The manuscript “Geochronological and Geochemical Effects of Zircon Chemical Abrasion: Insights from Single- Crystal Stepwise Dissolution Experiments” by McKanna and others is presenting step leaching U-Pb ID-TIMS data on three different zircon reference materials for two different temperatures and giving a recommendation for conditions to use by practitioners of chemical abrasion and an estimated minimum alpha dose at which Pb loss can occur. This study is a follow up from a previous study by the same authors on the same zircon samples that looked at structure of individual zircon grains of in response to chemical abrasion at different temperatures.  

The conclusions from this study are largely confirming what has already been published about the chemical abrasion conditions and its consequences (Huyskens et al 2016, Widmann et al. 2019). All three studies concluded that leaching at 180C is not high enough for most samples to effectively remove all Pb loss from a sample and a higher temperature is needed in most cases. This study, however, is the only one that has detailed structural observations of zircons before and after the leaching procedure and it can tie them together with the U-Pb observations. This also leads to the possibility of estimated radiation damage of the leached zones, which could not be done in the two previous studies. This is the main new angle of the study and these observations and discussions are unfortunately falling short in favour of documenting observations at length that have been made before. Thus, my recommendation is to focus a lot more on the new aspects of the study.

To cut back on the repetitions reorganising of the results section would help. For example, for each sample early leachates are enriched in LREE. Thus keeping the standard format of first heading “geologic setting and sample description” with a sub heading of the different samples is advantageous. It also makes it easier to compare and contrast the different zircons analysed in this study.   

The authors also looked for a tool to robustly identify zircons that have remaining Pb loss. Currently the identification of such analysis is very subjective and such a tool would make interpretations of a scattered dataset more robust. Unfortunately, neither of the indicators (U concentration,  Pb*/Pbc, or LREE enrichment) are effective tools for this task.

One of the recommendations of this paper is to look for the amount of radiation damage and tailor the leaching conditions this way. However, no information is given on how to determine the radiation damage prior to dissolution. In this study the radiation dose was calculated based on the analysis of U and Th in the leachate and residue of the zircons, which means after already performing the time-consuming analyses. In the prior study Raman was used to estimate the alpha dose for the zircons. Any method to determine the alpha dose would need to have a high spatial resolution for the entire volume of the zircon grain, since it was documented in the previous paper that these zones do exist also in the interior of some zircon grains.

Please provide all the calculated numbers that are used in the plots like alpha dose at Pb loss or LREE-I.

A description is needed how the amount of dissolved material for the calculation of alpha dose was determined.

Fig 12: could you distinguish between the different temperatures maybe using open vs filled symbols?

Fig 15: It looks like the samples leached at 210C overall had lower radiation damage compared to the ones used in the 180C experiment in this figure and this would need an explanation. My guess is that it has something to do with the fraction of material in each of the dissolution steps, but this is confusing at first. For me, the figures about the alpha dose are the most important in this study and the groupings in Fig 14 and 15 are so broad that they could be masking interesting details. A plot including alpha dose vs discordance for example on an individual analysis basis could be interesting.

Line 31:” However, Ultimately,…”

Line 49: “since the trajectory of Pb-loss follows Concordia” I think it should be the concordia

Lines 50-51: “the precision of 207Pb/235U dates is also lower than corresponding 238U/206Pb dates due to the shorter radioactive half-life of 235U and lower isotopic abundance (Corfu, 2013; Schoene,

2014).” The precision of the 207Pb/238U dates is lower not because of the shorter half-life. It is only lower due to the lower abundance, which in turn is due to the shorter half-life.

Line 59:” annealing zircon samples prior to leaching helps to minimize the unwanted isotopic

fractionation effects that plagued earlier leaching attempts” The main improvement is reducing elemental fractionation in the leaching steps.

Line 191: “included and altered grains…” Grains with inclusions

Line 210: “which had been redone as part of (Schoene et al., 2006).”   Brackets are in the wrong place

Line 522: frantzing should not be a verb

The paper reports the results of experiments on the chemical abrasion of zircon, using 3 samples of different age, and applying two different experimental protocols. The reported results include U-Pb isotopic ratios and ages, and abundances of a number of chemical element typical in zircon. The experiments evaluate the efficiency of the two different approaches in removing discordant zircon domains and isolating grains with closed isotopic systems.

The data will be interest to the geochronologists that use the U-Pb dating methods, especially the ID-TIMS community. It is not the first paper to report such experiments, but it adds some new perspectives, which will certainly be useful for a further advancement of the technique.

The paper is reasonably well prepared. There are some technical glitches, with figures inserted in the text and locally covering up portions of the text. I have put some suggestions and a number of comments directly in the annotated file.

The tables need some work to make them more useful and accessible for the readers. (1) It would be practical to assemble them all as separate sheets in just one file. (2) It would be practical to list U abundances, since they are mentioned repeatedly in the text. At present one has to use the Th/U ratios from one table and combine them with the Th abundances in another table to get an idea of the U contents. (3) There is no explanation of what (ppt) stands for (part per trillion, or per ton?), and ppt of what? Some solution? Because of this enigma the listed numbers do not mean anything directly. Further back in the table there are then absolute abundances in ppm. Please, put those in the front, and explain all the terms used. (4) Please list the 206/204 ratios. (5) The outcome of the experiments depends very strongly on the qualities and characteristics of the zircon  grains used, but the tables provide no information in merit at all. One may perhaps try to link the individual data to the information in the previous associated paper by these authors. I highly recommend putting a characterization of each grain in the table. Geochronologists know that no two zircons are born alike, and they know that successful dating is best done by a strict discrimination of the good from the bad. A lack of information on the tested grains strongly weakens the interpretations and lessons learned from the study.  

The discussion comprises are section linking a-dosage and degree of discordance, and its implications for the CA-application. It is certainly true that a-dosage and the relative radioactive damage are important factors affecting discordance. But it is also very simplistic, and not realistic, to reduce the degree of discordance to a straight function of a-dosage. Clearly, the textural factors, inclusions, and alteration play a major role, often regardless of U content. Some extremely high-U zircons, which would be destroyed in no-time by CA, can provide concordant U-Pb data if they are just treated gently by air abrasion,  demonstrating the relativity of these indicators. I would recommend that the authors reconsider and re-evaluate their discussion in merit.

There seems to be some confusing concerning the parameters used in the various calculations for data from the literature, such as the 238/235 ratio. I suggest adding a table listing the original information, and the equivalent values calculated with the same constants as in the present paper. It would be useful for the reader, but also a reminder for the authors, avoiding comparisons of apples and oranges.

Sept. 3^rd, 2023   F. Corfu