The study “Impact of water uptake and mixing state on submicron particles deposition in the human respiratory tract (HRT): Based on explicit hygroscopicity measurements at HRT-like conditions” by Man et al., examines the impact of considering hygroscopic properties of ambient aerosol into the deposition rates of particles in the HRT. It therefore uses measurements of particle size distribution (PSD) for hydrophobic and hygroscopic particles performed in the North China Plain at a RH of 98 %. The study shows the impact of a changed PSD due to hygroscopic growth of ambient aerosol. The results show the sensitivity on aerosol deposition in the HRT to PSD with a decrease in deposition around 24 % when considering hygroscopic growth of the aerosols in the HRT (with large effect for hygroscopic particles and rather little for hydrophobic).  

The study addresses an important topic and provides a promising analysis based on measurement data. In the current version of the manuscript there are no substantially new findings, however, the combination of measurements of size resolved hygroscopic and hydrophobic aerosol with HRT deposition modeling is a nice and interesting study. E.g. Ching and Kajino (2018) show comparable results for hygroscopic changes in the PSD. The study will profit from a more detailed comparison to previous studies, both on ambient aerosol properties and the HRT deposition. This comparison will also allow the authors to clearly define the novelty of their study. Furthermore, the study will benefit from some careful proofreading.  Nevertheless, the study is in my opinion worth to be published after addressing the below listed points.

General comments:

A sensitivity analysis on the different deposition mechanisms would be helpful. E.g., by turning off single deposition mechanism in the model to estimate the relative importance of the shift to larger sizes in the PSD due to hygroscopic growth. Also, an extension of the discussion of the results taking the different deposition mechanisms into account will be helpful.

Are there differences between the ICRP and the MPPD models? Why the authors chose the MPPD model rather than ICRP?

How does the K parameter change between previous measurements and studies with lower RH and this study with RH = 98 %. How do the results compare to the previous studies with lower RH measurements? What is the expected impact of the difference between the measurements at 98 % RH and the actual nearly saturated conditions in the lower HRT?

The PSD measurements are performed to an upper limit of 800 nm, but Fig. 1 shows values until 1 µm are these a result of a fit?

May I overread this information, but what hygroscopicity is assumed for the calculation of the adjusted PSDs (Fig. 1 and Fig. 5)?

The deposition dose which is one important quantity in this manuscript is highly sensitive to the exposure time. Since children are associated to a shorter exposure time the lower dose seems to be explained by that simple and rather arbitrary quantity. The deposition rate however, is not influenced by this quantity and thus may serve better to compare the health impacts of sub micrometer particulate air pollution. The exposure time seems quite high in my eyes (for adults and elderly around 4 hours of “resting” outdoors.

Is the time of hygroscopic growth considered in the MPPD modeling? Ching and Kajino (2018) pointed the importance of time in the hydration process of particles in the HRT out. The timescale until the particles reach their equilibrium size depending on RH can range in the scale of few seconds (Chuang 2003) and thus may play an important role in the upper respiratory tract.

Are in Fig. 5 shown deposition rates for “resting” (table 1) only? It would be interesting to combine different human behaviors with the ambient pollution (e.g. physical exercise in the evening hours when pollution is high).

Specific comments:

Fig.1: The figure would profit from including the range of the PSD. Also in the appendix a comparable figure for day and nighttime would be interesting for better understand the ambient aerosol conditions.

Fig.2: The figure would profit from including the range of the deposition fraction (percentiles or SD). I would appreciate a corresponding figure for children and elderly in the supplement.

Fig.3: Is it possible to include the red line (as indicator for the hygroscopic and hydrophobic fraction) in all columns?

Fig.4: Is for this figure the average PSD (Fig. 1) or the daily PSD (corresponding to Fig. 5) used? How does it behave with the particle fraction above 1 µm in the head region? Fig. 2 shows an increasing deposition fraction for larger particles and there is a significant fraction of particles above 1 µm after hygroscopic growth.

Fig.5: NO, CO, BC and OH should be introduced with the corresponding measurement method in the methods section. Is it possible to show a fraction of the BC mass to the mass of hydrophobic aerosol (calculated from the PSD)? This will be helpful in understanding the displayed data. The figure and interpretation would benefit from including the median deposition rate without considering hygroscopic growth as done in the other figures. For interpretation also, the total particle number concentration should be added and included into the discussion.

L40: This sentence reads as life expectancy declines, which is not the case (Haidong et al., 2019).

L.45: A reference for this statement will be beneficial.

L.74: this statement is quite general. There are also many measurements of hygroscopicity with e.g. CCN counters.

L.161: Is deposition and clearance not the output parameter of the MPPD model?

L.165: The particle range in model is until 1 µm but PSD in Fig. 1(b) reaches up to 1.2 µm. What is the expected effect of neglecting the tail of the PSD?

L318 and Fig.2: What is the potential explanation for the underestimation of particles below 20 nm in P?

L.347: The conclusion from the presence of hygroscopic newly formed aerosol is somehow confusing, since the study points out that hygroscopic aerosol is less effective deposited in the HRT. The factor is the larger number concentration, not the aerosol hygroscopicity (as mentioned above including the total number concertation in Fig.5 will make this clearer). I also wonder why the aerosol disappeared in the morning hours (e.g. Fig.S1 2014/06/28 around 8 am)?

L.356: Is the low boundary layer height meant as a general feature of the PBL or is it somehow measured or modeled during the campaign? A reference here would be nice.

L.363: This conclusion cannot be drawn that straight forward. What about the other aerosol during the peak concentrations of BC mass? How does the K value for BC compare to literature studies on the hygroscopicity of BC? An indication in Fig.S2 which are the “5 % highest BC periods” chosen and also the BC fraction compared to all aerosol mass will be beneficial.  

L.367: I do not see the benefit of the discussion; the PSD and PNC are not the same for freshly emitted and aged aerosol. Further, hydrophobic and hygroscopic particles can be emitted from the same source.

L.388: The discussion on the atmospheric oxidation capacity could be more addressed in the results section to strengthen this conclusion.

L.390: What are the strong primary emission sources over night?

L.392: This study does not provide insights about aging state, it shows the difference between hydrophobic and hygroscopic particles. Thus, I recommend to rephrase the sentence from “aged aerosol” to hygroscopic.

L.396: I do not see this aspect covered by the manuscript. Maybe a rephrasing closer to the actual research performed will help.

L.399: Data availability – maybe the authors want to consider a publication of the data on an open accessible research data repository after publication.

Technical comments:

There are some citations doubled in single sentences. Some of the cited literature seems to be not the primary literature.

L.140: Consider to reconstruct the sentence.  

References:

Chuang, P. Y. (2003), Measurement of the timescale of hygroscopic growth for atmospheric aerosols, J. Geophys. Res., 108, 4282, doi:10.1029/2002JD002757, D9.

Wang, Haidong et al. (2019), Global age-sex-specific fertility, mortality, healthy life expectancy (HALE), and population estimates in 204 countries and territories, 1950–2019: a comprehensive demographic analysis for the Global Burden of Disease Study 2019, The Lancet, Volume 396, Issue 10258, 1160 - 1203

In this work the authors describe measurement of the hygroscopic growth of externally mixed particles from the North China Plain. They use these data in conjunction with a lung deposition model to predict the effect of hygroscopic growth on deposition in the respiratory tract. The results show that in total, dose was reduced when hygroscopic growth effects were considered as the more numerous smaller particles, that deposit via diffusion mechanisms, deposited less effectively. Variations were seen across the size range, with smaller particles showing a reduced likelihood to deposit, while larger particles were more likely to deposit.  

Overall, this paper goes some way towards showing the importance of considering hygroscopic growth, but the extent of new insights is limited. The effects are reported to be rather small so an improved sensitivity analysis and consideration of uncertainties is needed to validate and support the conclusions. Some specific points towards this are detailed below:

·      Deposition fraction is on a particle number basis, and the conclusions connect the dose with the number of particles. The authors should considering reporting dose on a mass deposition basis, which will significantly increase the contributions of the larger particles on deposited dose.

·      Does the lung deposition model change the density of the particles as they grow due to water uptake? A density of 1.5 g/cm3 is high for hygroscopic particles at >90% RH. I suggest a sensitivity analysis be performed to compare the difference in deposition for 1.0 and 1.5 g/cm3 particle distributions.

·      How was the dry size of the particles determined in the hygroscopic growth measurements? Were any shape correction factors considered?

·      How accurate is the RH measured in the HTDMA? How stable is the RH? At the high RH of these measurements, even fractions of a % of RH can lead to significant changes in the size of the particles and will introduce uncertainty in the results. 

·      On line 103, HH-TDMA is referred to – what does the second “H” stand for?

·      Line 84 – a constant value of kappa with RH does not indicate an ideal solution. It indicates that the effective molar volume of the solute does not vary with RH.