The kinetics of condensation of D 2 18 O water vapor on D 2 16 O ice was studied in the temperature range 140-220 K. The measurements have been performed in real time using the pulsed valve technique in a lowpressure flow reactor. Ice samples were prepared by several different methods. The uptake coefficient γ was observed to decrease with increasing surface temperature, and varied from γ ) 0.06 to 0.8. A significant dependence of γ on the method of preparation of the ice has been found. At 180 K, γ ) 0.13 for singlecrystal ice, 0.18 for ice condensed from the vapor phase, and approximately 0.25 for bulk ice obtained from freezing liquid D 2 16 O. The uptake coefficient attained a value of γ ) 0.8 at 140 K for cubic ice prepared by vapor condensation at 140 K. The activation energy for evaporation of D 2 O at low temperatures (140-190 K) has been measured as 12.2 kcal/mol, whereas it decreases to 8.3 kcal/mol at higher temperatures (190-220 K). The rate of evaporation at 200 K corresponds to the loss of approximately 70 ( 10 formal monolayers per second. The experimental results suggest the formation of loosely bound water adsorbed to the surface of ice whose bond energy is estimated to be 4.0 ( 0.4 kcal/mol, independent of the type of ice.

The properties of the interface region of bulk, vapor-deposited, and single-crystal ice have been studied in a Knudsen cell flow reactor in the range 190−210 K using repetitive pulse experiments (RPEs) at variable frequency. Fluxes of surface-to-bulk loss in ice on single-crystal and bulk ice vary from 5 × 1011 to 1 × 1013 and from 5 × 1012 to 5 × 1014 molecule s-1 cm-2 for HCl and D2O, respectively. A positive activation energy for diffusional loss of E A = 3.0 ± 0.5 kcal/mol for HCl/ice and 5.3 ± 0.7 kcal/mol for D2O/ice has been measured. Complementary measurements (“dope and probe” experiments) of the HCl/ice interface region using the titration reaction of ClONO2 + HCl → Cl2 + ΗΝΟ3 are in good agreement with the diffusional loss measurements from RPEs. These experiments allowed the evaluation of the thickness h of the interfacial region, defined as the near-surface region of the ice where HCl is immediately available for titration at a high rate. We measured h = 100 ± 10 nm for single-crystal ice, 200 ± 50 nm for vapor-deposited ice, and 1000 ± 200 nm for bulk ice samples. The modeling of our results according to the laws of diffusion leads to values of the HCl diffusion coefficient D HCl ranging from (4.0 ± 1.0) × 10-14 to (2.8 ± 1.0) × 10-12 cm2 s-1 for single-crystal and bulk ice, respectively.

The uptake kinetics of HOBr on ice with and without HX has been measured in a Teflon-coated low-pressure flow reactor (Knudsen cell) at temperatures of 175-205 K. The values of the initial uptake coefficient γ 0 of HOBr on different types of pure H 2 O ice such as single-crystal ice, vapor-deposited ice, and samples frozen from liquid H 2 O range from 0.4 to 0.03 and reveal a pronounced negative temperature dependence with an activation energy of E a ) -9.7 ( 1.0 kcal/mol. The rate of HOBr uptake is independent of the type of ice. HOBr is able to sustain an equilibrium vapor pressure above an ice surface in the range from 185 to 210 K when sufficient HOBr has been adsorbed and is associated with an enthalpy change ∆H 0 r ) -9.4 ( 1.0 kcal/mol. Adsorbed HNO 3 has no influence on the HOBr uptake coefficient on ice, even when the amount of HNO 3 contaminating the surface of ice is as high as 10 formal monolayers. The interaction of HOBr with HX-doped ice in the temperature range from 180 to 215 K leads to rapid formation of BrX at values of γ 0 of HOBr being less temperature dependent than for HOBr adsorption on pure ice: E a ) -6.6 ( 2.0 kcal/mol has been found for the narrow temperature range from 195 to 215 K. The mass balance for BrCl from the reaction HOBr + HCl/ice is closed, in contrast to Br 2 , which is the primary product of the reaction HOBr + HBr/ice. The uptake rate coefficients for HOBr + HCl/ice and HBr/ice are not significantly different from each other over the range from 180 to 215 K and are equal to γ 0 ) 0.3 from 180 to 195 K, after which they drop with increasing temperature, but less so than for pure ice. At 205 K, γ 0 for HOBr + HX is typically a factor of 4 higher than γ 0 for HOBr uptake on pure ice. The atmospheric implications of these results are briefly discussed.

Pages 11739-11750. Equations 6 and 7 of the abovementioned publication contain a regrettable error in that R evap , the rate of evaporation of H 2 O from ice, has been confused with the NET rate of evaporation F 0 ) R evap -R cond which should have been used in the first place. The rate of condensation of water vapor, R cond , on the ice substrate may not be neglected in the used low-pressure flow reactor in the temperature range of interest. The correct equations describing the loss of H 2 O from the ice sample owing to hydrolysis of ClONO 2 , reaction R-2, and evaporation are therefore Table 5 displays the thickness of the interface h calculated according to eq 7 for the different ice samples. Table 6 presents the average values of h for the three types of ice examined in this work and the recalculated diffusion coefficients of HCl in ice, D HCl , as a function of the average interface thickness h. Although the numerical values of h are now significantly smaller than in the original publication amounting to approximately one-third of the value of the published ones, the main conclusion of the study remains unchanged. It may be summarized by stating that HCl is distributed over a layer tens of nm thick in an open interfacial structure whose HCl content rapidly interacts with the gas-phase such as reaction R-1. The resulting diffusion coefficients D HCl for the HCl/ice interface have decreased as a consequence of the increased average mole fraction X HCl within the interface thickness h. A forthcoming publication 51 will present a synopsis of the measured diffusion coefficients of HCl and HBr in ice obtained by using the DPE ("dope and probe") technique presented in this work. In this more recent study, 51 F 0 has been measured for every experiment such that the estimate according to eq 7a may be replaced with experimental data obtained under identical conditions to those of the measurement of the rate of halogen exchange, reaction R-1, and hydrolysis of chlorine nitrate, reaction R-2.

No abstract