NaCl Aerosol Particles

Jun 16, 2010 - loss of ozone was further analyzed in terms of a numeric model to explicitly ... in catalytic ozone destruction in the marine boundary ...
0 downloads 0 Views 1MB Size
Download PDF
J. Phys. Chem. A 2010, 114, 7085–7093

7085

Uptake of Ozone to Deliquesced KI and Mixed KI/NaCl Aerosol Particles Aure´lie Rouvie`re, Yulia Sosedova, and Markus Ammann* Laboratory of Radiochemistry and EnVironmental Chemistry, Paul Scherrer Institute, 5232 Villigen, Switzerland, Markus Ammann, OFLB 103, Paul Scherrer Institute, 5232 Villigen, Switzerland. ReceiVed: April 12, 2010; ReVised Manuscript ReceiVed: June 1, 2010

The kinetics of uptake of ozone to deliquesced potassium iodide (KI) aerosol particles has been investigated in an aerosol flow tube at 72-75% relative humidity, room temperature, and atmospheric pressure. The observed loss of ozone was further analyzed in terms of a numeric model to explicitly track the iodide concentration in the particles. This allowed retrieving a value Rb ) 0.6 ( 0.4 0.5 for the bulk accommodation coefficient (Rb). The second order rate constant in the bulk phase agreed with available literature (kb ) (1.0 ( 0.3) × 109 M-1 s-1) even for the high ionic strength conditions of the present experiments. As long as iodide remained in excess, the average uptake coefficient was γ ) (1.10 ( 0.20) × 10-2. Different experiments were performed where the iodide to chloride ratio, the ozone concentration, and the surface to volume ratio of particles were varied. In combination, the results obtained indicate that uptake was driven by fast bulk accommodation and reaction in the bulk for all conditions investigated. The results further suggest that ozone uptake is not limited by the bulk accommodation coefficient Rb under atmospheric conditions. 1. Introduction It is well established that iodine chemistry has an active role in the destruction of stratospheric and tropospheric ozone.1-4 The speciation of iodine in atmospheric aerosol is poorly understood, but atmospheric iodine chemistry studies have received increasing attention with respect to gas phase reaction kinetics due to its potential to initiate the reactive cycles involved in catalytic ozone destruction in the marine boundary layer.5-7 Some reactions with halogens can lead to the formation of new particles with potential consequences for cloud formation8-10 and hence for the radiative budget of the atmosphere. Iodine compounds play an important role in the chemistry of the marine boundary layer (MBL).11-13 Sea salt aerosols produced have a chemical composition similar to the seawater from which they are formed,14 which explains the presence of iodide in atmospheric aerosol samples.15 In addition, iodine compounds are often enriched in marine aerosol compared to their relative abundance in seawater.10 Dissolved inorganic iodine is known to be mainly present, in seawater, as iodide and iodate ions in varying proportion.15,16 Iodide is reported to be the dominant inorganic iodine species in samples from the North Sea coast.17 Biogenic emissions of iodine are mainly due to macroalgae species, where Laminaria digitata is known to be a strong accumulator of iodine and is also able to release and transform inorganic species.18-20 The photolysis of alkyl-iodides (CH3I, CH2I2, C3H7I), which are produced by metabolic processes and released from the sea to the atmosphere, and photolysis of molecular iodine lead to the production of reactive inorganic iodine species. Subsequent photochemistry and hydrolysis leads to the formation of iodide. Iodide provides a substantial aqueous sink to drive uptake of gases in contrast to other iodine derivatives,12 which provides a motivation to study the reactivity of iodide in the presence of * To whom correspondence should be addressed. Telephone: ++41 56 310 4049. Fax: ++41 56 310 4435. E-mail: [email protected].

ozone. Indeed, it was recently shown that iodide contributes significantly to the dry deposition of ozone to the ocean surface.2,21 Ozone is acting as an important oxidant in the gas and aqueous phases. Ozone reacts with a large range of inorganic ions in aqueous solution.22 The chemistry of chloride and bromide ions have been already studied in some detail.23 We know already that the reaction of ozone with iodide in seawater affects the global flux of iodine from the oceans.24 The reactive uptake of ozone on solid potassium iodide has been studied yielding a reactive uptake coefficient of γ ) (1.4 ( 0.7) × 10-4.25 However, marine aerosol over the oceans consists mainly of deliquesced liquid droplets under the prevailing humidity conditions of the marine boundary layer. In the past, a number of investigations of ozone to aqueous solutions containing iodide have been carried out using different methods.23,25-28 The bulk aqueous phase reaction is fast with a reaction rate constant around 1.2 × 109 M-1 s-123 and the lower limit of the bulk accommodation coefficient was estimated to be 10-2 from these studies. The bulk accommodation coefficient is defined as the probability that an ozone molecule colliding with an aqueous surface is actually entering the liquid in dissolved form. Iodide is also known to show the most pronounced preference for the aqueous solution-air interface among the halogenides due to its substantial polarizability.29-33 It could be expected that this also affects the heterogeneous reactivity with O3. The aim of the present work was to investigate the reaction of deliquesced particles composed of pure KI and of mixtures of KI and NaCl with gas phase O3 in an aerosol flow tube, in order to obtain an improved estimate for the bulk accommodation coefficient and to get a hint about possible surface effects. This is the first time this reaction has been investigated under aerosol conditions, which apart from the physical representation also more closely mimic the ionic strength conditions of marine boundary layer aerosol than other approaches with more dilute solutions. The experiments were assisted by a kinetic model to

10.1021/jp103257d  2010 American Chemical Society Published on Web 06/16/2010

7086

J. Phys. Chem. A, Vol. 114, No. 26, 2010

Rouvie`re et al.

Figure 1. Overview of the experimental setup.

track ozone and iodide concentrations as a function of time in the flow tube and to get better constraints on the relevant kinetic parameters. 2. Experimental Section A schematic representation of the experimental setup is given in Figure 1. Potassium iodide particles were produced by nebulizing an aqueous solution containing 5 g/L of potassium iodide salt into 5 L/min dry N2. The aerosol particles emitted were dried in a silica gel diffusion dryer (l ) 1.2 m). The particles were then exposed to a bipolar ion source (85Kr) to obtain an equilibrium charge distribution. An electrostatic precipitator removed all charged particles, and only neutral particles were passing on through the experiment to reduce losses of charged particles in isolating, but chemically inert tubing and in the flow reactor. For hygroscopic characterization experiments, a first DMA (differential mobility analyzer) could be placed after the ion source to obtain a monodispersed aerosol, which was neutralized again with another 85Kr source and the electrostatic precipitator. Experiments were performed at 72% RH (above the deliquescence humidity of 67%34), where the hygroscopic growth factor, that is, the ratio of the wet diameter to the dry diameter, is 1.3. We used a homemade humidifier to adjust the relative humidity by passing the flow of particles through a Goretex permeable membrane tube, which was kept in a temperature controlled water bath. The humidity could be adjusted to within (1% RH and was controlled by two capacitance humidity sensors. Under the conditions of the present experiments at 72% RH, an aqueous KI aerosol contains around 7.3 M KI. The background aerosol concentrations were lower than 20 particles/ cm3. The diameter of the particles used was in the range of 70-573 nm, and the size distribution was log-normal (log σ ) 0.25). Typical particle number, surface, and mass concentrations were: 5 × 105 particles/cm3, (1.6-6.6) × 10-4 cm2/cm3, and 1000-3550 µg/m3, respectively. Ozone was generated by irradiating a 0.5 L/min flow of a mixture of O2 and N2 in a quartz tube with an ultraviolet lamp (Pen-Ray 3SC-9, UV Products Ltd., USA), which has a resonance line at a wavelength of 185 nm. Then, the O3/O2/N2 flow was introduced to the aerosol flow tube through a movable injector and diluted by the aerosol flow or N2 (0.5 L/min) to a total flow of 1.0 L/min. In the absence of particles, the ozone concentration in the reactor was fixed for each experiment but could vary from one to the other from 70 to 300 ppb. The aerosol flow reactor was a Pyrex tube, 85 cm long, with an inner diameter of 2.5 cm. In order to minimize wall reactions, the tube was rinsed with ethanol and milli-Q water between each experiment. The flow tube was operated under laminar flow conditions at atmospheric pressure (970-980 mbar) and

ambient temperature (T ) 293 K). The temperature in the flow tube was maintained by making use of the efficient temperature controlled air conditioning system in the laboratory, which ensured that the DMA (see below) also had the same temperature, which was crucial to have the same relative humidity in the flow tube and the DMA. The ozone injector position, which was kept in the center of the flow tube by means of three PFA legs, could be moved to vary the reaction time from 2 to 25 s to obtain kinetic information. The movable injector was equipped with several lateral pinholes at its end to allow fast mixing of the two flows further downstream. At the exit of the flow reactor, part of the flow (around 330 mL/min) was directed to a Scanning Mobility Particle Sizer (SMPS) consisting of a DMA and a condensation particle counter (CPC), to measure the aerosol surface concentration of the aerosol. The sheath air for this DMA was taken from filtered sample gas such that the water vapor pressure in the DMA exactly matched that of the aerosol flow. The concentration of ozone was measured in the remainder of the flow downstream of the flow reactor with a photometric ozone analyzer (model ML 9810, Monitor Laboratories Inc., USA) after separating ozone from the particles by diffusion in an annular coflow device. This separation was necessary, because the aerosol interferes with the photometric ozone detection due to scattering and absorption. This separation device consists of a vertically mounted 60 cm long and 4.5 cm inner diameter acrylic glass tube with two concentric inlets a the bottom. The aerosol flow (about 670 mL/min) coming from the flow reactor passes through the inner inlet consisting of a 14 cm long and 3 cm inner diameter acrylic glass tube, the end of which exhibits a sharp edge to avoid turbulences at the point of mixing. A pure N2 flow is fed into the annular space between the separator tube wall and the inner inlet, adjusted to 670 mL such that the average linear gas velocities roughly match at the point of mixing. The principle of operation of this device is that under laminar conditions, gas exchange occurs by diffusion between the inner and outer flows, while particles remain in the inner part. At the top exit of the separator, a flow of 300 mL/min is pumped out through an orifice in the outer wall of the separator into the ozone analyzer. The residence time between ozone and particles in this device is 16 s under our flow conditions, thus the actual ozone-particle interaction times were from 18 to 41 s. The necessary prerequisite for proper performance is that this flow remained particle free (2 × 10 >(0.92-2.08) × 10-2 0.1 Rb ) 0.6 (

0.4 0.5

plotted against the ozone concentration in Figure 6. Therefore, as long as iodide is not depleted to less than 30% of its initial concentration, the bulk diffusion and reaction model explains the kinetic data very well within the range of ozone concentrations and particle sizes of the present study. As mentioned above, based on this data set, in many cases, the kinetics are consistent with a value for Rb of one. For individual experiments, this value is mostly well constrained. As an example, for experiment 5 we tried to run the model with a lower value of Rb ) 0.01, and we observed that it is not possible to fit the experimental data even by increasing the second-order rate constant to the collision limit. However, significant scatter is apparent among individual experiments. Several reasons may be responsible for this, such as oxidation of iodide in the nebulizer solution and thus a reduced initial iodide concentration in the aerosol, variations in humidity, variations in flow rates of the relatively complex flow system, and variation of the residence time in the coflow separation device, in which variable turbulent eddies could lead to variations in the separation time. Figure 7 summarizes the results obtained for mixed KI/NaCl particles. Note that the more conventional analysis of the effect of bulk phase concentration on uptake kinetics would be to plot the logarithm of the first-order rate constant against the logarithm of the bulk concentration. The slope of a linear regression to the data would then reveal whether a surface reaction (with the rate proportional to the bulk concentration) or a bulk reaction (with the rate proportional to the square root of the bulk concentration) prevails the kinetics. However, the significant iodide depletion occurring in the particles prevents us from doing this first-order analysis. Therefore, we test whether the analysis of the rate constant returned from each experiment leads to a consistent value for all concentrations. To calculate the iodide concentration for the mixed solution particles, we considered the solute mole fraction of iodide and chloride in the nebulized solution (xCl-,xI-). Then, by first measuring the size distribution under dry conditions for each experiment, we obtained the dry particle diameter (D0) and the dry particle volume, from which we deduced the total number of moles (iodide and chloride, nT0) in the mixed particles by taking into account the density and the mole fraction of each salt. Under humid conditions, where particles were deliquesced, we obtained a wet diameter and a wet volume (D, V), and we calculated the total concentration of solutes for the mixed aqueous particles (CT ) nTO/V). Finally, to obtain the iodide concentration we applied the mole fraction (xCl-,xI-) to this concentration. During these experiments the RH was adjusted to 75% in order to ensure that the mixed Cl-/I- particles got deliquesced. The volume growth factors also obtained from the analysis explained above were consistent with expectations for a volume-weighted average of the two individual solute component solutions. We neglected contributions from the low reactivity of chloride compared to iodide. Indeed, the second-order rate coefficient between chloride and ozone is very low (kb,Cl- ) 3 × 10-3 M-1 s-1).39

For Figure 7, we consider bulk reaction limited uptake as shown previously with our model. kb was obtained from adjusting the simulation assuming a bulk accommodation coefficient of 1 to the experimental data points in the same way as presented above. It shows that when assuming that Rb does not vary with iodide concentration, the kb value obtained from the experiment is not depending on the iodide concentration. The value of kb estimated previously is still consistent with the results obtained for mixed inorganic aerosol. To demonstrate the significance of this analysis, we calculated (dashed line in Figure 7) the dependence of kb of the iodide concentration for the case, where the measured ozone loss would have been due to a surface reaction, but which would have been erroneously analyzed as a bulk reaction. To construct this line, we created a data set using our model with the square root dependence of the uptake coefficient replaced by a linear dependence to determine the loss of ozone. This data was then reanalyzed as with the measured data explained above. This means that if we had a surface reaction driven system with equivalent rate at high iodide concentration, we would get a strong iodide dependence for the rate coefficient, if we would erroneously analyze it as a bulk reaction. Similarly to the case of pure iodide particles, this indicates that the bulk accommodation of ozone and bimolecular reaction with iodide in the bulk are sufficient to explain the present data of ozone uptake. Of course this would still leave room for surface reaction processes that involve three partners or depend on iodide coverage. At first glance, the absence of clear indications for a surface reaction is surprising, since it has meanwhile been well established that iodide is enhanced at the interface.31,33,40 However, this enhancement is restricted to the outermost interfacial layer, and is actually followed by a layer with depletion further below, with overall depletion near the interface, consistent with surface tension measurements.33 The reacto-diffusive length for ozone reacting with bulk iodide, that is, the characteristic length within which ozone is removed by reaction while diffusing into the interior of the solution,36 is 0.46 nm for the highly concentrated pure iodide particles and is a few nm for the low concentrations in the mixed particles. Therefore, the range of depths, which is probed by the present experiments, is larger than the depths to which these concentration variations occur near the interface. Hence, the average iodide concentration experienced by ozone diffusing into the particles over these depths is very close to the bulk concentration. Therefore, likely due to the strong reactivity of iodide, for the initial uptake kinetics of ozone to deliquesced iodide or mixed chloride/iodide particles, no surface reactivity seems to be important. Under our conditions, ozone loss was also likely not affected by the additional interfacial reactions suggested by Enami et al.7 The situation might be different for the less reactive bromide or chloride, where a surface reactivity was observed,41 but still not on the time scales of the present experiments.

7092

J. Phys. Chem. A, Vol. 114, No. 26, 2010

For the mixed Cl-/I- aerosol, the average uptake coefficient at time t ) 0 s returned by the simulation was found to decrease from γ ) 1.2 × 10-2 to γ ) 4.4 × 10-3 over the range 7.3-0.9 M of iodide and 0-6.1 M of chloride. Thus, decreasing the iodide concentration goes along with a decrease of the mass transfer of ozone to the bulk phase. An upper limit to the uptake coefficient of 1 × 10-4 for ozone to deliquesced sodium chloride particles was estimated by Abbatt and Waschewsky.42 Our results suggest that also for lowest amounts of iodide in presence of chloride, iodide remains the most important sink for ozone, consistent with known bulk phase chemistry. Comparison with Previous Results. The reactive uptake of ozone on solid potassium iodide was most recently estimated by Brown et al.25 reporting an average reactive sticking probability of (1.4 ( 0.7) × 10-4. They observed the formation of a thin layer of KIO3 at the surface,43 which then leads to deactivation. The significantly faster uptake kinetics observed in aqueous solution is an indication that iodide in its free ionic form is the reactive species, and the liquid medium allows O3 to diffuse and react in the bulk. The average of the uptake coefficients estimated in this study is equal to γ ) (1.10 ( 0.20) × 10-2, independent of the pH of the solution and of the scenario. There was no significant difference between the γ values in presence of sulfuric acid. The agreement between the experimental data and the simulation indicates reactivity is governed by accom0.4 , followed modation of ozone into the bulk, with Rb ) 0.6 ( 0.5 by diffusion and reaction in the bulk, with a second-order rate constant kb ) (1.0 ( 0.3) × 109 M-1 s-1. In the past, the uptake of ozone to aqueous solutions has been carried out using different methods. Table 3 shows a comparison of Rb and kb values reported in the literature. Previous studies indicate that Rb for O3 on iodide solutions is in the range of 2 × 10-3 to 1, and that bulk accommodation could be a rate limiting step.26 Our results appear to be consistent with the values of Rb and kb reported in the past, but we are able to provide a better constraint on the value of Rb. This is due to the fact that uptake as measured by the loss of ozone to suspended sub-micrometer particles is much less affected by gas phase diffusion than for the experiments with a large individual drop by Schutze et al44 or the droplet train by Hu et al. and Magi et al.,26,27 both techniques requiring complex corrections. A large bulk accommodation coefficient of ozone on aqueous solution is also in line with predictions by molecular dynamics simulations reported by Roeselova et al.45 In conclusion, the bulk accommodation coefficient is not the limiting factor for ozone uptake to sea salt particles. 4. Conclusions The reaction of ozone with deliquesced potassium iodide particles was investigated with an aerosol flow tube experiment with the analysis assisted by kinetic model simulations. This is the first time that this reaction has been investigated under aerosol conditions, that is, under high ionic strength conditions of relevance to sea salt aerosol in the marine boundary layer. The initial uptake coefficient is γ ) (1.10 ( 0.20) × 10-2 with a second-order rate coefficient in the bulk phase of kb ) (1.0 ( 0.3) × 109 M-1 s-1. The bulk accommodation coefficient is 0.4 . In mixed NaCl/KI particles, the estimated to be Rb ) 0.6 ( 0.5 initial uptake coefficients range from 4.4 × 10-3 to 1.2 × 10-2 for 0.9-7.3 M iodide concentration at 75% RH, in agreement with expected behavior for reaction in the bulk. A contribution by a surface process was not apparently contributing to the ozone loss. Exposing iodide to ozone in the form of sub-micrometer

Rouvie`re et al. aqueous aerosol particles leads to kinetics expected for more dilute bulk solutions. Therefore, the loss of iodide in real sea salt aerosol of the marine boundary layer can be described as a bulk aqueous reaction, with improved constraints on the rate parameters presented in this work. Acknowledgment. This work was supported by the Swiss National Science Foundation (grant no 200020-109341). We are grateful for the input of our colleagues S. Bruetsch, M. Birrer, and D. Piguet. References and Notes (1) Davis, D.; Crawford, J.; Liu, S.; McKeen, S.; Bandy, A.; Thornton, D.; Rowland, F.; Blake, D. J. Geophys. Res. [Atmos.] 1996, 101, 2135. (2) Chang, W.; Heikes, B. G.; Lee, M. Atmos. EnViron. 2004, 38, 1053. (3) Von Glasow, R. Nature 2008, 453, 1195. (4) Finlayson-Pitts, B. J. Chem. ReV. 2003, 103, 4801. (5) Tucceri, M. E.; Holscher, D.; Rodriguez, A.; Dillon, T. J.; Crowley, J. N. Phys. Chem. Chem. Phys. 2006, 8, 834. (6) Plane, J. M. C.; Joseph, D. M.; Allan, B. J.; Ashworth, S. H.; Francisco, J. S. J. Phys. Chem. A 2006, 110, 93. (7) Enami, S.; Vecitis, C. D.; Cheng, J.; Hoffmann, M. R.; Colussi, A. J. Chem. Phys. Lett. 2008, 455, 316. (8) Pechtl, S.; Schmitz, G.; von Glasow, R. Atmos. Chem. Phys. 2007, 7, 1381. (9) Saunders, R. W.; Plane, J. M. C. EnViron. Chem. 2005, 2, 299. (10) Saiz-Lopez, A.; Plane, J. M. C.; McFiggans, G.; Williams, P. I.; Ball, S. M.; Bitter, M.; Jones, R. L.; Hongwei, C.; Hoffmann, T. Atmos. Chem. Phys. Discuss. 2005, 5, 5405. (11) Davis, E. J. J. Phys. Chem. A 2008, 112, 1922. (12) Vogt, R.; Sander, R.; von Glasow, R.; Crutzen, P. J. J. Atmos. Chem. 1999, 32, 375. (13) McFiggans, G.; Plane, J. M. C.; Allan, B. J.; Carpenter, L. J.; Coe, H.; O’Dowd, C. J. Geophys. Res., [Atmos.] 2000, 105, 14371. (14) Blanchard, D. C. J. Geophys. Res., [Oceans Atmos.] 1985, 90, 961. (15) Gilfedder, B. S.; Lai, S. C.; Petri, M.; Biester, H.; Hoffmann, T. Atmos. Chem. Phys. 2008, 8, 6069. (16) Baker, A. R. EnViron. Chem. 2005, 2, 295. (17) Gabler, H. E.; Heumann, K. G. Int. J. EnViron. Anal. Chem. 1993, 50, 129. (18) Ku¨pper, L.J., C.; G.B., M.; C.J., P.; T.J., W.; Boneberg, E. M.; S., W.; M., W.; R., A.; D., G.; P., P.; A., B.; G.W., L. I.; P.M.H., K.; W., M.-K.; M.C., F. Proc. Natl. Acad. Sci. U.S.A. 2008, 105, 6954. (19) Chance, R.; Baker, A. R.; Ku¨pper, F. C.; Hughes, C.; Kloareg, B.; Malin, G. Estuarine, Coastal Shelf Sci. 2009, 82, 406. (20) Dixneuf, S.; Ruth, A. A.; Vaughan, S.; Varma, R. M.; Orphal, J. Atmos. Chem. Phys. 2009, 9, 823. (21) Oh, I.-B.; Byun, D. W.; Kim, H.-C.; Kim, S.; Cameron, B. Atmos. EnViron. 2008, 42, 4453. (22) Hoigne, J.; Bader, H. Water Res. 1983, 17, 173. (23) Liu, Q.; Schurter, L. M.; Muller, C. E.; Aloisio, S.; Francisco, J. S.; Margerum, D. W. Inorg. Chem. 2001, 40, 4436. (24) Garland, J. A.; Curtis, H. J. Geophy. Res., [Oceans Atmos.] 1981, 86, 3183. (25) Brown, M. A.; Newberg, J. T.; Krisch, M. J.; Mun, B. S.; Hemminger, J. C. J. Phys. Chem. C 2008, 112, 5520. (26) Hu, J. H.; Shi, Q.; Davidovits, P.; Worsnop, D. R.; Zahniser, M. S.; Kolb, C. E. J. Phys. Chem. 1995, 99, 8768. (27) Magi, L.; Schweitzer, F.; Pallares, C.; Cherif, S.; Mirabel, P.; George, C. J. Phys. Chem. A 1997, 101, 4943. (28) Muller, B.; Heal, M. R. Phys. Chem. Chem. Phys. 2002, 4, 3365. (29) Cheng, J.; Hoffmann, M. R.; Colussi, A. J. J. Phys. Chem. B 2008, 112, 7157. (30) Cheng, J.; Vecitis, C. D.; Hoffmann, M. R.; Colussi, A. J. J. Phys. Chem. B 2006, 110, 25598. (31) Jungwirth, P.; Tobias, D. J. J. Phys. Chem. B 2001, 105, 10468. (32) Ghosal, S.; Verdaguer, A.; Hemminger, J. C.; Salmeron, M. J. Phys. Chem. A 2005, 109, 4744. (33) Jungwirth, P.; Tobias, D. J. Chem. ReV. 2006, 106, 1259. (34) Woods, E, III; Kim, H. S.; Wivagg, C. N.; Dotson, S. J.; Broekhuizen, K. E.; Frohardt, E. F. J. Phys. Chem. A 2007, 111, 11013. (35) Sakamoto, Y.; Yabushita, A.; Kawasaki, M.; Enami, S. J. Phys. Chem. A 2009, 113, 7707. (36) Po¨schl, U.; Rudich, Y.; Ammann, M. Atmos. Chem. Phys. 2007, 7, 5989. (37) Kosak-Channing, L. F.; Helz, G. R. EnViron. Sci. Technol. 1983, 17, 145. (38) Winger, R. J.; Ren, L. Food Chem. 2009, 113, 600.

Uptake of Ozone to Deliquesced KI and Mixed KI/NaCl (39) Hoigne, J.; Bader, H.; Haag, W. R.; Staehelin, J. Water Res. 1985, 19, 993. (40) Ghosal, S.; Hemminger, J. C.; Bluhm, H.; Mun, B. S.; Hebenstreit, E. L. D.; Ketteler, G.; Ogletree, D. F.; Requejo, F. G.; Salmeron, M. Science 2005, 307, 563. (41) Mochida, M.; Hirokawa, J.; Akimoto, H. Geophys. Res. Lett. 2000, 27, 2629. (42) Abbatt, J. P. D.; Waschewsky, G. C. G. J. Phys. Chem. A 1998, 102, 3719. (43) Brown, M. A.; Liu, Z.; Ashby, P. D.; Mehta, A.; Grimm, R. L.; Hemminger, J. C. J. Phys. Chem. A 2008, 112, 18287. (44) Schutze, M.; Herrmann, H. Phys. Chem. Chem. Phys. 2002, 4, 60.

J. Phys. Chem. A, Vol. 114, No. 26, 2010 7093 (45) Roeselova, M.; Jungwirth, P.; Tobias, D. J.; Gerber, R. B. J. Phys. Chem. B 2003, 107, 12690. (46) Garland, J. A.; Elzerman, A. W.; Penkett, S. A. J. Geophys. Res. 1980, 85. (47) Bichsel, Y.; Von Gunten, U. EnViron. Sci. Technol. 1999, 33, 4040. (48) Troy, R. C.; Kelley, M. D.; Nagy, J. C.; Margerum, D. W. Inorg. Chem. 1991, 30, 4838. (49) Brimblecombe, P.; Clegg, S. L. J. Atmos. Chem. 1988, 7, 1.

JP103257D