reactivity with HO2, similar to that of effloresced inorganic aerosols, and olivine and fayalite to have a similar reactivity that was more than an order of magnitude greater than that of forsterite, thus demonstrating that the presence of Fe within the MSP is required for significant reactivity with HO2. Electronic structure calculations, conducted by Professor Plane, predicts that the difference in reactivity is associated mechanistic and energetic differences between the binding of HO2 to Fe and Mg sites, however, the positive dependence of γ(HO2) with RH and similar values of γ(HO2) for olivine and fayalite suggests that OH bridging groups or complexing with water molecules adsorbed to such sites, as with TiO2 nanoparticles, are adsorption sites for HO2. Taking the measurements made in this work and the likely dependence of γ(HO2) on temperature and RH, a value of 0.2 for γ(HO2) was applied by Dr Sandy James in WACCM-CARMA. This modelling study predicted reductions in the HO2 volume mixing ratio of up to 40% in the polar vortex. Impact to HO2 budgets in the mesosphere was found to be dependent on latitude, giving agreement with the presence of MSPs in the polar night.