Milsom et al. characterized the influence of ozone exposure on oleic acid films using quartz crystal microbalance (QCM), Raman spectroscopy, and white light interferometry (WLI) techniques. From the measured decreases in resonant frequency and dissipation factor of the oleic-acid-coated QCM sensors, the authors infer that oleic acid was oxidized and the droplet density increased; in separate experiments, the decay rate of oleic acid droplets was faster than that of mixed oleic acid / sodium oleate droplets. They observed evidence of coagulation and crust formation from the WLI sample images. The combination of techniques used here is proposed as a low-cost, field-deployable method for the measurement of the oxidation kinetics of other compounds.

General Comments

1. In complex/ambient samples (in the authors’ terminology, “real environmental films”, L220), the level of unsaturation and the corresponding ozone reactivity are likely to be much lower than those containing >50% oleic acid as studied here. Reducing the oleic acid fraction from 100% to 50% already reduced the ozone reactivity by a factor of 10 (Figure 3). What is the measured change in resonant frequency / dissipation factor of, for example, films containing 10%, 1%, and 0.1% oleic acid mixtures? How about for bare QCM sensors with no added films? There is also no discussion or method demonstration of controlled exposure of films to, for example, hydroxyl radicals which are a much less selective oxidant than ozone. It is not clear to me what the lowest sample ozone reactivity is that can be meaningfully measured with this technique, and because of that, its potential application to the measurement of oxidation kinetics in samples that don’t have significant levels of unsaturation (C=C bonds) seems limited or at best unknown.
2. In my opinion, the scope of the kinetic analysis is also limited. It seems like it should be possible to calculate additional - and arguably more useful – kinetic data products than the first-order loss rate of oleic acid; for example, I think the reactive uptake coefficient of ozone could be calculated from the first-order decay rate observed with QCM, the droplet sizes obtained from WLI samples, and the KM-SUB model. Please calculate the corresponding uptake coefficients for the results shown in Figures 2 and 3 and discuss these values in the context of previously measured ozone uptake coefficient values for pure and mixed oleic acid aerosols. Also, discuss any associated limitations in the technique that may prevent accurate retrieval of these values, as applicable.

Minor/Technical Comments

1. L56 – Suggest changing “by an order of days” to “by several days” or similar
2. L104 – In “the sect. S1”, delete “the” and capitalize “Sect.”
3. L130 – Suggest deleting “Only”, add “nonanal and nonanoic acid are the only oleic acid ozonolysis products known…”
4. L166 – change “A” to lower case
5. L189 – in the context of this discussion/figure, what are the specific morphological features define a “crust”?
6. L191 - It would be useful if a label/annotation could be added to Figures 5b and 5d to indicate where the authors think a crust formed - to the untrained eye, this may not necessarily be obvious.

The manuscript by Milsom et al. presents kinetic measurements of oleic acid ozonolysis using a low-cost quartz-crystal microbalance with dissipation (QCM-D). The authors observed decreases in both resonant frequency and dissipation. The retrieved first-order reaction rate agrees well with Raman spectroscopy data. Overall, I think this is an interesting study, which well demonstrates that the low-cost QCM-D instrument can be used to investigate atmospherically relevant multiphase reaction kinetics. However, I do have a few concerns about the interpretation of the results.

Line 122-128: “This observation suggests an increase in mass per unit area on the QCM crystal surface via the Sauerbrey equation, which states that the mass per unit area deposited on a QCM crystal is inversely proportional to the crystal’s measured resonant frequency…

The apparent increase in mass per unit area observed during ozonolysis could be due to an increase in film density. This has been observed previously for oleic acid ozonolysis, where density increases from 0.89 to 1.12 g cm-3 with increasing ozone exposure, presumably due to ozonolysis products having higher densities…”  

My concern is that the initial film of the liquid oleic acid film may have a large dissipation, therefore the QCM frequency does not follow the linear relationship of the Sauerbrey equation. In this case, the decrease in resonant frequency does not necessarily mean an increase in mass per unit area. For example, Chao et al. (ACS Omega, 2020) reported that the resonant frequency of QCM increases during the solid-to-liquid phase transition induced by the deliquescence of salts, although the actual mass largely increases. It seems to me that the QCM-D results reported in this study behave just like the reverse process. I think the phase transition from liquid to solid-like state during oxidation played a major role in driving both f and D changes, and the density change only had a small effect.

In addition, the authors should consider adding baseline measurements for bare sensors, and report delta D and delta F data relative to the baseline. The data should not be interpreted with Sauerbrey equation if the D value is large relative to the baseline. Also, reporting data from different overtones can help to verify of the Sauerbrey equation is applicable. Alternatively, the authors could try measurements with lower sample mass, such that the Sauerbrey equation is valid and the mass change during oxidation can be determined.

Reference:

Chao, H.-J., Huang, W.-C., Chen, C.-L., Chou, C. C. K., and Hung, H.-M.: Water Adsorption vs Phase Transition of Aerosols Monitored by a Quartz Crystal Microbalance, ACS Omega, 5, 31858-31866, 10.1021/acsomega.0c04698, 2020.