We thank Trystan Herriott for his thoughtful comments and suggestions on this manuscript. We agree with Trystan and will provide a more detailed discussion on the impact of matrix effects on uncertainties for LA-ICP-MS analyses. To aid in this discussion, we plan to collect new zircon profilometry data to assess the differences in pit depths between treated and untreated aliquots of FC1, R33, and MIGU-02. This will help us understand how laser coupling and ablation rates may be impacted between treated and untreated aliquots. We will also add an additional section to the introduction to address the significance of Pb-loss in different high-T/low-T conditions.

We also agree with Trystan (and other reviewers) for the need of more quantitative comparison of treated and untreated detrital zircon aliquots. We have added several assessments of similarity between treated and untreated aliquots using DZStats (Saylor and Sundell, 2016) and have calculated the percentage of total grains comprising peak age populations for treated and untreated aliquots. These new comparative assessments have helped quantify the changes in peak shape (height, width) between treated and untreated aliquots and have been incorporated into the discussion of the revised manuscript. Please see other referee comment replies for additional details.

We will add in some final statements in the conclusions section addressing future work related to this method. This will focus on the need for future testing and more analyses to better understand how CA impacts the youngest detrital zircon peak age populations and potentials for improving maximum depositional age (MDA) calculations.

Line by line comments:

Line 36: We will mention the importance of MDA calculations from detrital zircon analyses in the introduction. The importance of MDA calculations and how CA-LA-ICP-MS can potentially improve these calculations will also be outlined in the conclusions.

Table 1/Figure 1: We will double check to make sure all appropriate CA-ID-TIMS vs. ID-TIMS superscripts were used in Table 1 and make sure they are properly labeled in Figure 1.

We thank Marcel Guillong for their careful review and comments on this manuscript. In agreement with their comment (and others), we have changed the title of the manuscript to: Minimizing the effects of Pb-loss in detrital and igneous U-Pb zircon geochronology by CA-LA-ICP-MS. We believe this change better reflects the key findings of this study. Improved accuracy has been observed in analyses of chemically abraded igneous zircon (this study, von Quadt et al., 2014). Demonstrating this same improvement for detrital zircon is much more difficult given that we cannot re-date every analyzed zircon by a second method, like ID-TIMS. Instead, we have to infer that the detrital zircon will respond to chemical abrasion in the same manner as the igneous zircon and minimize the effects of Pb-loss. This should improve accuracy in detrital zircon measurements and lead to predictable changes in DZ spectra. Namely that age populations will sharpen and move toward slightly older dates as Pb-loss is mitigated. Zircon populations with high degrees of radiation damage may also be preferentially dissolved in the process and lead to changes in the relative importance of each age population. We did note some of these changes in the manuscript. To address the concern of the reviewer and to better quantify changes between treated and untreated detrital zircon aliquots, we have added several assessments of similarity between treated and untreated aliquots using DZStats (Saylor and Sundell, 2016). Additionally, the proportion of grains within each age peak (% out of total for that sample) has been calculated and labeled in the revised manuscript. This comparative assessment of the DZ samples has helped quantify the changes in the heights and widths of peak age populations and has helped better frame our discussion of how CA affects detrital zircon age spectra.

In reference to the comment: “accuracy and precision of reference materials analyzed using CA and non CA seems not improved…”:  We have also gone back through in the revised manuscript to clarify language in Section 3.1. Here, the reviewer’s main concern is that accuracy and precision of reference materials was not improved between treated and untreated aliquots. However, we note that we did not necessarily anticipate this would be the case, because reference materials used for U-Pb geochronology are dominantly concordant (i.e., Pb loss is rare or absent) and ‘well-behaved’, and these are reasons why they are chosen as reference materials in the first place. Instead, the main objective of analyzing all these reference materials after performing chemical abrasion was to demonstrate that the accuracy and precision of our U-Pb data would not be negatively affected, and hence that the dates from our unknows are reliable over a wide age range, We did, however, note some slight improvement in the U-Pb systematics of some reference materials, given that fewer analyses were discarded due to discordance. However, these materials were Proterozoic in age and are primarily used for their homogenous ²⁰⁷Pb/²⁰⁶Pb which is less sensitive to recent Pb-loss. Overall, the behavior between treated and untreated reference materials is similar, and thus the objective of these analyses (i.e., demonstrate our method does not negatively affect accuracy) was met. Again, this result is expected because reference materials are selected due to homogenous isotopic compositions and excellent behavior during analysis. For the purposes of this manuscript, these results demonstrate that chemical abrasion does not systematically bias our U-Pb results.

In reference to the comment that the “comparison for MIGU-02 is not entirely fair as for the CA about 150 zircons were used vs non CA only 35 grains being used…”:  We acknowledge that there is a balance to be struck when deciding whether to utilize chemical abrasion prior to LA-ICP-MS analyses. For samples with significant radiation damage, there is always the possibility that the entire sample will dissolve. Running a high-n on highly damaged zircon might ultimately yield enough concordant analyses to make a confident age determination. However, the concordant analyses for our metamict igneous sample MIGU-02 were inaccurate by up to -11% for the ²⁰⁷Pb/²⁰⁶Pb dates and -21% for the ²⁰⁶Pb/²³⁸U dates.  The chemically abraded aliquot didn’t have these issues. For many felsic igneous samples zircon yields are not an issue, so performing the CA treatment on many crystals–even if the majority dissolve–as a means to optimize analytical time on the LA-ICPMS and enhance accuracy is, in our experience, a worthwhile approach. We think that chemical abrasion’s efficacy at reducing Pb-loss, its relative ease and low cost in the laboratory, and the possibility of optimizing ‘beam time’ by only focusing on those grains that will yield concordant results, make it a worthwhile step in U-Pb zircon analyses by LA-ICP-MS. However, we will better acknowledge the potential drawbacks for samples with high degrees of radiation damage in the revised manuscript.

To address the comment on how CA may impact laser coupling with the zircon (line 73-74), new optical profilometry data was collected to quantify the depth and shape of laser ablation pits in treated and untreated grains, and will be incorporated into the revised version of this manuscript. This will help us better understand how CA influences laser coupling and ablation rates between treated and untreated aliquots. Although zircons are mounted and polished, Crowley et al. (2014) and McKanna et al. (2023) shows that chemical etching and 3D porous texture can occur throughout the zircon crystal interior.

A few notes on addressing inaccurate wording: We agree with the reviewer and in accordance to their comment to avoid use of the term standard when referring to reference materials, we went back and removed any language using the term standard. Also in accordance with the statement that rank order plots are not ranked, we revised all rank order plots to be ranked by age (youngest to oldest) in Figure 2 and all supplementary material of reference materials. TE concentrations and how these respond to the CA treatment are beyond the scope of this study, so we implemented no changes here, as we do not have the necessary data to address it.

Other Figure/Line revisions:

Figure 3A. Scaling changed to better display ellipses. If we were to plot the discarded analyses, then no detail would be observed. This would make it hard to see any detail in the concordant analyses.

Figure 4, 8, and 9: How was the U concentration quantified.

Uranium concentrations, as reported in our original submission, were semiquantitative, calculated using simple standard-sample bracketing relative to the average U concentrations for our primary reference material. Quantification of trace element concentrations using LA-ICPMS are rigorously done using internal normalization relative to a stoichiometric element (e.g., Zr or Si), but we were not able to do this with our method. This was because measuring isotopes in the Zr or Si mass range would have required a magnet jump with the Element2, which would significantly slow down our analyses. The reviewers are correct that this should be better explained, so in our revised manuscript we have included additional clarifications about our analytical method and the reason why quantification was not done using an internal standard. Furthermore, because the U concentrations we reported in our original manuscript were not strictly quantitative, and the CA treatment of reference materials has the potential to skew the semiquantitative calculations originally performed, we have modified our manuscript to avoid the use of U concentrations in the text and figures. We now base our observations on the 238U cps for each analysis, reported after performing a simple inter-session normalization for instrumental sensitivity. We explain this procedure in greater detail in the revised manuscript, but in brief: we note that the 238U cps of our SL crystal were very homogeneous between and within runs of treated and untreated aliquots, so we used these as reference to normalize the cps of 238U for all sessions. By removing minor variations in sensitivity using this simple approach, we now focus our discussion on the effects of chemical abrasion as a function of 238U cps rather than U concentration. While this approach does not affect our general conclusions, it does resolve two key issues: i) removes the need to build our discussion around U concentrations, as these were not determined quantitatively; ii) removes possible inaccuracies introduced by the effects that chemical abrasion of reference materials can have on U (semiquantitative) concentrations calculated by simple standard-sample bracketing.

Figure 5 and 6: We have added in the proportion of analyzed grains (out of total) that comprise the age populations to these two figures and quantified similarity between the treated and untreated aliquots using DZStats (Saylor and Sundell, 2016).

Line 350: This was a type-o. Thank you for catching this mistake! It has been corrected in the revised version to ²⁰⁷Pb/²⁰⁶Pb age.

We thank David Chew for their careful review and comments on this manuscript. We address each bullet point in the original review below:

1. In agreement with this comment (and others), we have changed the title of the manuscript to: Minimizing the effects of Pb-loss in detrital and igneous U-Pb zircon geochronology by CA-LA-ICP-MS. We believe this change better reflects the key findings of this study (see Marcel Guillong’s comment/reply).
2. In reference to the comment on seeing data on pit depths for treated versus untreated aliquots: We are currently collecting new zircon profilometry data of treated and untreated aliquots of MIGU-02, FC1, and R33 to see how pit depths vary. This new data will be incorporated into the revised version of this manuscript to better assess how CA impacts laser coupling and ablation rates between treated and untreated aliquots. We think it is an excellent suggestion to compare the pit depths of treated and untreated MIGU-02 zircons with their age, 238U cps, and concordance.
3. We have completed a more quantitative assessment of detrital zircon age populations between treated and untreated aliquots of Rora Med and NM8A. We used DZStats (Saylor and Sundell, 2016) to calculate similarity, cross-correlation, and likeness values. Additionally, we agree with the reviewer’s suggestion to calculate the percent of grains (out of the total) defining each peak. These will be labeled on revised Figures 5 and 6 and the new quantitative assessment will be incorporated into discussion of how CA impacts peak height and width.

In regard to the reviewer’s comment on CA selectively dissolving high U grains with reference to the 1890 peak age population of Rora Med: To address this comment, we wanted to better quantify the decrease in peak height from the treated to the untreated aliquot and compare U concentrations associated with this peak age population. The total percent of zircons making up the 1890 peak age population in the untreated aliquot of Rora Med is ~7.4% compared to a decrease to ~3% of total grains in the treated aliquot. This decrease correlates to the observed change in peak height on Fig. 5. To determine if CA preferentially dissolved zircons of high U associated with this peak age population in the treated aliquot, in our revised manuscript we now compare the 238U cps between treated and untreated aliquots. As discussed in the response to reviewer Guillong’s comments, in our revised version we modify the way in which U is discussed, avoiding calculated U concentrations and focusing on sensitivity-normalized 238U cps. Doing this better elucidates the effects that chemical abrasion had on Rora Med, as follows:

We updated Figure 9 to compare the 238U cps between treated and untreated aliquots of Rora Med (and NM8A). In the updated version, the treated aliquot of Rora Med has overall lower 238U cps values compared to the untreated aliquot of Rora Med. This difference is especially noticeable for the 1800-2000 Ma and the 2600-2800 Ma age populations. For example, in the untreated aliquot, there are 70 grains associated with the 1880-1900 Ma age population. About 33% of the grains between 1880-1900 Ma make up the highest 238U cps values (238U cps > 2,171,019) for the entire aliquot, but these grains only represent ~2.5% of the total aliquot (23/920 grains). In contrast, there are a total of 33 grains in the treated aliquot for this age population, and ~42% of zircons (14 total grains) in the 1880-1900 Ma age population make up the highest 238U cps values (238U cps > 1,355,272). This represents ~3% of all zircons in the treated aliquot (14/1035 grains). Overall, there are significantly less zircon grains of the 1880-1900 Ma age population in the treated aliquot, and the overall 238U cps values are lower. This difference is likely due to mitigation of Pb-loss in the treated aliquot, but overall, the range of U values associated with the 1890 peak age population in both treated and untreated aliquots suggests that CA does not selectively dissolve only high U grains.

1. In this study, the separates that were bulk chemically abraded were 100% zircon. However, author Donaghy has implemented this method on detrital zircon samples for other research and has used separates that were ~85-90% zircon. The acid dissolution does effectively remove the other mineral phases and leaves a sample that is 100% zircon.
2. We agree with the reviewer’s suggestion here to add in more discussion on the presence of xenocrysts and antecrysts and their effect on how they shift the calculated age from the reference age (scatter in Figure 1; % difference from reference age). As suggested by this reviewer and others, we will draft a new figure that shows the % bias for each run of treated and untreated aliquots of the 13 reference materials. This figure will highlight reduced scatter in CA aliquots. With respect to the 91500 crystal, the aliquot we used was obtained from the International Association of Geoanalysts (https://iageo.com/zircon-91500/) so we believe its origin is robust. That said, in past research, the 91500 reference material has shown substantial negative age offset (Gehrels et al., 2008; Schoene et al., 2014), but the origin of these offsets has remained enigmatic. So, while the offset in this study is not entirely surprising, we expanded our discussion to briefly highlight this on our revised manuscript.

Minor Comments:

Figure 1: We have modified our approach re. reporting U (see response to reviewer Guillong), and U concentrations are no longer used in the text or figures.

Section 2.2: We will add in the fluence on each sample, including the repetition rate and total amount of shots for the LA-ICP-MS methods.

Line 152: We will add in more language in the methods section outlining the term ‘round robin’ in context of analysis setup for this study.

Line 164: We made suggested change from “the dates of CA and non-CA…” to “the weighted mean dates…”.

Figure 2 and Line 221 – Reviewer caught a type-o in the n-value listed in Figure 2 compared to the correct value written in line 221. Figure 2 listed the wrong total number of analyses (n=35 versus n=20). Figure 2 was changed in the revised figure and no other changes were necessary as calculations were based off a n=20 value.

The grid lines in Figure 4 and Figure 10 will be modified to be a lighter weight and reflect the appearance of Figure 9.

We thank Matthew Horstwood for their careful review and comments on this manuscript.

1. ‘Please state the uncertainty level in all figures’: We will add the uncertainty levels to all figures/supplementary materials in the revised manuscript.
2. Uncertainties and MSWDs will be quoted to 2 significant figures with the ages/ratios/values quoted to the same 2 significant figures.
3. All metadata tables for the LAICPMS and IDTIMS results will be included in our revised version.
4. The reference values used for FC-1 are the CA-ID-TIMS results of Ibañez-Mejia and Tissot (2019), which were obtained using material from the same outcrop as that used for this LA-ICPMS study. We have clarified this in the revised manuscript.
5. ‘Imaging and targeting to avoid zonation/inclusions/xenocrysts..’: Although CL imaging was used for all samples in this study, we will add in a more detailed discussion on the usefulness of CL imaging during LA-ICP-MS zircon analysis to avoid inclusions/xenocrysts/zoning.
6. We appreciate the drafted figures and recommendation for tabulating the % bias of the treated and untreated aliquots of reference materials to compare. This is an excellent recommendation and we will incorporate this suggestion as an additional figure and provide a more detailed discussion in the revised text on how CA is reducing the scatter while not changing the overall % bias. Additionally, we will tighten up vague language surrounding statements such as ‘improved resolution and precision’ and ‘CA improving the resolution of concordancy’. We will provide a better discussion on how we are defining these terms, and with new comparative assessments of treated and untreated aliquots, will be able to better quantify these changes in the discussion.
7. In regards to the comment on quantifying the shape and resolution of peak age populations in detrital zircon samples: We have completed a more quantitative assessment of detrital zircon age populations between treated and untreated aliquots of Rora Med and NM8A. We used DZStats (Saylor and Sundell, 2016) to calculate similarity, cross-correlation, and likeness values. Additionally, we calculated the percent of grains (out of the total) defining each peak to better quantify how peak heights and widths vary between treated and untreated aliquots. These will be labeled on revised Figures 5 and 6 and the new quantitative assessment will be incorporated into discussion of how CA impacts peak height and width. Please see comment reply to Chew and Guillong.