General comments

Panji and Isaacman-VanWertz present a novel inlet that enriches concentrations of organic gases, allowing for detection of trace gases by low-cost sensors. Not only is this a generally helpful method but the presented material, equations, and experiment setup is straightforward and can easily be replicated, understood, and modified. My comments are mostly minor, with slightly more moderate comments about applicability to on-line measurements. I recommend publication after these comments are addressed.

Specific comments

Line 25-26: Can you elaborate on this a little more? Does the “majority of the analytical instruments” include these low-cost sensors or are you referring to the more expensive ones in the preceding sentences?

Line 35-36: This sentence as written does not tell us much. What are the major findings/limitations of this reference relative to what you are presenting here? It would be helpful to set up what advancements you are making for preconcentration of on-line sampling.

Line 134: What is this airtight material, its dimensions, the sub-atmospheric pressure you get it down to, and the pump you used?

Line 163-165: Based on the given reference you should be able to estimate this change in permeation rate, right? Can you provide that here for your molecules?

Line 186-187: This is related to the above comment. This can be estimated or even measured. Can you provide an estimate to help put a range on your error?

Figure 4: Are these the 1σ of triplicate experiments? If so, please state that in the caption.

Line 220-221: Why is this an appendix section and not just the last part of the results and discussion section? You set up your introduction to say that low-cost sensors have a higher need for preconcentration for on-line sampling and here is an example of that. I would suggest including it in the main text. 

Appendix B: In your introduction you say that low-cost sensors fall short in measuring in the ppt range but your setup for on-line, real-time sampling does not address that. At best you can achieve a 1 bar pressure differential assuming you are drawing in air at ambient conditions without a pressure drop and your container is at vacuum. This leads to ~100% enrichment for more trace VOC according to Table A1 and Figures 4 and 5 which won’t be detected by your miniPID2 if average concentrations are hundreds of ppt. If you decrease the sample flowrate or increase the inlet length your latency in detection may be too long and impractical. Is there a calculated estimate you can provide for an increase in enrichment through heating? That would be helpful for others who want to modify this.

Line 232-235: This is related to the above comment, but I see the flowrate of sample into a detector as a major limitation of this method since this method works best with <2 sccm of sample flow. For many VOC-detecting instruments this is an issue since they normally introduce flow on the orders of hundreds of sccm and therefore you would need to dilute any enrichment you gain. Do you have any examples of instruments with sccm-level flowrates that would work with this method?

Line 249: There is not a large change in enrichment with concentration but there is a notable change. Do you have an explanation for why increasing concentration increases enrichment? If these changes are all within error, can you state that and what the error range is based on the assumptions you mention throughout the text?

Line 250-252: What is meant by testing at ambient conditions? Isn’t that what this test was representative of? Or do you mean in conditions where the concentrations may fluctuate on a shorter timescale?

Technical corrections

Line 11-12: I suggest changing “man-made” to “human-made”.

Line 68: “… so it is an ideal…”?

Line 73: can you briefly explain this unit conversion to get 3.35 x 10^-11?

Figure 6: Some labels are boxed. You should revise to remove the boxes.

Figure B2: In the last line of the caption “benzene” is misspelled.

The study by Panji and Isaacman-VanWertz presents an innovative technique for enriching organic gases, utilizing a commercially available pre-concentration inlet system. The authors have theoretically characterised and validated the inlet design parameters, operating conditions, and inlet efficiency, thereby enabling the predictable enrichment of a majority of atmospheric gases. The approach is intriguing and valuable, with the results demonstrating its potential as a pre-concentration system for low-cost sensors. The paper is well-articulated, and the methodology is clearly described. However, I believe that any study introducing a new sampling approach must demonstrate its applicability with a real-life example, a section that is currently missing from this manuscript. I recommend acceptance of the paper upon addressing the below comments.

General comment

A real-world application of this method is crucial for a technical study of this nature. A simple demonstration, such as a diel cycle of ambient air sampling, could significantly enhance the paper's impact and convince readers of the method's practicality.

Specific comments

L7. Could you elaborate on the choice of such low flow rates for your experiments? While the enrichment is evidently more efficient within this range, these low flows are not common in VOC sampling and in fact, they can be challenging to work with. Additionally, it would be beneficial if you could discuss the uncertainties associated with measuring flows at these low rates.

L26-35. You may want to add the cryogenic trapping in empty stainless steel tubes, which would however require very low temperatures and the use of liquid nitrogen (Apel et. al. 2016, Bourtsoukidis et al., 2017)

L114-115. Ambient air is typically characterized by a complex atmospheric mixture, particularly at low mixing ratios. Could you provide more details on how you plan to separate individual compounds in such a complex mixture? This information is essential for understanding the real-world applicability of the system.

L130-131 & L166-170. In your experiments, you've used pressurised air for sample delivery. Could you discuss the potential limitations this might pose for atmospheric sampling? Are there specific scenarios or applications where this method would be particularly useful or challenging? For instance, I understand that this method might be well-suited for canister analysis.

L153 & Figure 3. There seems to be a significant delay (approximately half an hour) before the desired enrichment is achieved. However, in L174 you state that the residence time was 1 minute at 1 sccm. Could you provide some insight into the reasons for this discrepancy? Is this delay a result of the changes in the experimental setup?

L194. From the figures, it appears that this 'slightly higher' difference might be statistically significant. Please elaborate in further discussion.

Line 217: Given that Flame Ionization Detectors are highly sensitive with detection limits in the lower ppt range (down to 1-2ppt), the rationale for using such high concentrations in your experiments is not clear from this statement. While the demonstration of functionality is understood, the need to pre-concentrate such high concentrations could be better justified. It's worth noting that low-cost sensors using alternative methods might be more practical and easier to use at these high concentrations.

Lines 234-235: The impact of temperature on the method's effectiveness seems to warrant at least more detailed discussion. Could you elaborate on how temperature variations, such as those occurring within a diel cycle, might affect the performance of your method? This is particularly important in understanding the method's robustness in real-world conditions.

References

Apel, E. C. and UCAR/NCAR – Earth Observing Laboratory: Trace Organic Gas Analyzer (TOGA) for HIAPER. UCAR/NCAR – Earth Observing Laboratory, https://doi.org/10.5065/D6DF6P9Q, 2016.

Bourtsoukidis, E., Helleis, F., Tomsche, L., Fischer, H., Hofmann, R., Lelieveld, J., and Williams, J.: An aircraft gas chromatograph–mass spectrometer System for Organic Fast Identification Analysis (SOFIA): design, performance and a case study of Asian monsoon pollution outflow, Atmos. Meas. Tech., 10, 5089–5105, https://doi.org/10.5194/amt-10-5089-2017, 2017.