Thanks to Dr. Niedermann for a careful reading of this paper. All three of these corrections can easily be made in a revised version.

As regards the grain size, both standards were analyzed as they were supplied, without further grain size separation. However, neither material is extremely well sorted (CRONUS-A was prepared from a 0.25-0.8 mm size fraction, and CoQtz-N from 0.25-0.71, and the grain size distribution appears to have become wider after HF etching), so the jars were thoroughly mixed to homogenize the grain sizes before each aliquot was taken. It is not known whether there is a grain-size dependence of the Ne-21 concentration in either material (although this is unlikely for a number of reasons), but thorough mixing is probably a good idea regardless because of the small size of the aliquots used for Ne-21 measurements. 

Thanks to PH Blard for a careful review of this paper. The review includes a number of comments, which, as suggested by the reviewer, can be easily resolved, as follows: 

1 – Statistics: could you try processing the data considering a statistical model that assumes two different sources of variability (analytical uncertainty + inter-aliquots heterogeneities), following for example Vermeesch et al (QG, 2015)?

The reasoning here is that it is not possible to disprove the hypothesis that all samples belong to one population characterized only by measurement uncertainty, or, equivalently, that the simplest possible uncertainty model is adequate. For example, the chi-squared p-value is 0.66, so the hypothesis is not rejected at high confidence. Thus, proceeding to a more complex uncertainty model is not justified. 

If one did as suggested, the overdispersion would be indistinguishable from zero. 

2 – Nucleogenic ²¹Ne: author discusses the possible presence of nucleogenic ²¹Ne, mentioning a number of 2 x 10⁶ at/g. However, since most of the extraction steps clearly fall on the spallation lines in the 3-neon isotopes space, I don’t understand why this hypothesis should be considered as likely. Additional calculations considering the U-Th concentrations of the quartz, as well as the thermal history of this rocks, might also help to place independent constraints on this potential nucleogenic ²¹Ne production.

In the manuscript, this issue was condensed into the remark '...it is not possible to discern from Ne isotope ratios alone whether measured excess ²¹Ne is all cosmogenic or may include a nucleogenic component' (lines 154-155), without further explanation.

To explain this further, the reason that this is not possible is because nucleogenic neon can be the result of (α,n) reactions on both ¹⁸O (producing ²¹Ne) and ¹⁹F (producing ²²Ne), and, therefore, nucleogenic neon can have any ²²Ne/²¹Ne ratio imposed by the F/O ratio at the location where the neon is produced. The O concentration in quartz is stoichiometric, but F concentrations in quartz are generally unknown. Neon isotope data from quartz in the literature show displacements of Ne isotope ratios both above the cosmogenic-atmospheric mixing line (indicating nucleogenic ²²Ne/²¹Ne greater than cosmogenic ²²Ne/²¹Ne) and below (nucleogenic ²²Ne/²¹Ne less than cosmogenic ²²Ne/²¹Ne). If  nucleogenic and cosmogenic ²²Ne/²¹Ne production ratios are similar, then nucleogenic neon would simply displace isotope ratios along the atmospheric-cosmogenic mixing line, and, no matter how precise the measurements, it would be impossible to tell whether ²¹Ne and ²²Ne excess was cosmogenic or nucleogenic. Obtaining a constraint on the nucleogenic ²²Ne/²¹Ne ratio would require either (i) measurement of the F/O ratio in the quartz (which might not be conclusive, because nucleogenic neon trapped in fluid inclusions might not have been produced in situ), or (ii) measurement of Ne isotopes in a shielded sample of the same lithology. As neither of these are available for CoQtz-N, the nucleogenic ²²Ne/²¹Ne ratio in this sample cannot be determined, which means that cosmogenic and nucleogenic ²¹Ne and ²²Ne cannot be distinguished on the basis of Ne isotope ratios. 

Although this is interesting, it is not critically relevant to the subject of the paper, as noted in lines 163-164 of the manuscript. Thus, no change to the manuscript is proposed here. 

3 – It would be useful to add in the Introduction more information about the quartz material itself: location of the rock, sieved granulometry, detailed about the quartz separation and cleaning (type of HF etching: external rim removal? rinsing?)

This information is all provided in the source papers describing the preparation of the quartz standards (Jull 2015 and Binnie 2019). In the interests of keeping the present paper short, it was not repeated. However, as noted in the response to the other review, the information that the samples were measured exactly as supplied can be added. 

4 – I didn’t find the blanks values in the data spreadsheet available as a supplement. Although I understand that the author treats blanks as part of the sample because of their atmospheric isotopic signatures, I however think this is a useful information, that needs to be provided for each analytical system system.

These data are not in the supplement because the empty Ta packets were screened in different analytical sessions, prior to the measurements described here. Basically, Ta tubes were only screened when a new batch of tubing was received, and not in every analytical session. In all cases the amount of neon released from empty tubes was in the range of 0-50,000 atoms ²¹Ne, with, as noted, isotope composition indistinguishable from atmosphere. This information can be added to the text. 

Other minor comments and suggestions:

Line 5: A "low" concentration is a subjective statement that is not informative. Please provide ²¹Ne concentrations range.

This is true, but the intended point of this sentence is simply that the concentration in CoQtz-N is lower than in other standards. The actual concentrations are provided in lines 9-10. 

Line 16: Quote the relevant articles there (Jull et al., 2015; Vermeesch et al., 2015).

These articles are cited in lines 18-23. 

Line 27: Add a point (.) after the ending parenthesis.

Corrected. 

Line 31: Maybe quote here (Blard, CG, 2021), a paper in which I list several additional sources of inaccuracies (section 4.5), notably internal leaks, uncomplete release of the analyzed gas from the cold trap, or non-linear behavior of mass spectrometer sensitivity.

To clarify this comment, although it is correct that Blard 2021 as well as Blard 2014 and Vermeesch 2015 propose various reasons for the discrepancies, these references do not include the observation that the offset is constant among samples with widely varying concentrations, which is consistent with only some of these possibilities. For example, incomplete trapping/release or nonlinear sensitivity would be unlikely to be constant across different samples with different properties and ²¹Ne concentrations, but instead would be expected to be correlated with, for example, Ar or He abundance (for the trapping efficiency hypothesis) or total Ne pressure (for the nonlinearity hypothesis). The observation of a consistent offset across variable ²¹Ne concentrations does not appear in the Blard 2021, Board. 2014, or Vermeesch 2015 references, but instead in the Balco 2019 and Balter-Kennedy 2023 references. 

Regardless, as all of these papers are extensively cited throughout the manuscript already, no changes to the text are proposed here. 

Line 38 to 40: Same comment that the one I did for the abstract: provide average concentrations for these standards.

These are given in line 47. 

Line 45: Here, more information about the CoQtz-N standard would be useful: 1) Where to get this std, 2) available amount?, 3) granulometry?, 4) was it etched with HF to remove the external rim of the grains (an important issue regarding the nucleogenic production)?

As noted above, a revised manuscript can clarify that the samples were measured exactly as supplied, without further preparation. All these points are covered in detail in the Jull and Binnie papers describing the quartz standards, and have not been reproduced here. 

Line 66 to 69: Please provide here 1) the ranges of the observed ionization ratios (Ar⁺⁺/Ar⁺ and CO₂₊₊/CO₂⁺) and 2) The amplitude of the CO₂⁺⁺ and Ar⁺⁺ corrections on the measured ²²Ne and ²⁰Ne in this dataset.

This procedure, including the characteristics of the BGC MAP-II, is thoroughly discussed in the Balco and Shuster 2009 reference, and to keep the present paper as concise as possible, this discussion was not reproduced here (the CO₂⁺⁺/CO₂⁺ ratio for the LLNL system was reported in line 98 because it has not been mentioned in any previous reference). With respect to the amplitude of the corrections, corrections on mass 20 for ⁴⁰Ar⁺⁺  and on mass 22 for CO₂⁺⁺, for the BGC analyses described in this paper, were equivalent to 0.15-0.22 Gatoms ²⁰Ne and 2.5-2.7 Matoms ²²Ne, respectively. The importance of these corrections can be obtained by reference to the supplementary tables, but of course it is difficult to describe this concisely in a meaningful way because the importance of the correction is highly variable with the amount of gas in the sample. For example, for a blank analysis or a high-temperature step with minimal gas release, the correction is obviously nearly 100% of the signal on both mass 20 and mass 22, but it would not be very helpful to state that "the correction was up to 100% of the measured signal". Regardless, the equivalent abundances above can be added to the paper. 

Line 73: "small amounts" is a subjective statement. Please provide the relative contribution of these hot blanks to the samples, as you did just above for the cold blanks. This is all the more important as hot blanks represent the real extraction conditions of the samples. I understand that you decided to treat these hot blanks as an atmospheric contribution released along with the atmospheric neon contained in the samples. However, it is important and useful to provide the value of these blanks (in atoms and/or in blanks/samples contribution) for readers.

As noted above, the range of Ne amounts found in the hot blanks can be added to the text. 

Line 82 to 84: I don't understand this reasoning. Why a 3 steps heating would prevent cross-heating of the nearby samples, notably at 1100 °C? Should we understand that you first performed the 450°C on all samples before going to higher T steps? In that case, please indicate more clearly which procedure has been applied.

Correct: all samples are heated to 450 first, then all samples to 850, etc. This can be clarified. 

Line 97 to 98: What is the CO₂⁺⁺ contribution on the measured ²²Ne signal?

For the LLNL system, the correction is equivalent to 2-12 Matoms ²²Ne. The higher variability at LLNL than at BGC is related to gettering efficiency during sample processing. 

Line 107: What is the amplitude of the gas standard adjustment after normalizing it against a 320 Mat/g concentration for CRONUS-A?

There is not really a meaningful answer to this, because comparison to an approximate value that was not expected to be accurate except by accident would not be very informative. As it happens, the approximate estimate was about 10% different from the CRONUS-A estimate, but it's not clear what this means because it is unknown whether the uncertainty in the approximate guess was bigger or smaller than 10%. In any case, this issue is not relevant to the paper -- it is not necessary that any approximate estimate ever existed, and for purposes of this work, the calibration is completely defined by the CRONUS-A measurements. 

Line 124 to 125: I am not sure that this statement is correct. What is the contribution of the atmospheric ²¹Ne correction to the total uncertainty, notably for the new standard quartz analyzed here? For samples bearing less than 20 Mat/g of cosmogenic ²¹Ne, this correction is generally a larger source of uncertainty than the reproducibility of the ²¹Ne measurement itself: >5% here, looking at Figure 1 and the results table.

To clarify this comment, another way of describing this observation is that the equation used to compute excess ²¹Ne (in line 118) has the property that the uncertainty in excess ²¹Ne increases nonlinearly with the fraction of ²¹Ne that is atmospheric (²¹Ne2_atm/²¹Ne_tot). This is true regardless of the source or magnitude of the uncertainties. In this case, the primary source of uncertainty is the reproducibility of the ²¹Ne signal in the standards, which is propagated into an uncertainty in the ²¹Ne sensitivity and thence into an uncertainty on N_21,tot in the equation in line 118. For the specific example of CRONUS-A (for which ²¹Ne2_atm/²¹Ne_tot is about 0.1-0.15), the overall uncertainty in excess ²¹Ne is, as noted, about 1.8% (BGC) or 1% (LLNL). For CoQtz-N. ²¹Ne2_atm/²¹Ne_tot is about 0.75, so even with the same relative input uncertainty on N_21,tot, the output uncertainty on excess ²¹Ne is much bigger (5-6%). Thus, the reviewer's statement that the atmosphere subtraction is commonly responsible for most of the uncertainty in excess ²¹Ne is correct, but it is not in conflict with the text. The point here is that the actual size of the uncertainty depends strongly on ²¹Ne2_atm/²¹Ne_tot just because of the form of the equation, but, for any specific value of ²¹Ne2_atm/²¹Ne_tot, the largest contributor to the total uncertainty is always the reproducibility of the standards. Given that the text refers specifically to CRONUS-A, it appears to be correct as written, and no change is proposed. 

Line 144: Here, do you refer to the ²²Ne/²⁰Ne ratio, that is indeed indistinguishable from the atmosphere? The ²¹Ne/²⁰Ne ratio seems on the contrary to be statistically larger than the air endmember, as discussed further below.

Well, both. The statement 'high temperature steps have composition indistinguishable from atmosphere when each analysis is considered individually' is equivalent to saying that most of the 68% uncertainty ellipses overlap the atmosphere endpoint. Although the 95% confidence ellipses are not shown, all of the 95% ellipses for the high-temperature steps would overlap the atmosphere endpoint. The second part of this sentence then states that all the high-temperature steps have 21/20 ratios slightly above atmosphere, which is equivalent to saying that all the black dots for the high-temperature step in the figure are to the right of the atmosphere endpoint. As this is the same thing that the reviewer is saying,  it appears that this review comment and the paper are in agreement. 

Line 154-155: I don't understand this reasoning: on the contrary, if extraction steps fall on the 3 isotopes cosmogenic line, it is highly probable that the excess ²¹Ne is cosmogenic, and not nucleogenic (at least for your low T extraction steps).

As discussed above, without any constraint on the 22/21 ratio of a hypothetical nucleogenic component, it's strictly not possible to say this. A mixture of cosmogenic, atmospheric and nucleogenic neon could lie either on or off the cosmogenic-atmospheric mixing line.