|The authors have done a very thorough job of responding to the reviewer comments, as well as the comments by Paul Ziemann, including performing a new experiment. This paper should be published once the authors address the minor issues below.|
The addition of the CIMS measurements prior to when the lights were turned on in Fig. 3 is very useful. However, it seems as if Fig. 3 is showing the non-background subtracted signals. The authors should note this in the caption.
Regarding the comparison between the current study and the work from Ziemann and co-workers, there remains an unfortunate discrepancy. It is very difficult to reconcile the results from the two studies, and there is overlap in the vapor pressure ranges considered, even if the functional group types here are more diverse. Certainly, this issue is not going to be worked out from this single study. Nonetheless, in the authors response they note that they tentatively attribute the discrepancy to differences in mixing between the two chambers. I think that there are two main arguments against this being the predominant aspect.
1. The theoretical loss rate of gases, absent accommodation limitations, is governed by mixing and diffusion. This has not been measured here for the gases, but this group has previously reported wall loss rates for particles as a function of size. These particle wall loss rates provide constraints on the potential for differences in mixing to be responsible for the differences. Although the Caltech chamber may operate in a static mode (no active mixing), the loss rates of particles are nonetheless substantial, even for the smallest particles for which sedimentation is not an issue. I believe the consideration of these particle wall loss rates places bounds, at least approximate ones, on mixing effects. The particle loss rates reported in Loza et al. (2012) for the smallest particles sampled (~10 nm) were on the order of 10 4 s 1. This is much larger than many of the values observed in the current study, suggesting that diffusion (i.e. mixing) is not a limiting factor.
2. Although the authors rightly note that Ziemann and co-workers use “active” mixing, the action is very limited in scope. Specifically, it appears that after injection into the Ziemann chamber, a small fan is turned on for a few 10’s of seconds before it is turned off for the remainder of the experiment. Thus, the “active” mixing is only at the very, very beginning of the experiment. In the current study, the authors fill the bag and allow the system to sit for ~1hr prior to turning on the lights. Mixing, i.e. homogenization, will occur during this period thus that at the point when the lights are turned on the system may, in fact, be reasonably well mixed. As such, the differences in the experiments may not be as large as one might think.
My point is not that one of the studies is right or wrong, but simply that I don’t think that differences in mixing will ultimately prove to be the final answer to the disparity between these studies. The authors have obviously done their due diligence in that they performed an additional experiment as recommended by Ziemann and this certainly provides further confidence in the results from the current study. In any case, I would suggest that the authors update their “tentative attribution” statement. Hopefully this discrepancy will be fully resolved through future studies, as this is an important topic.