tions indicates that this phenomenon also occurs with hydronium ions in water (see eactions within or on the surfaces of ical constituent. It has been recognized for the figure). -) liquid or solid particles are known to some t h e that HCl volatilizes from sea .salt With chloride ions predicted to occupy a i t p l a y an important role in the strato-aerosols as a result of displacement by substantial hction of the sea salt aerosol sphere, the atmospheric layer from about strong acids, such as

more »

... N03 or H2SO4 gases surface, incoming gas-phase OH radicals 10 to 50 kilometers above Earth's surface. (6). More recently, a combination of atmocan encounter Cl-ions to form a surface Much less is known about such heteroge-spheric measurements, laboratory experi-OH-Cl-complex, which is then assumed to neous chemistry in ments, and kinetic simulations has shown undergo either self-reaction or decomposi-'~&anced online at the troposphere, the that reactive chlorine-containing gases (C12, tion. The self-reaction releases molecular ' r~n c e m a g . o r g / c g i / layer from Earth's HOCl, CINO2, and ClBr) may be released chlorine to the gas phase. When the rates of -.&1tenVfuW288/5464/285 surface up to the from sea salt aerosols by heterogeneous self-reaction and decomposition of the stratosphere. What processes (7, 8). Episodic destruction of complex are adjusted to fit experimental little is known has focused on the ubiqui-boundary layer ozone in the Arctic has been data, the observed rate of laboratory protous aqueous droplets, which have general-attributed to catalytic cycles involving BrO duction of molecular chlorine can be reproly been assumed to function as well-mixed radicals (9,lO). Satellite observations show duced by this mechanism (I). Moreover, the microscopic chemical reactors. But this resulting concentration of view is likely too simplistic. On page 301, Monte Carlo simulation of a cluster of water C1 atoms falls in the right Knipping et al. (I) use a combination of molecules and one hydronium (H30+) ion. This range when compared with laboratory experisimulation, performed by lsamu Kusaka and David ambient data, namely about ments, kinetic mod-Oxtoby of the University of Chicago, uses the po-104 to lo5 molecules/cm3. eling, and molecular larizable model potentials for hydronium ion and Other highly concentratdynamics simula-a water molecule proposed by Kozack and Jordan ed inorganic salt particles tions to show that re-( 74). In the model, the hydronium ion and the waare also ubiquitous in the actions at the droplet ter molecule each Cany a single isotropic polan'zatmosphere. Aerosol sulsurface may also abiliw. The simulation used a spherical cell with a fates and nitrates, together contribute in an im-radius of 30 A, centered around the oxygen site of with their associated portant way to tropo-the hydronium ion and containing 22 water and water, the spheric heteroge-molecules. The hydronium ion prefers the surface largest category, by mass, of neous chemistry. of the cluster because it cannot be incorporated inorganic anthropogenic easily into the hydrogen bond network of the wa- There is an enorter molecules. The preferential surface migration is particulate matter. It is mous variety of tropospheric probably caused by the mismatch in H -0 4 bond lmOthat such pmicles surfaces, ranging from sea salt, mineral angles, rather than the size of the ion itself. exert a radiative effect on dust, and combustion-generated aerosols to climate, but from a chemi-Earth' s surface itself Together with the be-cal point of view they have wild-array of gas-phase species in the that tropospheric air masses enriched in generally been assumed to function simply troposphere, this leads to a wide range of BrO are always situated close to sea ice, as reservoirs for their gaseous molecular potential heterogeneous interactions. supporting the hypothesis that autocatalytic counterparts. Do ionic surface enrichment Knowledge of tropospheric heterogeneous release of bromine from sea salt gives rise mechanisms also occur in such particles? If chemistry is almost certainly incomplete, to substantial BrO formation (11) . sb, what are the implications for gas-phase but some key processes have been identi-During halogen liberation from sea salt, tropospheric chemistry? These questions fied. These include conversion of NO, (NO gaseous C12 is generated from aqueous sea will likely receive serious attention in the + NO2) to HN03 by the reaction of N20S salt particles in the presence of gas-phase laboratories of atmospheric and aerosol with moist aerosols (2) and heterogeneous hydroxyl (OH) radicals (12). The generachemistry for some time to come. hydrolysis of N203 to form nitrous acid tion of molecular chlorine was presumed to (HONO) (3). Photochemical production of be a result of the interaction of gas-phase References and NOW formaldehyde at the air-snow interface con-OH radicals with the brine solution, but the l: ~M l i l i K ; ' , P $ , " $ 4 f~1 1~~~i 2 K~1~~~ Photostitutes an important local HCHO source in observed rate of C12 production could not h m h l m ffw UK in Strat0-c w w ( E v a L I Almb surface (4). Recently, it has been be explained on the basis of known bulk No. 12,~et ~ropulsion u*ratory.~asadena,C4 IW . E speculated that mineral dust, long believed aqueous-phase chemistry. There is evidence ktgEtA1b.~,%ze2g/;d' to be a nonreactive substrate, can interact that large, polarizable anions preferentially 5. F. J. Dentener etd.1. a o p h p ~e s . 101. 22869 (1996). f with N205, 03, and H02 radicals to infiu-concentrate at the surface of aqueous solu-6. E. Er i bsonp Te11us1*375 (lgS9). ence gas-phase chemistry (5). 7. A. A. P. Pszenny et dl., Geophys. Res. Lett 20, 699 tions of inorganic salts, in contrast to small, (1993). Despite its benign source, sea salt nonpolarinble cations, which remain in the 8. C.\Ol~eetaL G w h~ Res.Lett25.3831(1958) 9. LA Bade et aL, Nature 334,138 (1988). % aerosol has proved to be a surprisingly bulk of the solution phase (13). Knipping et " , , , " pba rrmr,H, 450 ( l W, ) . 3 complex and interactive tropospheric chem-al. (I) performed molecular dynamics sim-11. T. Wagner and U. Plan, ~a t u n 395,466 (1998). 8 ulations of NaCl-H20 to inves& 12. K.W. OUm Science 279.74 (1998). 13. P. Jungwirth.]. Phys. Chem. A 104,145 (2000). The author is in the Department of Chemical Engi-gate to what extent this effect also Wcurs in 14. R E. Kozack and P. C. Jordan, 1. Chem Phys %, 3120 neering and the Division of Engineering and Applied seawater and whether this might expfain the (1992); R E. Kozack and P. C. Jordan, 1. Chem. f h~ 96,3131 (1992). Science, California Institute of Technology, Pasadena, generation of Cl2-Their results show that 15. t h, k 1. ~~~~k~ and . o~o~y p d i n g the CA 91125, USA.