Record of low-temperature aqueous alteration of Martian zircon during the late Amazonian

Several lines of evidence support the presence of liquid water on Mars at different times. Among those, hydrated minerals testify to past aqueous weathering processes that can be precisely studied in Martian meteorites such as NWA 7533/7034. Bringing constraints on the timing of weathering of the Martian crust would help understand its evolution, the availability of liquid water, and the habitability of Mars. Here we present a new method based on U–Th–Pb isotope systems to assess if zircon crystals underwent low-temperature aqueous alteration, such as exemplified by Hadean-aged detrital crystals from Western Australia. Data for NWA 7533 zircons show evidence for aqueous alteration and modeling of U–Th–Pb isotope system evolution indicates that the latest alteration event occurred during the late Amazonian (227–56 Ma). This finding largely expands the time duration over which liquid water was available near the Martian surface, thereby suggesting that Mars might still be habitable.

In contrast the Guiteeau and Flauhaut model proposes that the anomalous Th/U ratios are due to decoupling of the U-Th-Pb system in the Martian zircon by low temperature alteration by aqueous solutions and attribute discordance of the Martian zircons to loss of radiogenic Pb. In this model the Th and U are rearranged within the zircon. There is no mention in the model of gain of U and Th from the environmental solutions so, as far as the present reviewer can see, the two models are in no way similar and there is no basis, as suggested in the paper, to draw parallels between the two models.
Consequently the present reviewer cannot recommend the paper for publication in its present form. It could well be that the anomalous Th/U in Martian zircons is the result of interaction with the hydrosphere along the lines of the fluid catalyzed, decoupling -Pb loss model. It is up to the authors to provide convincing evidence in support of this model. But their decoupling model is nothing like the Jack Hills zircon weathering model, which involves addition of U and Th from weathering solutions without significant loss of radiogenic Pb, and in any revision of the paper the differences between the two models should be pointed out or the JH weathering paper could be mentioned in passing or not at all, with the revision focusing on a description of the decoupling model as a new stand-alone concept.

Some specific comments
Line 41. The oldest terrestrial zircon is ~ 4375Ma. The reviewer has no idea how the authors extracted an age of 4438 Ma from the reference Valley paper .
Line 44-45. "Radiation damage accumulates" Line 47. Can be mobilized Iine 48. Remove " aging": Iine 50. As far as the reviewer is aware references 17-19 do not deal with oxygen isotopes Line 52 What are advanced signs of radiation damage?
(line 68 -71 and also line 72-77) . These sentences misrepresent what is described in the Pidgeon et al. paper. They interpret the Pidgeon et al. paper as saying that the anomalous Th/U ratios are associated with areas of high radiation damage in the zircon and are the result of "decoupling" of Th from U. However, it is reported in the Pidgeon et al. paper that anomalous Th/U, enhanced U and Th concentration, elevated δ18O and common Pb , anomalous OH and U-Th-Pb zero age discordance are found in SIMS analyses overlapping cracks and are the result of materials including water being added to the zircon during penetration of aqueous fluids into cracks and metamict parts. It is not correct to attribute the anomalous Th/U results of the Jack Hills to decoupling. This term is not mentioned in the Pidgeon et al. paper .
Also, The present authors deal only with the Th/U ratios. The Pidgeon et al. paper looks at a combination of parameters besides Th/U , including U and Th contents, Th-U-Pb discordance, common Pb content, δ18O, and location of SIMS analysis on the grain -e.g. whether the SIMs analysis overlaps a crack or a zone. The present authors would need to review these additional parameters (where the data are available) to properly assess the likelihood that the Jack hills zircon low temperature fluid interaction mechanism could have affected the Martian zircons.
line 68 73 and elsewhere. Decoupling is the fundamental mechanism referred to in the Guitreau Flahaut paper and the term " decoupling ": needs to be defined. Does it refer to independent movement of U-Th and Pb within a zircon during low temperature fluid interaction? Is the zircon viewed as a closed system to U, Th and Pb? Or is it envisaged that radiogenic Pb can escape during weathering interaction?. What about U and Th ? Is it envisaged that these elements are contained within the zircon or can these also be leached out of the zircon during low temperature aqueous interaction?
Line 72 to 75. This raises the issue of the α-dose shown Fig2 It should be kept in mind that zircons can be annealed in nature and it has been shown that for the JH zircons the α-dose determined from the 207/206 age and the measured Th and U is a gross overestimate of the dose needed to produce the observed radiation damage in the zircon. This is referred to briefly in your method section. Line 82. The sentence "Lead mobility which accounts for most cases of U-Pb discordance". contradicts the conclusions of the Pidgeon et al. paper that discordance of the Jack Hills zircons is due to U and Th gain. There are a number of hydrothermal experiments that demonstrate episodic loss of radiogenic Pb from radiation damaged zircon . These experiments result in annealing of the zircon. In low temperature weathering it is envisaged that solutions infiltrate the zircon mainly through cracks but also into radiation damaged zones ( to be demonstrated) where they precipitate trace elements including U and Th but cause no annealing of the radiation damage and little if any loss of radiogenic Pb .(except possibly in some damaged zones which is yet to be demonstrated). Essentially all the U-Th-Pb discordance is the result of U and Th gain and not Pb loss.
Line 86 -104 Presents arguments for limited Th/U variation in natural systems.
Line 96. This process would probably also remove much of the U and Th Line 97-100. With the possible exception of ICPMS data, Pb-Pb and U-Th-Pb ratios are routinely corrected for common Pb using 204 Line 105 the sentence starting. " Available data presented in supplementary material (Fig.2S, Table  S3) etc. Is one of a number of examples of the use of undefined terms such as " first percolation point" ( used elswhere) and unexplained statements such as " expected from current knowledge of lunar surface evolution" and would be incomprehensible to anyone not totally familiar with radiation damage and lunar literature 105 and on. This is difficult to review as on the one hand it interprets the anomalous Th/U as evidence low temperature alteration " much as what was experienced by crystals from Western Australia " and on the other applies the data to "constrain the age of decoupling between Th/U and Pb isotopes", which is not what was experienced by crystals from Western Australia. With this in mind the reviewer has not extended the review into the section which sets out to determine the age of weathering.
Reviewer #3 (Remarks to the Author): Review of "Record of low-temperature aqueous alteration of Martian zircon during the late Amazonian" by Guitreau and Flahaut.
The timing of aqueous alteration on Mars is of fundamental importance in planetary science and astrobiology. The present study investigated this issue through analysis of previously published U-Th-Pb isotope data for zircon in two ancient Martian meteorites. Specifically, the authors constrain the timing of secondary Th/U fractionation in Martian zircon by comparing measured Th/U with "time-integrated Th/U" calculated from measured 208Pb/206Pb (i.e., ratio of 232Th and 238U daughter nuclides), given that such secondary Th/U fractionation has been documented in aqueously altered terrestrial zircon (Pidgeon et al., 2017). The analytical results revealed that Th/U fractionation has occurred in geologically young ages within one Martian zircon grain (NWA 7533-Z2). Assuming that the zircon grain initially had a Th/U ratio between 0.5 and 1.0, the timing of Th/U fractionation was estimated as between 564 and 42 Ma. This led the authors to argue that aqueous alteration should occur during the late Amazonian on Mars.
The subject addressed in the present study is of wide interest and the approach is novel. I am afraid, however, that the secondary Th/U fractionation observed in zircon NWA 75333-Z2 may be due to recent aqueous alteration on the Earth rather than Mars. Considering ample evidence of terrestrial alteration for the Martian meteorite (Lorand et al., 2015), this possibility should be critically examined before the manuscript is acceptable for publication. The possibility might be excluded if the minimum age estimate of 42 Ma for the Th/U fractionation was robust. Unfortunately, however, the age estimate depends highly on the assumed initial Th/U ratio in the zircon grain: the age of secondary Th/U fractionation can be down to ~0 Ma when the initial Th/U ratio is assumed to be ~2.0, which is well within the range of measured Th/U ratios in other Martian zircon grains that have not undergone recent Th/U fractionation (Fig. 3). Alternatively, positive ∆17O values obtained from zircon NWA 75333-Z2 (Nemchin et al., 2014) may be taken as evidence for alteration on Mars. Please note, however, that the elevated values could be analytical artifacts due to increased 16OH/16O ratios, as observed for altered terrestrial zircon grains (Pidgeon et al., 2017). Thus, I feel that the conclusion of late Amazonian alteration on Mars is still unsound.
Additional comments: Solubility of U Th/U fractionation during aqueous alteration is a result of the higher solubility of U in fluid than Th. Note, however, that the U solubility is dependent on pH and fO2 conditions. For instance, it is well known that U has been essentially insoluble under reduced surface conditions on the Archean Earth. Thus, it needs to be evaluated whether Th/U fractionation is viable during aqueous alteration under Mars surface conditions. Radiation damage or fracture It is certainly true that radiation damage is an important factor for controlling the effect of secondary alteration on zircon chemistry. In the case of Martian zircon NWA 75333-Z2, however, fractures in analytical spots rather than radiation damage seem to be the main cause for the elevated measured Th/U ratios (see images of the grain presented in supplementary information of Nemchin et al., 2014). Indeed, Pidgeon et al. (2017) demonstrated that analyses which overlap a fracture yield highly elevated Th contents and Th/U ratios, corresponding well with the data for NWA 75333-Z2. 1

Note: line numbers given in our responses referred to the final version of the manuscript (with track-changes accepted).
Reviewer #1 comments: 1) Authors did not get to the Martian zircons until 3 pages into the main text. It would be better to add a few introductory sentences at the beginning of the main text.
Response: As suggested, we added some text to introduce Martian zircons lines 51-55, just after describing the zircon mineral specificities.
2) It is unclear how authors derived the alteration age of zircons to be between 42 and 564 Ma? These ages appeared rather suddenly on line 132 to 135. Since this is the crux of the results, please explain how the ages are calculated.
Response: We have further developed explanations about our model and its results at lines 178-217 to make our point more explicit. In addition, we have modified the caption to Figure 5 (where the time-integrated Th/U can be related to the alteration age) to assist the reader in understanding the results of our two-stage model. In addition, results for the three-stage model are now presented as a table (Table 1)  Response: As suggested, sample numbers were inserted next to the data points. 4) Natural communication allows more pages. It would be helpful for authors to move some supplementary materials to the main text, so that the outlines of how Response: Thank you for this suggestion. As proposed, we have increased the text length to add some key details about the model within the main body of the article. We believe it makes the manuscript much clearer and allowed us to address the comments of the second and third reviewers at the same time (see below).
Minor comments: Line 19: should be "Evidence exists that supports the " Response: Thank you for spotting this. We changed the text accordingly.
Response: We have added this reference as suggested.
Reviewer #2 comments: Guitreau and Flahaut propose that anomalous Th/U ratios in some zircons from a Martian meteorite can be explained by interaction with weathering solutions in much the same way as anomalous Th/U in ancient zircons from the Jack.
However the explanation of these authors of a decoupling of the Th-U-Pb system of the Martian zircons as a result of weathering is very different from the mechanism proposed for the weathering of the Jack Hills zircon by Pidgeon et al.. Weathering of the Jack Hills zircons involves transport of U and Th and other elements present in surrounding weathering fluids into the radiation damaged zircons via cracks and possibly by percolation into highly metamict domains. This results in an increase in the U and Th concentrations and the measured Th/U. U-Th-Pb discordance is interpreted as the result of U and Th gain not Pb loss and any loss of radiogenic Pb is viewed as insignificant.
In contrast the Guitreau and Flauhaut model proposes that the anomalous Th/U ratios are due to decoupling of the U-Th-Pb system in the Martian zircon by low temperature alteration by aqueous solutions and attribute discordance of the Martian zircons to loss of radiogenic Pb. In this model the Th and U are rearranged within the zircon. There is no mention in the model of gain of U and Th from the environmental solutions so, as far as the present reviewer can see, the two models are in no way similar and there is no basis, as suggested in the paper, to draw parallels between the two models.
Consequently the present reviewer cannot recommend the paper for publication in its present form. It could well be that the anomalous Th/U in Martian zircons is the result of interaction with the hydrosphere along the lines of the fluid catalyzed, decoupling -Pb loss model. It is up to the authors to provide convincing evidence in support of this model. But their decoupling model is nothing like the Jack Hills zircon weathering model, which involves addition of U and Th from weathering solutions without significant loss of radiogenic Pb, and in any revision of the paper the differences between the two models should be pointed out or the JH weathering paper could be mentioned in passing or not at all, with the revision focusing on a description of the decoupling model as a new stand-alone concept.
Response: We completely agree with the reviewer about Jack Hills zircons. In fact, we realize that the reviewer misinterpreted our message, because nowhere in the main text it was written that we interpreted Pidgeon et al. paper to say that Pb-loss is the cause for U-Pb discordance in Jack Hills zircons. However, it is true that the wording in the method section may have been misleading and we apologize for this issue. We have rewritten the text in a more explicit manner and have defined terms that the reviewer was confused about.
Line 41. The oldest terrestrial zircon is ~ 4375Ma. The reviewer has no idea how the authors extracted an age of 4438 Ma from the reference Valley paper. Line 68 -71 and also line 72-77. These sentences misrepresent what is described in the Pidgeon et al. paper. They interpret the Pidgeon et al. paper as saying that the anomalous Th/U ratios are associated with areas of high radiation damage in the zircon and are the result of "decoupling" of Th from U. However, it is reported in the Pidgeon et al. paper that anomalous Th/U, enhanced U and Th concentration, elevated ∂18O and common Pb, anomalous OH and U-Th-Pb zero age discordance are found in SIMS analyses overlapping cracks and are the result of materials including water being added to the zircon during penetration of aqueous fluids into cracks and metamict parts. It is not correct to attribute the anomalous Th/U results of the Jack Hills to decoupling. This term is not mentioned in the Pidgeon et al. paper .
Also, The present authors deal only with the Th/U ratios. The Pidgeon et al. paper looks at a combination of parameters besides Th/U, including U and Th contents, Th-U-Pb discordance, common Pb content, ∂18O, and location of SIMS analysis on the grain -e.g. whether the SIMs analysis overlaps a crack or a zone. The present authors would need to review these additional parameters (where the data are available) to properly assess the likelihood that the Jack hills zircon low temperature fluid interaction mechanism could have affected the Martian zircons.
Response: As stated above, our initial description was unclear and probably clumsy, as our message was not understood. We have made significant changes to the text, added a clear description of the model, and defined the term "decoupling " (which is in the end, the discrepancy between the measured Th/U and the time-integrated Th/U). We totally agree with Pidgeon et al. conclusions about Jack Hills zircons and we are sorry if they were not well reported in our initial version of the text. We have significantly modified the text so as to accurately represent Pidgeon et al. conclusions.
Line 68-73 and elsewhere. Decoupling is the fundamental mechanism referred to in the Guitreau Flahaut paper and the term "decoupling ": needs to be defined. Does it refer to independent movement of U-Th and Pb within a zircon during low temperature fluid interaction? Is the zircon viewed as a closed system to U, Th and Pb? Or is it envisaged that radiogenic Pb can escape during weathering interaction? What about U and Th? Is it envisaged that these elements are contained within the zircon or can these also be leached out of the zircon during low temperature aqueous interaction?
Response: In the present manuscript, decoupling is not used to describe a mechanism but only used as a descriptive term. It merely illustrates the fact that measured and time-integrated Th/U ratios are different, no matter which mechanism or process is responsible for it. We have made this definition more explicit in the text (l.83-84).
Line 72 to 75. This raises the issue of the alpha-dose shown Fig2 It should be kept in mind that zircons can be annealed in nature and it has been shown that for the JH zircons the alpha-dose determined from the 207/206 age and the measured Th and U is a gross overestimate of the dose needed to produce the observed radiation damage in the zircon. This is referred to briefly in your method section.
Response: We have emphasized the inaccurate nature of radiation doses in the revised manuscript (e.g., l. 91-98).
Line 82. The sentence "Lead mobility which accounts for most cases of U-Pb discordance". contradicts the conclusions of the Pidgeon et al. paper that discordance of the Jack Hills zircons is due to U and Th gain. There are a number of hydrothermal experiments that demonstrate episodic loss of radiogenic Pb from radiation damaged zircon. These experiments result in annealing of the zircon. In low temperature weathering it is envisaged that solutions infiltrate the zircon mainly through cracks but also into radiation damaged zones ( to be demonstrated) where they precipitate trace elements including U and Th but cause no annealing of the radiation damage and little if any loss of radiogenic Pb .(except possibly in some damaged zones which is yet to be demonstrated). Essentially all the U-Th-Pb discordance is the result of U and Th gain and not Pb loss.
Response: In this sentence we were precisely referring to cases when zircon annealing occurred because we are discussing if Pb-loss could produce the decoupling we observe. This sentence was not at all referring to alteration such as seen by the Jack Hills zircons. We totally agree about the U and Th increase and have made this idea more explicit in the text.
Line 86 -104 Presents arguments for limited Th/U variation in natural systems.
Response: We were only referring to observed Th/U in natural zircon. We have clarified this point.
Line 96. This process would probably also remove much of the U and Th 5 Response: We respectfully think different as U and Th substitute relatively well to Zr in the zircon lattice. Therefore, U and Th would still be present. It is true, though, that U and Th concentrations would be moderate.
Line 97-100. With the possible exception of ICPMS data, Pb-Pb and U-Th-Pb ratios are routinely corrected for common Pb using 204.
Response: We have added this idea in the revised text (l.133-134).
Line 105 the sentence starting. " Available data presented in supplementary material (Fig.2S, Table S3) etc. Is one of a number of examples of the use of undefined terms such as " first percolation point" ( used elswhere) and unexplained statements such as "expected from current knowledge of lunar surface evolution" and would be incomprehensible to anyone not totally familiar with radiation damage and lunar literature Response: We have defined first percolation point (lines 43-48) and have tried our best to be more explicit in the text (e.g., we have inserted a couple of sentences on lunar zircons lines 140-144).
Line 105 and on. This is difficult to review as on the one hand it interprets the anomalous Th/U as evidence low temperature alteration "much as what was experienced by crystals from Western Australia " and on the other applies the data to "constrain the age of decoupling between Th/U and Pb isotopes", which is not what was experienced by crystals from Western Australia. With this in mind the reviewer has not extended the review into the section which sets out to determine the age of weathering.
Response: As stated before, we have made significant modifications to the main body of the text to clarify our model and interpretation of the observed "decoupling", we are grateful for the comments which helped us rework the manuscript in a much clearer way and hope the reviewer is satisfied with these changes.

Reviewer #3 comments:
The subject addressed in the present study is of wide interest and the approach is novel. I am afraid, however, that the secondary Th/U fractionation observed in zircon NWA 75333-Z2 may be due to recent aqueous alteration on the Earth rather than Mars. Considering ample evidence of terrestrial alteration for the Martian meteorite (Lorand et al., 2015), this possibility should be critically examined before the manuscript is acceptable for publication. The possibility might be excluded if the minimum age estimate of 42 Ma for the Th/U fractionation was robust. Unfortunately, however, the age estimate depends highly on the assumed initial Th/U ratio in the zircon grain: the age of secondary Th/U fractionation can be down to ~0 Ma when the initial Th/U ratio is assumed to be ~2.0, which is well within the range of measured Th/U ratios in other Martian zircon grains that have not undergone recent Th/U fractionation (Fig. 3). Alternatively, positive Δ17O values obtained from zircon NWA 75333-Z2 (Nemchin et al., 2014) may be taken as evidence for alteration on Mars.
6 Please note, however, that the elevated values could be analytical artifacts due to increased 16OH/16O ratios, as observed for altered terrestrial zircon grains (Pidgeon et al., 2017). Thus, I feel that the conclusion of late Amazonian alteration on Mars is still unsound.
Response: We agree that this is a very important issue to address. The question of whether alteration observed in some martian meteorites (e.g., precipitated carbonates or sulfates in shergottites) could be of terrestrial origin has been an ongoing debate for decades. However, we think that the reviewer's points are not a real issue in the case of our zircons, because the analyzed domains that show Th/U decoupling do not show anomalous Δ17O values but rather regular mass-dependent fractionation along the Martian fractionation line. Therefore, the data are robust and not biased by analytical artifacts. Furthermore, assuming an initial Th/U ratio of 2 for the Martian zircons, as proposed by the reviewer, is simply impossible as the initial Th/U cannot be larger than the time-integrated one given that these crystals experienced an increase in Th/U ratio. Moreover, careful examination of textural and chemical characteristics of most well-behaved magmatic zircons indicate that their measured Th/U range between 0.4 and 0.8, which is very consistent with high-precision ID-TIMS measurements that gave an even narrower range of 0.6 to 0.7. This range is perfectly compatible with Th/U in terrestrial magmatic zircons, as well as in pristine lunar zircons. Therefore, we judge that a Th/U of 1, contrary to what the reviewer suggests, marks an absolute upper-limit for Martian magmatic zircons. Consequently, we respectfully think that comments made by reviewer 3 have no bearings on our study. Regarding the robustness of determined ages, we have refined our model so as to provide a more robust and more realistic estimate and we now think that our results are definitively conclusive.
Additional comments: Solubility of U Th/U fractionation during aqueous alteration is a result of the higher solubility of U in fluid than Th. Note, however, that the U solubility is dependent on pH and fO2 conditions. For instance, it is well known that U has been essentially insoluble under reduced surface conditions on the Archean Earth. Thus, it needs to be evaluated whether Th/U fractionation is viable during aqueous alteration under Mars surface conditions. Response: The reviewer raises an important point here. In fact, Martian conditions are oxidizing and therefore perfectly consistent with an alteration event like the one we envisage to happen on Mars. We have added some text discussing redox conditions on Mars and the possibility for U to be mobilized.
Radiation damage or fracture It is certainly true that radiation damage is an important factor for controlling the effect of secondary alteration on zircon chemistry. In the case of Martian zircon NWA 7533 Z2, however, fractures in analytical spots rather than radiation damage seem to be the main cause for the elevated measured Th/U ratios (see images of the grain presented in supplementary information of Nemchin et al., 2014). Indeed, Pidgeon et al. (2017) demonstrated that analyses which overlap a fracture yield highly elevated Th contents and Th/U ratios, corresponding well with the data for NWA 7533-Z2.