Technical Comments

Response to Comments on "Reactions at Interfaces As a Source of Sulfate Formation in Sea-Salt Particles"

See allHide authors and affiliations

Science  30 Jan 2004:
Vol. 303, Issue 5658, pp. 628
DOI: 10.1126/science.1092750

Our paper (1) demonstrated that the oxidation of chloride at the interface will be a new source of alkalinity that can modulate the acidification of particles during the day when gaseous OH is present. We proposed that this modulation of the acidity of sea-salt particles could potentially affect the rate of uptake and oxidation of SO2 [S(IV)], which is very sensitive to pH.

Keene and Pszenny (2) suggest that we have overestimated the influence of the production of alkalinity from the interface reaction of OH with Cl on S(IV) uptake and oxidation. This is based on field measurements of pH, nitrate and sulfate concentrations in marine aerosol particles, and model calculations of their acidification. To make a first-order estimate of whether the modulation of pH we propose could be significant, we examined the uptake and oxidation of SO2 in freshly generated sea-salt particles under conditions of a remote marine environment where concentrations of nitrogen oxides and therefore of nitric acid are low. Biogenically produced dimethyl sulfide and other organic sulfur compounds, however, are present in remote atmospheres (3, 4) and serve as sources of sulfur dioxide and sulfuric acid.

We do not suggest that the proposed interface mechanism will keep all sea-salt particles alkaline over the course of an entire day under all conditions, but rather that it will modulate the rate of titration of the alkalinity by acids during midday when gas-phase OH concentrations increase. The contribution to modulating particle pH will therefore vary diurnally, as well as with particle size, because the reaction occurs at the interface. The integrated impact of this interface process will obviously depend on such factors, as well as whether the air mass is significantly polluted, in which case acidification by anthropogenic nitric and sulfuric acids will occur more rapidly. A variable fraction of particles would be expected to become acidic over the course of a day or even less, depending on the particular conditions. However, observation of acidic sea-salt particles, particularly in air masses affected by anthropogenic emissions, does not rule out a decrease in the rate of acidification during the day from the interfacial OH-Cl reaction.

Our mechanism is therefore not inconsistent with the data cited by Keene and Pszenny (2). For example, Huebert et al. (5), cited by Keene and Pszenny, stated that their marine air mass samples from Santa Maria, Azores, “did not always represent pristine marine air, since it may include some continental air, as well as occasional emissions from San Miguel Island when the wind was northwesterly.” Similarly, measurements at Bermuda are often affected by continental emissions; for example, Keene et al. (6), also cited in the comment by Keene and Pszenny, described the sampling for these pH measurements as having been carried out “under moderately polluted conditions.”

Keene and Pszenny (2) state that “most fresh sea-salt aerosols are acidified within seconds to tens of minutes via incorporation of acids and precursors.” This appears to be based primarily on model calculations applied to polluted conditions. However, model calculations for remote regions such as Christmas Island by Erickson et al. (7), cited by Keene and Pszenny, predict that there are insufficient acids to titrate the alkalinity, and the pH of even the smallest particles is calculated to remain above 6 [figure 6 in (7)]. Of course, model predictions of both remote and polluted atmospheres can only reflect the model inputs, and none of these models includes interface chemistry.

Sander et al. (8) state that our “extrapolation of the laboratory atmospheric conditions is erroneous,” based on diffusion limitations for OH and the values we assumed for the accommodation coefficient and concentrations of OH, as well as the concentration and size of sea-salt particles. We respectfully disagree. The impact of diffusion from the gas phase depends on particle size, of course, becoming less important for smaller particles. The amount of airborne sea salt and its size distribution depends particularly on wind speed, altitude, and proximity to surf zones (914); although we chose 1-μm particles for simplicity, there are many small sea-salt particles that, in terms of number concentration, exceed the larger particles by one to two orders of magnitude. Although we considered only gas-phase OH as the source of this free radical at the interface, this most certainly underestimates the amount of OH available for the reaction; an increasing body of evidence suggests that additional sources of the hydroxyl radical at the interface exist. For example, recent molecular dynamics (MD) simulations (15) show that the nitrate ion clearly prefers the air-water interface compared with solvation in the bulk; photolysis of nitrate is an efficient source of OH (16) and hence is likely a direct source of interfacial OH that is not limited by diffusion. Furthermore, MD calculations suggest that OH may itself prefer the interface (17), which would lead to an increase in the importance of its surface reactions, including that with Cl. These are obvious examples; others will doubtless become apparent as theoretical approaches to the production, segregation, and reaction of species at interfaces receive increasing attention. A major research effort on these interfacial processes is currently under way (18).

The simplified version of MOCCA applied by Sander et al. (8) predicts that the impact on particle sulfate could be 10% when a diffusion limitation is included (likely an underestimate, given the existence of direct sources and enhancement of OH at the interface), and about 60% without the diffusion limitation. These calculations assume an initial pH of 5.3, whereas fresh sea-salt particles are alkaline. As stated in our original article, large-scale models devoted to understanding SO2 and its oxidation in the atmosphere commonly tend to underpredict sulfate in the boundary layer by ∼ <20%, and gas-phase SO2 concentrations tend to be overpredicted by larger amounts (1922).

We believe that the values we adopted in our first-order estimate are quite reasonable. Only lower limits for the accommodation coefficient of OH on water have been experimentally measured, and these are large. For example, Takami et al. (23) reported that the accommodation coefficient was “>0.01 and was possibly close to unity” at 293 K. This would not be surprising, given that the mass accommodation coefficient of gas-phase water on liquid water at 280 K has been measured to be 0.17 (24), and that of OH would be expected to be at least as large. As discussed above, the sea-salt concentration and size distribution varies with a number of parameters (914), and the values of 10 particles cm–3 and a radius of 1 μm fall well within the range of reasonable values for a first-order approach.

Finally, Sander et al. (8) cite a daytime global OH average of 2 × 106 cm–3 based on a modeling study. We prefer to use experimental measurements of OH concentrations that have been made in the troposphere, which commonly show a peak OH concentration of about 107 OH cm–3, symmetrically distributed about local noon (25). A value of 5 × 106 cm–3 for a typical daytime value therefore seems reasonable. This is supported by measurements of OH in the marine boundary layer. For example, Mauldin et al. (26) reported that around Christmas Island, “typical midday average OH concentrations observed in the boundary layer under clean, cloudless conditions were (6-8) × 106 cm–3.” Similarly, Davis et al. (27) reported median values for OH concentrations in the range of (4-8) × 106 cm–3 at high noon in the marine boundary layer. We therefore judge an OH concentration of 5 × 106 cm–3 to be reasonable for our first-order calculations.

A model is currently being developed (28) that will incorporate interface chemistry, along with appropriate NOx-VOC-SOx chemistry, for quantitative evaluation of the importance of reactions at the air-water interface of particles in the atmosphere. Inclusion of a variety of interfacial species and processes, along with size-segregated particles, will allow a more realistic evaluation of the contribution of this and other such surface reactions to the atmospheric chemistry of both gases and particles. However, it is clear from our experiments and the developing theoretical work that this exciting and rapidly developing field of reactions at interfaces has the potential to play a significant role in the chemistry of the atmosphere, including the marine boundary layer.

References and Notes

Navigate This Article