We thank the reviewer for the valuable comments which have helped us improve the manuscript. We provide here answers and mention the changes that have been made to address the referee’s concerns. The referee’s comments are in normal font while the replies are given in bold.

Reviewer 1

Hellen et al., in their paper “Measurements of C10-C15 biogenic volatile organic compounds (BVOCs) with sorbent tubes”, investigated the performance of different configurations of adsorbent tubes in sampling BVOCs with respect to several parameters, including adsorbent type, tubing material, temperature, storage and ozone filtration techniques. They also cross-compared their results with a subset of measurements made at another research facility in a different institution (i.e. IMT Nord). The paper is well organized, properly structured and well written. I was also pleased to see that the authors thoroughly considered recent papers on this topic (as there are several!) and effectively identified areas of improvement to address through their paper. The topic of their research is quite relevant for offline measurement groups, especially those making ambient gas-phase measurements. Hence, I should think that this paper is very informative for relevant members of the scientific community, and also for those groups who may be considering using adsorbent tubes in future offline studies. However, I do have some concerns. I suggest publication of this work after these concerns are properly addressed:

 My major concern is regarding the provided explanations for the observed differences in analyte recoveries between Tenax TA and Carbopack for different experimental conditions. I think that the explanations are a bit superficial and do not yet fully get to the scientific detail that will inform a confident decision making of future users of these techniques. For example, in line 354, the authors note that none of the studied compounds broke through the Carbopack B tubes, while some did in case of Tenax TA. At the same time, Figure 2 presents those analytes collected with Carbopack B tubes showed maximum losses over a 1-month storage period, while the losses were comparatively lower with Tenax TA. This seemed a bit counterintuitive to me. Yet, assuming that the results are all consistent, the authors should be able to provide some scientific explanation for why species would not break through an adsorbent material during sampling but show relatively higher losses from the same material when stored for a long period.

-Carbopack B is a stronger sorbent for hydrocarbons than Tenax TA. Therefore, breakthrough volumes of Carbopack B are expected to be higher. However, storage stability does not only depend on the strength of the sorbent material, but also on the sampled compounds volatility and reactivity. In Carbopack B sorbent, some of these BVOCs seem to isomerize when interacting with the sorbent and therefore their stability in the sorbent is lower even without storage. As it was shown in the section 3.3, comparison of Carbopack B tubes with other tubes and online sampling showed that Carbopack B seemed to suffer on some losses already without long storage time. A comment on this was added to the manuscript into section 3.1 on lines 351-353 (new version of the manuscript). While it is possible to keep sampling volumes low enough for Tenax TA samples, it is not possible to avoid deviations found for the Carbopack B tubes. Therefore, in the ‘Summary and conclusions’ -section we recommend Tenax TA as a sorbent for the studied BVOCs.

In lines 333-334, isomerization of terpenes is suggested to be a source of high recovery of camphene from Carbopack tubes. Figure 2 shows that such high recovery of camphene was not observed in tubes other than Carbopack, suggesting material dependent behavior. Please explain this in more detail. Does Carbopack facilitate the isomerization process in some way?

-We do not have detailed knowledge on the reactions occurring in the Carbopack B, but for all Carbopack B samples some of the monoterpenes (e.g. β-pinene) suffered losses while camphene was formed at the same time. This was especially clear in the intercomparison studies, where even if camphene was not flushed into the tubes, camphene peak was found from all Carbopack B tubes. Contrary to Carbopack B tubes, no camphene peak was observed in Tenax TA tubes when the same sample procedure was followed.

In line 350, authors note that the breakthrough volumes were lower for self-packed tubes at 30% RH. Indeed, table 4 shows that the differences can be as high as 4-5x. This is an important information that warrants detailed explanation because self-packing can be impacted by biases pertaining to human activity (e.g. packing density achieved). Are the authors certain that this (or similar) difference would exist if another group replicated the self-made tubes? It is suggested that newer material in commercial tubes could explain the higher retention, but I would suggest expanding and providing a 2-liner scientific description of analyte retention in an adsorbent material. It seems that the retention tests with 70% RH were done in sequence after the 30% RH tests (lines 351-352), and then results were more comparable (Table 4). However, I imagine that the commercially packed adsorbent material would not age enough in sequential tests or does it? Also column 2 of table 4 shows a major drop in breakthrough volume when going from 30% to 70% for the commercial Tenax TA tubes. However, a drop of this magnitude is not observed in column 3 for self-packed tubes with the same material. Please provide a brief explanation for this and for the role of humidity in retention efficiency of different species.

-We added a description of analyte retention in an adsorbent material and more discussion on the impact of humidity into the manuscript section 3.2 on lines 369-387 (new version of the manuscript): ‘ For the commercial tubes breakthrough volumes for RH 70% were lower than for RH 30 %. The impact of RH was not detected for self-packed tubes. Retention of the compounds on the sorbent is based on the competition of the analyte with other molecules in air. Even though Tenax TA is a hydrophobic sorbent at high humidity some water may still be adsorbed, or water may condense onto the tube impacting the sorption of other compounds. In earlier studies impact of RH on Tenax TA retention was not clear since often Tenax TA was studied in multibed sorbents. In their breath sample studies Wilkinson et al. (2020) found that high relative humidity during sampling generally reduces the ability of Tenax TA/Carbograph sorbent tubes to capture volatile compounds. However, since the impact was more pronounced for the less volatile compounds with higher breakthrough volumes, they speculated that impact is other than breakthrough. Ho et al. (2017) found significant impacts of RH on retention of C₂-C₅ aliphatic hydrocarbons in multisorbent (Tenax TA- Carbograph 1 TD- Carboxen 1003) tubes. The breakthrough volumes for C₂ aliphatic compounds were reduced by 13 – 22fold under 90 % RH and the main impact was expected to be through the less hydrophobic sorbent Carboxen 1003.  For >C₅ compounds no breakthrough was detected. In the studies of Maceira et al. (2017) humidity problems were demonstrated with carbon-based tubes, while Tenax-based tubes did not display any influence. Our tests with commercial tubes at 30 % humidity were done for the totally new tubes and this could also explain the higher retention. In the tests under 70 % humidity conducted after the tests of 30 % humidity, any difference between commercial and self-packed tubes was not that clear. However, tests with ReGaS2 generator with commercial tubes at RH 30% conducted later show also bit higher breakthrough volumes, but not as high as for the first tests (Fig. 3), indicating some influence of RH. Increasing the temperature of the sorbent tube to 10 °C above the temperature of the air sample could reduce interferences of the humidity (Karbiwnyk et al. 2002).’

 Figure 2: The error bars on the Carbopack tubes are quite broad irrespective of the storage temperature. Interestingly, in some cases, the errors are larger for tubes stored at -20 C. I would have expected reduced losses at -20 C relative to +20 C and thus smaller errors. It would be good if the authors could briefly explain this trend in section 3.1.

-We added to the manuscript into section 3.1 (lines 348-351 in the new version of the manuscript): ‘Even though it would be expected that compounds are more stable at -20 ^oC relative to +20 °C, sometimes even larger deviations were observed in Carbopack B at -20 °C. Some deviations were detected also in the tubes analysed immediately possibly due to incomplete desorption or reactions on the sorbent surface.’

Lines 363-385: This reads like literature review. Further, since your results are consistent with previous work described in these lines (and stated as such in line 383), I suggest consolidating this paragraph more and stating more clearly how your work is novel in the aspect that is the focus of the discussion in this section.

 -We modified section 3.2 (lines 398-408 in the new version of the manuscript) and added a comment that in most earlier studies only a few monoterpenes (a- and β-pinene) have been studied while here several different BVOCs were studied. Modified version: ‘Estimates of breakthrough volumes can be found in the literature, but most earlier studies included only α-pinene, β-pinene and/or limonene (Arnts, 2010; Gallego et. al., 2010; Sheu et al., 2018) while in our study several BVOCs were tested. These earlier results are in accordance with our findings on α- and β-pinene having lower safe sampling volumes in Tenax TA (4–7 L) than limonene (≥ 38 L). In addition to this, our study found that other C₁₀–C₁₅ BVOCs have also higher safe sampling volumes except camphene, which was breaking through at same volumes as α- and β-pinene. In studies on Tenax TA in multibed tubes together with stronger carbon-based sorbents (Veenas et al. 2020, Helin et al. 2020, Komenda et al, 2001), much higher breakthrough volumes (even >24 L) were detected even for α- and β-pinene. Also, this is in accordance with our results on breakthrough volumes in Carbopack B tubes being very high (>160 L) for all studied compounds. Due to lower amount of each sorbent in multibed tubes, especially α- and β-pinene may be breaking through the Tenax TA into the carbon-based sorbent at very low sampling volumes. Behaviour of the compounds in these carbon-sorbents should be also considered as shown by losses and possible isomerization of some terpenes (e.g. β-pinene) in Carbopack B in this study.’

Figure 6: Why do you see up to 20% enhancement in signal when an empty tube is used ahead of the sorbent tube (Figure 1)? I would have expected either the same result or some upstream losses to the surface of the empty tube making the relative difference either negative or zero. Interestingly, the error bars on some compounds with positive differences are bounded in the positive domain (i.e. lower error bound > 0), especially for Carbopack-based measurements. Does this mean all replicates consistently showed enhancement in signal when an empty tube was used upstream? If so, why? All these aspects should be explained in section 3.4.

-This test set up had higher than average uncertainty, which was shown in table 1. This was caused by the fluctuations in sampling flows and higher than normal drifts in MS signal. To explain this we modified the first paragraph of section 3.4 (lines 469-473 in new version of the manuscript): ‘The results show that the relative differences with different tube materials, were between -15 % and +20 % for most compounds. This is expected to be within uncertainty of this test set up. Higher than average uncertainty (Table 1) was caused by the fluctuations in sampling flows and drifts of the MS response.  However, there were clear losses of some sesquiterpenes (β-caryophyllene, β-farnesene and α-humulene) and oxygenated BVOCs (linalool and 4-AMCH) on SS tubes (Fig. 6). For these SS tubes slightly higher amounts of p-cymene and limonene were also detected.’

Lines 521-522: Please explain why the probability of isomerization of beta-pinene would depend on adsorbent material. Similarly, explain the opposite trend for alpha-pinene. This will be useful for future users.

-In Tenax TA no additional terpene peaks were detected. It only happened in Carbopack B. We added a comment on this into section 3.6 on lines 554-555 (the new version of the manuscript).

Lines 534-537: The rationale provided here about the impact of transportation sounds vague, particularly when explaining the 23% difference for alpha-pinene. I would imagine that the tubes were transported in a controlled environment. Please provide more detail regarding the role of transportation in causing this discrepancy. Again, this is important for other users since being able to sample in remote locations is a major advantage of offline techniques and transportation is a key element between sampling and eventual laboratory analysis of a collected sample.

-Samples were transported through regular post without environment controlling (e.g. cooling). We added a comment on differences of the laboratories to the manuscript into section 3.6 and lines 570-575 in the new version of the manuscript: ‘Some differences between IMT and FMI were higher (23 % for α-pinene in Tenax TA) than the calculated uncertainty (7 % for both laboratories, Table 1, Supplement S2). A possible reason is the different calibration method used and their associated uncertainties; FMI calibrated their TD-GC-MS with methanol solutions while IMT used a traceable NPL gas standard to calibrate their TD-GC-FID. Comparisons of liquid methanol standard and NPL gas standard in section 3.3 indicated that this can cause up to 12 % difference for α-pinene results. However, to a lesser extent, transportation through regular post without cooling may have had also impact on the compounds sampled on the FMI tubes.’

Minor points:

Line 100: Suggest renaming “Experimental” to “Experimental techniques” or something similar. “Experimental” on its own is incomplete.

-corrected

Table 1: Font correction required.

-corrected

Lines 320-321: The symbol for degrees should be in superscript. Please check at other places.

-corrected

Line 498-499: Please provide citation or supporting data.

-this sentence has been removed as suggested by the other reviewer

Line 524: Avoid adjectives – “nice”.

-the sentence was reformulated (Section 3.6, lines 558-560 in the new version of the mansucript): ‘When results of FMI and IMT were compared taking into account uncertainties, the results for α-pinene with both sorbents (except 15-min loading with Carbopack B); and for β-pinene with Tenax TA (except 30- min loading, Fig. 9) were in agreement.’

We thank the reviewer for the valuable comments which have helped us improve the manuscript. We provide here answers and mention the changes that have been made to address the referee’s concerns. The referee’s comments are in normal font while the replies are given in bold.

Reviewer 2:

Overall assessment:

Despite the increasing use of online measurement techniques (e.g., PTRMS, online GC), sorbent tubes remain an important tool used by the scientific community to quantify BVOCs (isoprene, monoterpenes, sesquiterpenes, etc.). The authors present results on breakthrough volumes and storage stability for different VOCs and with different sorbent and tube materials. These results are important and should be made available to the wider scientific community. The manuscript is well-written, and the methodology and data interpretation conducted by the authors appear sound. The results are clearly presented in figures/tables and support the conclusions of the study. I would suggest several revisions to further improve the manuscript (see below).

General comment:

Why not also evaluate the measurements of isoprene using sorbent tubes, which is by far the most abundant BVOC (accounting for about 50-70% of global BVOC emissions)?

-Focus of this project was less volatile biogenic compounds, which have not been studied as much as isoprene.  Isoprene measurements with online and offline techniques are better established. Terpene measurements including oxygenated BVOC, are more challenging and up to now few online and offline measurement studies focused on studying this large panel of species. Moreover, it is well-known that the tested Tenax TA tubes are not suitable for isoprene measurement.

Line-by-line comments:

L16: Replace “oxy BVOCs” with “oxygenated BVOCs” for greater clarity.

-corrected

L17: “bornylacetate”: there should be a space between “bornyl” and “acetate”. Please change this throughout the manuscript as well.

-corrected

L24: What do you mean by ±20 °C? Do you mean from -20 °C to +20 °C? Also see Line 555.

-this was corrected to ‘at -20 °C and at +20 °C’.

L50: Maybe also mention PTR-Q-MS (quadrupole mass spectrometer)?

-we changed the sentence into ‘Proton transfer reaction mass spectrometers (PTR-MSs) both quadrupoles and time-of-flights (TOFs)) have been used…’

L68: Change “Most commonly sorbent,” to “Most common sorbent”

-corrected

L104: delta-3-carene is written incorrectly.

-corrected

L107: What do you mean by “Each compound was weighed in 500 mL of methanol”? Please revise this sentence to make it clearer.

-For clarification the sentence was changed to ‘For producing standard solutions ~30 mg of each BVOC was added to the  500 mL of methanol, which was further diluted into six different amount-of-substance fractions (a.k.a. amount fractions).’

L114: The word “online” is repeated twice.

-corrected

L124: Please use proper superscripts for the degree symbol in 320°C and for the units for flow rate.

-corrected

Table 1: What are the LOQs for Carbopack B tubes?

-LOQs for Carbopack B tubes were added to the table 1.

L153: Do you mean “dual stage regulator”?

-yes, this was corrected.

L155: How did you sample the sorbent tubes directly from the outlet of the gas cylinder valve? Did you use a pressure or flow regulator to limit the flow going through the sorbent tube?

- Clarification for this was added to the section 2.2 (lines 160-162 in the new version of the manuscript): ’Sorbent tubes were sampled from the outlet of the gas cylinder valve using a T-connector and a pump (N 86 Laboport, KNF) with the sampling flow of ~100 ml min^-1. The main standard gas flow was kept at ~200-500 ml min^-1 during sampling.’

L214: How did you control the relative humidity?

-We added a clarification on this to the section 2.5 (lines 234-236 in the new version of the manuscript): ‘The air was humidified for desired level by bubbling a fraction of air through the ultrapure water (Milli-Q Gradient, Molsheim, France). The RH was measured with a Vaisala HMI 33 device (probe HMP 35, Vaisala, Helsinki, Finland).’

L215: This sentence in unclear: “Since no blank was detected in blank sorbent tube tests, zero was used as one point.”

-due to unclarity the sentence was removed

Figure 1: What is the model/manufacturer of the sampling pump used to pull the air stream through the sorbent tubes? This should be mentioned somewhere in the methods section.

-this was added to section 2.1

Table 2: “diameterfilters” should be two words.

-corrected

L405: Is there a different way to generate sesquiterpenes of varying amount fractions without using the methanol injection system? It would be nice if you could also perform the linearity tests on sequiterpenes.

-There are no gas standards or reliable permeation tube methods available for sesquiterpenes. However, with our methanol solution injections directly into sorbent tubes under nitrogen flow, we have found similar calibration curves as for the monoterpenes, which gives some confidence for their calibration even though traceability cannot be achieved.

L422: Do you have an explanation/guess for why most of the compounds had severe losses in Carbopack B tubes? Is Carbopack B not a suitable sorbent for analyzing mono- and sesquiterpenes? This should be addressed somewhere in the manuscript.

-we added into Summary and conclusions -part (lines 588-589 in the new version of the manuscript) sentence ‘As shown by the losses of sesquiterpenes, oxygenated BVOCs and some monoterpenes during storage, in comparisons with online sampling and in laboratory intercomparisons, Carbopack B is unsuitable sorbent for the studied BVOCs.’

L495: Can you provide an explanation/guess as to why limonene is affected more than a-pinene and b-pinene?

-Higher reactivity of limonene could be a reason. This was added to the manuscript section 3.5 (lines 527-528 in the new version of the manuscript).

Figure 8 legend: Use Greek letters for a-pinene and b-pinene, or write in full, e.g., alpha-pinene, beta-pinene.

-The legend of Figure 8 is corrected:

L547: Replace “where out of the calibration range” with “were out of the calibration range”.

-corrected

L561: Maybe state which specific BVOCs suffered the greatest losses with MnO2 scrubbers.

-this was added

L585: “were content”: Do you mean “were contained”?

-yes, this was corrected

We thank the reviewer for the valuable comments which have helped us improve the manuscript. We provide here answers and mention the changes that have been made to address the referee’s concerns. The referee’s comments are in normal font while the replies are given in bold.

Reviewer 3:

This manuscript describes a series of experiments characterizing solid adsorbent tubes for the sampling and thermal desorption of biogenic volatile organic compounds.  Overall, this is a nice piece of work and a well conducted study.

What I wonder, though, is what the groundbreaking new findings of this work are? Most of what is described in this paper has been done in prior studies and been published in one way or another. The findings that are presented in this manuscript mostly confirm the prior literature. Of course, there is some value in this, however, the paper could be strengthened if the authors were to explicitly emphasize what the new findings from this work are and how they advance the science in this field.

-Even though there are some studies on suitability of sorbent tubes for BVOCs, usually only a few most common terpenes (e.g. alpha and beta-pinene) are studied and very little is known on the suitability of the sorbent tube sampling for oxygenated BVOCs.  Performance of the tube sampling method may vary a lot even for different monoterpenes. We tried to clarify this in the introduction.  Ozone traps and tube materials have been studied, but not for this range of BVOCs, neither particle filters were thoroughly addressed for BVOCs in the literature to the best of our knowledge. During recent years it has become clear that more BVOC measurements (in addition to isoprene) are needed for air quality monitoring in rural areas in a climate change context, and in urban areas, where terpenes can be markers of some anthropogenic sources. Therefore cost-effective, simple and reliable sampling methods are needed. In addition to sorbent tubes, many of these results can be used also for choosing the most suitable set-up, cold trap and optimizing sampling times and flow for online sampling of BVOCs (breakthrough, selection of ozone scrubbers, tube materials, particle filters…). Carbon-based sorbents (e.g. Carbopack B) are still sometimes used for measurements of terpenes and quite often in multibed tubes. This study clearly shows that Carbopack B is not suitable for these compounds and even when using multibed tubes one should take care that these BVOCs are not breaking through the Tenax TA into the carbon sorbent.

There are a few places where the language is a bit inaccurate. I encourage the authors to carefully and critically review the entire manuscript text.  A few suggested corrections are presented below.

-The new version of the manuscript was sent to language check

Specific comments:

Line 1:  Measurements of atmospheric C10-C15 biogenic …..

-title changed as suggested by the reviewer

13:  …typically present at sub parts per billion mole fractions in

-this was added

15:  … simple sampling apparatus is needed for sample collection.

-this was added

24/25:  It would be better to report breakthrough volumes normalized to the amount of adsorbent.

-Since breakthrough volumes may also depend on the shape of the sampler and since all the other tests are based also on this certain type of sampler, we did not normalize these to the amount of adsorbent. It is easier for the users of the tubes to have these as volumes. We added a comment that these volumes are for the used type of tubes. Also in EN 14662-1 standard they use volume/tube.

28:  Be more specific than ‘greater impact’.

-We added relative difference values

34:  The global atmospheric burden of biogenic volatile …..is ~ 10 times

-this was corrected as suggested

56:  In my understanding of terminology, I would not label sorbent sampling-chromatography methods as ‘in-situ’.

-That is true, sorbent tube sampling with chromatographs is not in-situ, but these chromatographs are often also used in-situ. Then the chromatograph is at the sampling site and the sample is taken directly into the cold trap of the thermal desorption unit of the chromatograph. Those are generally called as in-situ/on line measurements. We tried to clarify this in the manuscript in section 2.2 (lines 162-163 in the new version of the manuscript).

85: There are quite a number of options for selective removal of ozone from the sample air, see for instance [Helmig, 1997]. Why and how did the authors select and decide to investigate the methods described in their study?

-these are commonly used at European air quality monitoring stations and are suitable for sorbent tube sampling.

107:  What do you mean by “Each compound was weighed in 500 mL of methanol”?

-Clarification on this was added to the manuscript into section 2.1 (lines 114-115 in the new version of the manuscript):  ‘For producing standard solutions ~30 mg of each BVOC was added to the 500 mL of methanol, which was further diluted into six different amount-of-substance fractions (a.k.a. amount fractions).’

110:  How high was the air flow rate?

-Flow rate varied between 1 to 2 L/min depending on the test. This was added to the section 2.1.

114:  …online online ….

-corrected

125:  This seems like a very short tube conditioning time?

-Freshly packed sorbent tubes were conditioned for longer periods, typically hours. This was re-conditioning time after every use. We added this remark into the manuscript.

153:  Please give the standard preparation date and please specify the regulator that was used.

-these were added to the manuscript section 2.2

154:  Please give flow rate and sampling length time.

-They were added to the manuscript section 2.2

159: How was the noise level determined if there were no peaks detected in blank runs?

-There is always background noise in the chromatograms, but no peaks deviating from the noise were detected in the blanks. Signal-to-noise ratio was calculated from the sample and calibration runs by integrating the noise next to the actual peak.

167: What makes you believe that the storage behavior of diluted methane solutions injected onto the cartridges is the same as for collected gas samples (without a large excess of methanol)?

-Similar as to the calibration tubes, the methanol was flushed away before the storage and analyses. Excess methanol would impact the analyses even without storage. We added a comment on this to the manuscript section 2.3 (lines 196-197 in the new version of the ,amuscript).

184:  Please provide more detail on the permeation device.

-permeation device was described in more details in the supplement S1 and we added a reference to the supplement into the manuscript on line 213.

250:  ….sodium thiosulfate ( …..

-corrected

251:   ….manganese dioxide …

-corrected

252:  These flow rates vary by more than a factor of ten for the different scrubbers that were tested.  How do you know that the comparison results are due to the materials tested and not from the different flow rates?

-It is true that these results depend on the flow rates. In addition to possible losses of VOCs, the flow rates depend on the ozone removal capacity of the scrubbers. While for the recoveries of VOCs high flows would be beneficial, with too high flow ozone is not removed efficiently anymore or ozone removal capacity of the scrubber remains efficient during very short time period (ozone removal capacities have been studied in earlier studies e.g. Hellén et al. 2012). The flows used for different scrubbers have been optimized based on this and are representing the flows used in the field. We have commented this in the results section 3.5 (lines 560-562 in the new version of the manuscript).

Table 2:  Please explain the impregnation procedure.

-added to the manuscript

Table 2:  Please explain the coating procedure with KI.

-The coating procedure is as follows:

“The potassium iodide ozone trap was home-made prepared by using a one meter copper tube (1/4 inch diameter) coated internally with a semi-saturated solution of potassium iodide (KI) consisting of 12.75 g of KI dissolved in 20 mL of demineralized water, for 3 hours. Then, the tube is emptied from the remaining solution, before being dried under a constant flow of zero air (0.75 L min^-1) for at least 3 hours.” Short version of this was added to the Table 2

275-282:  It would fit better to include the description of this instrument in the methods section.  Please also provide the details of the focusing traps.

-This part was moved to section 2.2.

Here are additional details regarding the three modules of the trap summarized in the table below:

TableXX: Detailed description of the traps present on the preconcentrator.

To clarify this we modified the description of the instrument now in section 2.2 lines 181-186 in the version of the manuscript as follows: ‘Samples entered the instrument through a preconcentrator provided with three traps. In a first step of the preconcentration, samples passed through a cold multi-sorbent trap (Tenax TA and glass beads) at -135 °C to eliminate the water in the sample. Then the trap was heated to 120 °C and the sample passed along a second additional cold trap of Tenax TA at -55 °C where it was cooled and then heated to 180 °C to remove the CO₂ present in the samples. Finally, the sample reached a third trap where a flow of liquid nitrogen at -200 °C passed, then the sample was heated to 150 °C.’

Table 3:  Please provide the part numbers of the filter materials that were tested.

The part number of each type of PM filter was added to Table 3 into the manuscript.

Table 3. Type of particle filters used for the tests.

307:  Provide all detail on the NPL standard (including preparation date).

-The date of issue  and comment on the standard composition was added to the manuscript section 2.9 on lines 322-324 (new version of the manuscript): ‘The calibration of the TD-GC-FID/MS was performed with a NPL (UK, date of issue 20^th March 2020) calibration standard containing NMHCs from C₂ to C₉ (including toluene) and 3 monoterpenes (α-pinene, β-pinene, limonene) at around 4 nmol mol^-1.’

346:  I question that breakthrough test results from samples that have such a high excess of methanol are applicable to gas standard samples.

-To ensure that the methanol was not affecting the results, we measured the breakthrough volume also with the permeation unit. As the gained results do not differ significantly, we assume the methanol plays a minor role (if any).

385:  Please discuss the effect of humidity in more depth. There definitely seem to be some differences.  And 70% RH isn’t really all that high.  For enclosure experiments with BVOC sampling one will often have sustained higher RH levels.

-There were differences for the commercial Tenax TA tubes, but not for self-packed.  However, tests with commercial tubes at 30 % humidity were done for the totally new tubes and this could explain the higher retention. We added to the manuscript section 3.2 (lines 369-385 in the new version of the manuscript): ‘. For the commercial tubes breakthrough volumes for RH 70% were lower than for RH 30 %. The impact of RH was not detected for the self-packed tubes. The retention of the compounds on the sorbent is based on the competition of the analyte with other molecules in air. Even though Tenax TA is a hydrophobic sorbent at high humidity some water may still be adsorbed, or water may condense onto the tube impacting the sorption of other compounds. In earlier studies impact of RH on Tenax TA retention was not clear since often Tenax TA was studied in multibed sorbents. In their breath sample studies Wilkinson et al. (2020) found that high relative humidity during sampling generally reduces the ability of Tenax TA/Carbograph sorbent tubes to capture volatile compounds. However, since the impact was more pronounced for the less volatile compounds with higher breakthrough volumes, they speculated that impact is other than breakthrough. Ho et al. (2017) found significant impacts of RH on retention of C₂-C₅ aliphatic hydrocarbons in multisorbent (Tenax TA- Carbograph 1 TD- Carboxen 1003) tubes. The breakthrough volumes for C₂ aliphatic compounds were reduced by 13–22-fold under 90% RH and the main impact was expected to be through the less hydrophobic sorbent Carboxen 1003.  For >C₅ compounds no breakthrough was detected. In the studies of Maceira et al. (2017) humidity problems were demonstrated with carbon-based tubes, while Tenax-based tubes did not display any influence. Our tests with commercial tubes at 30 % humidity were done for the totally new tubes and this could also explain the higher retention. In the tests under 70 % humidity conducted after the tests of 30 % humidity, any difference between commercial and self-packed tubes was not that clear. However, tests with a ReGaS2 generator with commercial tubes at RH 30% conducted later show slightly higher breakthrough volumes, but not as high as for the first tests (Fig. 3), indicating some influence of RH. Increasing the temperature of the sorbent tube to 10 °C above the temperature of the air sample could reduce the interference of the humidity (Karbiwnyk et al. 2002).’

396:  … b) Recovery …..

-corrected

396:  What was used as reference (100%) for these tests?

-100% was the expected amount in the NPL standard gas. To clarify this figure caption was changed to: ‘Figure 4. a) Relative difference between the measured and expected amount in NPL gas standard with different online mode sampling times. b) Relative difference between the measured and expected amount of NPL gas standard flushed into different tubes (C=commercial, S=self-packed) or analyzed with online mode of the instrument (online). Error bars show the standard deviation between the three replicate samples. For both tests the instrument was calibrated with methanol solutions of BVOCs.’

406-409:  This points towards difficulties and inconsistencies you had with the SQT analyses?  Please either detail and explain or maybe just drop SQT completely from the manuscript?

-Actual analyses or calibrations were not a problem. The only problem was production of stable SQT containing air. Therefore, all other results and tests should be reliable also for them. Comment on methanol injection system on lines 406-409 was removed, since it is not clear that it implies to the system producing terpene rich air and it is not the system we used for calibrating the instrument.

408:  What do you actually really mean by ‘online’ sampling?  Sorry, that is still not really clear to me.

-In online mode, sample is taken directly to the cold trap of thermal desorption unit without first concentrating it to the sorbent tube. We clarified this in the manuscript in section 2.2 in lines 153-154 (new version of the manuscript).

464: Mention in figure caption that these are the results for the sodium thiosulfate ozone scrubber.

-this was added

473:  Pollmann

-corrected

499: This recommendation seems subjective. What really matters is the sampling volume through the filter and the particle loading. 

-This comment was removed from the manuscript.

Figure 8:  These values are all positive.  Does that mean that all of the tested VOCs are artificially generated on the filters? Wouldn’t one expect losses, i.e. negative numbers?

-Most of these are still within uncertainties. The main observed differences come from the used stainless steel filter where the compounds from earlier measurements may have been released during the tests. For the other filters, only limonene shows high differences (around 5%) which may be due to its high reactivity and may be due to the sampling and analyses to a lesser extent.

538: This really is an essential question that I would like to see addressed in much more depth:  How comparable are the liquid standard versus gas standard calibration results?

-As discussed in section 3.3 and shown in figure 4a, deviation between liquid versus NPL gas standard calibration was less than 12 % for alpha-pinene, 3-carene and 1,8-cineol. However, compared to traceable certified NPL gas standard calibration, liquid calibration is prone to more errors and should be always compared to traceable gas standards if possible. Stable gas standards are available only for a few BVOCs and response of MS may vary between the BVOCs. Therefore, methanol standard was chosen for this study, where performance of the tubes were studied for many BVOCs. Lack of gas standards is also reason why comparison of liquid vs. gas standard was conducted only for a few compounds.

We tried to clarify this by modifying the sentences in section 3.6 on lines 570-574 (new version of the manuscript): ‘Some differences between IMT and FMI were higher (23 % for α-pinene in Tenax TA) than the calculated uncertainty (7 % for both laboratories, Table 1, Supplement S2). A possible reason is the different calibration method and their associated uncertainties; FMI calibrated their TD-GC-MS with methanol solutions while IMT used a traceable NPL gas standard to calibrate their TD-GC-FID. Comparisons of liquid methanol standard and NPL gas standard in section 3.3 indicated that this can cause up to 12 % difference for α-pinene results.’

Figure 9: Sampling volumes would be more meaningful than sampling times.

-we corrected this to the figure

Helmig, D. (1997), Ozone removal techniques in the sampling of atmospheric volatile organic trace gases, Atmospheric Environment, 31, 3635-3651.

-the reference was added