|This paper invokes the idea of an organic film over an aqueous aerosol to explain observed low ammonium to sulfate molar ratios, despite excess gas phase ammonia. The paper has substantially changed since the first version, but again these authors chose to analyze aerosol pH effects through the use of molar ratios, which has been shown to be problematic. |
The paper now focuses on the observation that both currently used thermodynamic models, ISORROPIA and E-AIM tend to predict higher molar ratios than what is observed. They have added some interesting plots (e.g., Fig 1 is quite nice) and arguments to support the idea that the lower molar ratios may be due to the organic film. Some issues should be considered and clarified in the text prior to publication.
A major issue is clarifying the logic of this hypothesis and its implications. If NH3 is not in equilibrium with the particle phase, the premise of this work, and given it is the most important fine mode base (in this case, ANS, the only base) it follows that the thermodynamic models cannot be used, as they assume equilibrium. The model predictions are then incorrect, including pH etc. Of coarse it raises the issue how the models do so well predicting actual NH3/NH4+ partitioning over a wide range of ambient conditions, something that should be explained (more on this below). Because this has very large implications, impacting many already published paper, this implication should be very clearly stated in the paper.
A second major point is the authors should explicitly state how the organic film impedes the uptake of NH3, is it through a low accommodation coefficient, or maybe low solubility of NH3 in the organic film? A curious thing is that H2O and NH3 have nearly identical molar masses making many of their properties similar (diffusivities, etc). How these two molecules could behave so differently when interacting with the film should be directly addressed. More details below.
1). A simpler explanation for the model vs measured molar ratio discrepancy could simply be that modeled pH vs observed molar ratios respond differently to treaing the fine mode as a bulk property. When averaged over all sizes, observed molar ratios may be sensitive to this assumption, whereas actual pH less so. It has already been shown that molar ratios and pH are not necessarily related in a simple way. This possible explanation for the molar ratio discrepancy (or this complication when interpreting the data) should be noted in the paper.
2) Issues with imprecise statements. Eg, in the Abstract, lines 18 to 22 (and other related discussions in the paper, eg first sections of Intro.), regarding ammonia concentrations and molar ratios. The statements are based on the extreme end-members of processes that are asymptotic (E-AIM solution in Fig 1), and the ambient condition is between the two extremes.
Line 29-30 statement that molar ratios decrease with decrease sulfate is incompatible with theory is, I believe, not correct, it is predicted by the model (see more on this below).
3) The authors assert the film will lower the rate of NH3 uptake so that equilibrium is not achieved since the time scales to reach equilibrium become long. This would seem to then invalidate the use of the thermodynamic models in general since the model assumes equilibrium for all semi-volatile species, including water vapor. To argue that an organic film affects NH3 mass transport dynamics, affects on other species should also be considered. The thermodynamics is a multi-component system, the various species behavior being inter-connected. For example, if NH3 time scales to reach equilibrium become large, what happens to other key species, such as H2O, HNO3, etc interacting with the organic surface. See discussion below on diffusivities of various species through an organic film affecting reactive uptake. The conclusion from below is that water will also have a similar time scale to reach equilibrium as ammonia, by the mechanism proposed here, and nitric acid will take about 2 times as long to reach equilibrium. How this will be handled in the CTM and how this is reconciled with existing data and publications on water uptake of particles, etc, should be specifically discussed, along with a detailed analysis of exactly how an organic film will affect reactive uptake. That is, exactly what is the mechanism that slows NH3, but not other species? Maybe it is differences in accommodation coefficients. If so, is this reasonable given measured accommodation coefficients? It should be explicitly stated what the possible physical explanation is for this proposed resistance to NH3 uptake.
4) pH matters, not molar ratios, when concerned about acidity effects. The goal of all this work, as noted by the authors in the abstract, is to predict the effects of aerosol pH on other aerosol properties or processes. Eg, from the abstract, where it states that … uptake of ammonia has important implications for aerosol mass, hygroscopicity, and acidity. It also states that … decrease sulfate aerosol will not have the co-benefit of suppressing acid-catalyzed secondary organic aerosol (SOA) formation. The key is aerosol pH, not molar ratios. As already noted, a reported analyses [Guo et al., 2016; Guo et al., 2015], suggests that in terms of partitioning of semi-volatile species, ISORROPIA works fairly well. That is, the models provide accurate predictions of properties of interest related to pH, something that molar ratios cannot do, as far as I can tell.
5) One aspect the film achieves, according to this work, is that in the CTM model NH3 is predicted better (in addition to molar ratios, which is discussed above). Given the large uncertainties in NH3 emissions, could this simply be fortuitous? How robust a test of the model is the use of measured versus predicted NH3? Accurate prediction of gas particle partitioning of semi-volatile species would seem a better approach. For example, can the CTM accurately predict NH4+/NH3, HNO3/NO3- partitioning, as has been done with ISORROPIA (see various references)?
6) An explanation for the issue of decreasing trends of sulfate and molar ratios discussed in this manuscript is given elsewhere [Weber et al., 2016]. It can simply be explained by ammonia volatility, which should be pointed out in this paper.
Mass Transport Limitations in Aqueous-Phase Chemistry (the following is based on Seinfeld and Pandis, 2nd Edition, 2006, Chapter 12, Section 12.2).
In trying to understand the physical details associated with the proposed interaction between NH3 and an organic film, the following analysis of characteristic times for equilibrium was undertaken, following the argument put forward in the paper. A comparison is made between NH3 and H2O since they have nearly identical molecular weights.
If the rate-limiting step for mass transfer of a gas, say NH3 in this case, from the gas to the particle surface and then through the organic film to the bulk water of the particle interior, (ultimately leading to conversion of NH3 to NH4+), is gas transport through an organic film on the perimeter of the particle, the characteristic time (τ) for the system to reach equilibrium will be in proportion to:
τNH3 α Rp2/(γ DNH3-aerosol), (from S & P equation 12.61)
where Rp could be thought of as the thickness of the film, γ the accommodation coefficient and DNH3-aerosol the diffusivity of NH3 in the organic layer. If we assume that the gases of interest will all have fairly similar accommodation coefficients, all being fairly sticky, and effective Henry’s law constants (similar solubility in the OA phase) the time scales comes down to diffusivity (S&P Eq 12.62), or roughly the ratio depends on sqrt(molecular weight) of the solute (Eq 12.61).
Water (18 g/mole) has a very similar molecular weight as NH3 (17 g/mole) so time scales to reach equilibrium for these two species should be roughly equal. The ratio of equilibration time scales of NH3 and H2O due to the organic film is very roughly then = sqrt(18/17) = 1.03, or approximately 1. Thus, with similar accommodation coefficients and solubility in the OA, if NH3 is not in equilibrium due to the film, water is also not, and will behave very similar to NH3. This would contradict many studies demonstrating that in the ambient atmosphere water is in equilibrium, thus NH3 should also be, by this analyses, or the assumptions are wrong.
Consider a different semi-volatile species, say HNO3 (molecular weight=63 g/mole). In this case the time scales for equilibrium for HNO3 relative to NH3 will be sqrt(63/17) = 1.9. Thus nitric acid takes about 2 times as long to reach equilibrium. How can this be reconciled based on ambient data showing good agreement with measured partitioning of nitric acid and nitrate under most conditions (eg, [Guo et al., 2016])? How would this be included in the overall CTM?
Note that all other semi-volatile species of importance when assessing aerosol pH (e.g, HCl MW=36 g/mole, organic acids…) will have larger molecular weights than NH3, so it will take even longer for them to reach equilibrium. Maybe this all can be explained by differences in mass accommodation coefficients? In fact, it seems the only plausible way.
If possible, the authors should attempt to explicitly state in the text by what mechanism the NH3 is selectively impeded, as these could be measured to test the hypothesis of the paper. Overall, the situation is getting very complicated relative to a straight thermodynamic analysis (assumes equilibrium) that appears to agree with observations of liquid water content and partitioning of semi-volatile species sensitive to pH.
Guo, H., A. P. Sullivan, P. Campuzano-Jost, J. C. Schroder, F. D. Lopez-Hilfiker, J. E. Dibb, J. L. Jimenez, J. A. Thornton, S. S. Brown, A. Nenes, and R. J. Weber (2016), Fine particle pH and the partitioning of nitric acid during winter in the northeastern United States, J. Geophys. Res. Atmos., 121(17), 10,355-310,376.
Guo, H., L. Xu, A. Bougiatioti, K. M. Cerully, S. L. Capps, J. R. Hite, A. G. Carlton, S.-H. Lee, M. H. Bergin, N. L. Ng, A. Nenes, and R. J. Weber (2015), Predicting particle water and pH in the southeastern United States, Atmos. Chem. Phys., 15, 5211–5228.
Weber, R. J., H. Guo, A. G. Russell, and A. Nenes (2016), High aerosol acidity despite declining atmospheric sulfate concentrations over the past 15 years, Nature Geoscience, 9(10.1038/ngeo2665), 282-285.