The heterogeneous reactions of HNO3(g) + NaCl(s) -HCl(g) + NaNOs(s) (eq 1) and N205(g) + NaCl(s) -ClN02(g) + NaNO3(s) (eq 2) were investigated over the temperature range 223-296 K in a flow-tube reactor coupled to a quadrupole mass spectrometer. Either a chemical ionization mass spectrometer (CIMS) or an electron-impact ionization mass spectrometer (EIMS) was used to provide suitable detection sensitivity and selectivity. In order to mimic atmospheric conditions, partial pressures of HN03 and N2O5 in the range 6 x Torr were used. Granule sizes and surface roughness of the solid NaCl substrates were determined by using a scanning electron microscope. For dry NaCl substrates, decay rates of HN03 were used to obtain y(1) = 0.013 f 0.004 (lo) at 296 K and '0.008 at 223 K, respectively. The error quoted is the statistical error. After all corrections were made, h e overall error, including systematic error, was estimated to be about a factor of 2. HC1 was found to be the sole gas-phase product of reaction 1. The mechanism changed from heterogeneous reaction to predominantly physical adsorption when the reactor was cooled from 296 to 223 K. For reaction 2 usingdry salts, y(2) was found to be less than 1.0 x at both 223 and 296 K. The gas-phase reaction product was identified as ClN02 in previous studies using an infrared spectrometer. An enhancement in reaction probability was observed if water was not completely removed from salt surfaces, probably due to the reaction of N2O5(g) + H2O(s) -2HN03(g). Our results are compared with previous literature values obtained using different experimental techniques and conditions. The implications of the present results for the enhancement of the hydrogen chloride column density in the lower stratosphere after the El Chichon volcanic eruption and for the chemistry of HCl and HN03 in the marine troposphere are discussed.

No abstract

The uptake of HCl in water ice and nitric acid ice films has been investigated in a flow reactor interfaced with a differentially pumped quadrupole mass spectrometer. These studies were performed under experimental conditions that may mimic the polar stratosphere. The HCl uptake in ice films at 188 and 193 K was determined to be in the range of 8.7 X 1013 to 1.8 X 1015 molecules/cm2 (if the geometric area of the flow reactor, 290 cm2, was used in the calculation) when HCl partial pressures of 7 X 10-8 to 6 X 1 0 6 Torr were used. On the basis of a model which accounts for the total surface area of the films, the true surface density could be a factor of 25 lower than that calculated by the geometric area. A slightly higher uptake was observed at the lower temperature of 188 K. The uptake of HC1 in ice was significantly enhanced by using an HCl partial pressure greater than 1 X Torr. The observation was found to be consistent with the formation of the hexahydrate or the trihydrate of HC1 according to the phase diagram of the HCl/H20 system. The uptake of HCl in nitric acid ice at 188 K was determined to be in the range of 8.0 X 1013 to 5.3 X lOI4 molecules/cm2 at a HCl partial pressure of 4.5 X lo-' Torr. Measurement of both H N 0 3 and H2O vapor pressures was made to positively identify the formation of nitric acid trihydrate (NAT) surface according to the phase diagram of the HNOs/ H2O system. The HCl uptake in NAT is comparable to that in water ice in the present experiment, but significantly smaller than the previously reported values by Mauersberger and his co-workers. Implications of these results for the heterogeneous chemistry of the polar ozone depletion are briefly discussed.

Laboratory studies of sticking coefficients of H2O, HCl, Cl2 and HNO3 on ice and heterogeneous reactions of ClONO2 with ice or HCl/ice have been performed in a fast flow reactor. A quadrupole mass spectrometer with electron impact ionization was used as a detector. Measured sticking coefficients are: 0.3 (+0.7. −0.1) for H2O, 0.4 (+0.6, −0.2) for HCl, <1.0 × 10−4 for Cl2, and 0.3 (+0.7, −0.1) for HNO3 at 200 K. The reaction probability of ClONO2 on ice was found to be 0.06 (±0.03) while HOCl was observed as a sole product in the gas phase. With HCl present in ice, the reaction probability of ClONO2 is greatly enhanced, approaching 0.27 (+0.73, −0.13) while molecular chlorine was found to be the major product in the gas phase. Another reaction product was nitric acid which remained in the solid phase. Since the polar stratospheric clouds contain ice particles or possibly HCl/ice particles on the surface, the present results should be a major factor in producing the observed springtime ozone depletion in the Antarctic stratosphere.

In a recent article Hanson and Ravishankara (HR) obtained data for the reaction probabilities (7s) of N205 and ClONO2 on ice films of varying thickness.' They interpreted the observed change in y as evidence that no corrections for internal diffusion were needed and, thus, by implication that their films were effectively nonporous. H R also calculated y vs thickness by using the theory of surface reaction and pore diffusion2 and obtained results that did not agree with their experimental data. The purpose of this comment is to point out the following: (1) the strong evidence that vapor deposited ice films can be highly porous;(2) that porous films do not always show a large increase in reactivity or uptake capacity with thickness; (3) that H R made assumptions that resulted in the poor fit of their data; and (4) the need for characterization of the ice films used.(1) Ice films formed by deposition from the vapor phase have been widely used to simulate stratospheric cloud surfaces for laboratory measurements of reaction and uptake rates. To obtain intrinsic surface reaction probabilities that can be used in atmospheric models, we need to know the area of the film surface that actually takes part in the reaction. If the films are smooth and nonporous, the geometric area can be used, as was done in all of the early work using these ice films. However, studies of the morphology of films deposited under conditions (temperature, pressure, deposition rate, sample size, film thickness, substrate) similar to those used for the rate measurements show that the films have temperature-sensitive surface areas much larger than the geometric area3 and that they are composed of loosely consolidated granules with diameters of a few micrometers or le~s.3.~ The large internal areas of such films consist of the surface areas of the individual ice granules. The loose packing of the granules makes these films highly porous and allows rapid gasphase diffusion into the interior of the film. For porous films, the observed rates are affected by the internal surface area, and corrections, which account for the interaction of surface reaction and pore diffusion, are required in order to extract intrinsic Y S .~ The magnitude of these corrections is less than a factor of 3 for an observed y > 0.1. However, for y < 0.1, the corrections become larger and the use of geometric areas for porous films results in ys that are upper limits to the true values.(2) An increase in reactivity with thickness implies that a film is porous, but the converse is not always true. That is, a porous film can show little or no change in reactivity with increased thickness depending on its morphology. This can be seen by considering a simple model based on our studies of these films using optical and electron microscopy and gas adsorption (BET) In this approximation, the film comprises spherical granules stacked in layers. For a porous film, the relation between the observed probability, y(obs), which is calculated by using the geometric area, and the true value, ...