Mueller et al Review for GChron

Overall this concise paper is really interesting and provides a new method for analyzing olivine samples from young lava flows for more precise cosmogenic 3He exposure age dating when the standard crush/fusion methods can’t resolve the 3He cosmogenic component amongst a persistent mantle He component.

A few points and then line by line comments;

Although probably covered by the blanks (re-extracts) done after each step heat temperature step….I just worry about the 3He/4He ratios of the lower T step heating steps possibly having a higher R because the 3He is successfully diffusing out of the matrix faster than 4He because of the difference in mass. Perhaps a little more discussion could be added on why this worry is likely unfounded? A little on diffusion rates of 3He and 4he with varying T. And/or consider publishing the re-extract data so readers can see that there was zero (or very low) 3He and 4He on the re-extract at 800C? This is really important because the 800C step is the only one that really contributes 3Hec (as stated) by this step heat method.

This is not a huge point, but I’m not sure that the two-step method is always done as the authors suggested, namely that the step after crushing always involved completely powdered olivine/pyx sample. Often researchers just use the crushed leftovers (or sometimes uncrushed phenocrysts) and it works fine as there isn’t much mantle gas, the crushing step opened the most of the hot inclusions and/or the exposure duration and site PR were high enough to overcome some left over mantle gas and get a high % 3Hec. Some labs will try using in vacuo crushed sample material first and then power if needed as the powdering step brings with it a host of problems as the authors state.

Line by line comments:

18 – should give the temp step ranges of the three ranges used rather than just 700-1400C

20 – ‘youngest’…. how? Morphologically, relative age relationships?

24 – I’d remove the ‘for decades’ to avoid confusion

35-37 – I’d list mantle He first and the other sources after that as mantle is the most important for young volcanics

55-65 – I’d just mention that this newer variant of the isochron method was introduced in Blard 2021 and has different axes that the ‘traditional’ isochron method of Cerling and Craig, 1994;  Blard  and Pik , 2008; etc.

66 – again, ‘youngest’ , how do you know they’re the youngest?

69 – not sure what ‘following work on peridotites’ means – has this step heat correction already been done and published before? Or is this saying that the peridotites analyzed here have already been studied which aided in developing this new method? Either way, perhaps rewrite this.

75 – 14C ages are typically given as either 14C yr BP (if uncalibrated) or cal yr BP (if calibrated). The reference that is cited should tell which.

86 – there it is—what youngest means…

88 – I’m not exactly sure what ‘primary depositional features’ are on a lava flow? Maybe primary lava flow morphology, or something like ropy texture suggests that it is a flow top with minimal weathering.

101-105 – so the olivine samples were a mix of phenocryst and xenocryst olivine, as well as possible olivine from ultra-mafic xenoliths? Are they all seeing the same mantle gas environment? All the ways you calculate the 3Hem they look the same at least.

113 – multiple problems with exponents and units in this line

120 – is this a resistance furnace?

125 –  second time with these exponents and units, maybe not a mistake but intentional. Describing a concentration as 10^-3 Matoms is a little confusing, similarly 10^3 M atoms is as well.

145 – so again with the exponent and M – why not just say 10^12 instead of 10^6 M…

159 – here the exponent then M units use is confusing for me with the uncertainties. The first uncertain figure is typically the last significant figure (this is followed very well throughout the manuscript), so 2.35 +/- 0.04 is ok, but not 2.35 +/- 0.14, which should be 2.4 +/- 0.1; here, with the 2.34 x 10^4; the 0.34 part is unnecessary as the number the uncertainty is on has sig figs to the hundredths place x 10^6 which is the 2 in 2.34 x 10^4. Putting the units in just scientific notation without the M makes this clear (like done with fcc and ncc STP/g). Ok, ill give some here, often we provide more digits in uncertainties that would be given by a hard application of that rule. But maybe just one more, not two. For exposure ages… often ages are something like (in ka) 10.2 +/- 1.2 rather than 10 +/- 1; but something like 10.21 +/- 1.23 is too much….the first example would especially be ok if there as another age that was 10.2 +/- 0.9 or some age/uncertainty pair for that suite of samples with a first uncertain figure in the tenths….

I wont comment on this use of sci notation and then Matoms anymore but it continues in later sections. AS this is more of an editing thing I’ll leave it up to the editor whether to demand a correction on this but I’m for not forcing the use of M atoms when the units are something much larger than 10^6. Its just awkward. I get why they do it, 3Hec are often in M atoms! But still off putting for me.

170 – the crush results in this table are not Matoms/g but rather Matoms. No leeway on this, you don’t know how much mass was actually crushed.

172-174 – this assumption is likely untrue and I’m not sure its ‘curious’ to test it in this way (which I’m ok with – testing it in this way that is…), but rather interesting, valuable, or illustrative?

183 – the Thirumalai 2011 paper cited in the figure caption is not in the ref list.

185 – might be worth it here to remind readers that these step heat samples were powdered.

205 – maybe before mantle in the end of the line put ‘persistent’, so its ‘persistent mantle Helium’

207 – crush measurements are M atoms

237 – the 3 on 3He isn’t superscripted

241 – I think these two points could be fleshed out a little more, just a few more sentences, on what a more detailed step heat should be and why it comes at a cost of analytical precision? For readers who will try this method your insight into next steps will be valuable.

255 – and this directly shows the increase in precision using an isochron vs. crush/fusion as suggested by Blard and Pik, 2008; Blard 2021

264 – not sure which PR calculator you’re using, is it the FORMER Cronus Earth calculator, the Balco one…sometimes called the UW calculator, but not sure what it’s routinely called now, I’d just put the web address for clarity.

271 – for the average of the two ages, how is the uncertainty determined? If you average the ages the standard deviation might be better for the uncertainty (about 0.2). If you averaged the 3Hec and determine an age then seems ok to use the internal uncertainty associated with the exposure age calculator (more like 1.1 that is used). Should just say.

283 – Table 3, not a big thing but for most direct measurements you are giving three sig figs. Could do the same for the shielding?

288—maybe remind readers of the association of TS peridotite to these samples? And you set up and used TS as shorthand for Twin Sisters earlier, maybe either use it consistently after that or just don’t worry about shorthand for a location you only mention a few times.

Figures

Fig 1 – I find some of the colors hard to match back to the legend, they are clearly labelled though.

Fig 2 – maybe should put equation of that line inside the figure space rather than in the caption, and include statistical tests on that regression:  r2, p, MSWD etc. Should also note that the line is extrapolated beyond the two data clusters.

Fig 3 – this figure is key to the whole paper and is really interesting. I would shorten the lines, to maybe like +/- 50 deg C around the T that you actually measured 3He and 4He at. As is, it looks like you are declaring that there is some sort of degassing domain change at 900 C and 1200 C but I don’t think you’re really trying to say that?

Please find my review in the attached PDF

Review by P.-H. Blard

This article provides an interesting method based on a sequential heating of olivines from surface lava flow samples. This technique permits to separate the cosmogenic 3He and the mantle 3He component. Their dataset obtained on 2 samples convincingly indicate that diffusion at 800°C preferentially release cosmogenic 3He, permitting to determine this cosmic-ray produced 3He with a better precision than using the standard crushing-melting method. This better precision is possible because the mantle component is not released at “low” temperature, reducing the impact of the uncertainty of the magmatic correction. However, although I think that this pilot study is interesting and should be published, I have major criticisms about the radicality of several conclusions of the authors, who write that the isochron method and the standard crushing-fusion methos yields imprecise or flawed ages. They cannot be so definitive and negative about these well-established methods, for two reasons: first, they didn’t apply the heating step method on the same samples that those processed by the crushing-melting methods, and second, they build an isochron without using samples (or aliquots) that have been exposed to the same “dose” of cosmic rays. The most plausible conclusion is rather that their samples have various exposure ages (either because the lava have different ages, or because the sampled surfaces suffered differential erosion or shielding). I encourage authors to revise their manuscript taking into account this main criticism and also the other points listed below.

Please also take into account these other minor points:

Line 40: “"low" is rather imprecise. Provide actual range of U and Th concentrations.”

Line 42: This belief is not really accurate: young lava (< 100 ka) may require correcting for radiogenic 4He accumulation, because what matters is the 3Hecos/4Herad production ratio, that is independent from the lava age (e.g. Blard and Farley, 2008; Blard, 2021).  Olivines that bear significant amount of U (> 0.1 ppm) require more than 10% correction (R factor lower than 0.9, see for example figure 9 in Blard, 2021). In other words, a "young" lava does not necessarily imply that the radiogenic 4He impact on the magmatic 3He correction is negligible (Dunai and Wijbrans, 2000 = 5% correction; Blard et al., 2006 = up to 12% correction, even in lavas younger than 200 ka ; see R factors in Table 2 in Blard and Farley, 2008).

Line 42: What is a “young” lava?

Line 49: Maybe add a  caution here to mention that dry powdering under atmosphere may yield adsorption of non negligible amount of atmospheric helium on the phenocrysts/xenocrysts (Protin et al., 2016).

Line 51: Quote a source for the value of this atmospheric isotopic ratio.

Line 56: Figure 5 in Blard 2021 shows a modeling of the magmatic helium impact on the final 3He_cos uncertainty.

Line 68: Note that this approach involving a step heating to selectively release cosmogenic 3He at around 1000°C is not a newly developed method. As written in Niedermann 2002: "The few papers which report stepwise heating data for cosmogenic He show the major release of 3Hec from mafic and ultramafic minerals (olivine, pyroxene) below ~ 900-1100°C (Kurz 1986a; Staudacher and Allègre 1991, 1993a; Sarda et al. 1993; Schäfer et al. 2000)."

Line 75: Lava-dammed lakes are not shown on figure 1, could you please add them on this map? It would be very useful to clearly show the stratigraphic relationships between lavas and lakes.

Line 85: “Cosmogenic sampling”. Since "cosmogenic" means "produced by cosmic rays", I suggest to reword.

Line 90: A picture of a sample would be very useful here.

Line 91: What is the typical sample thickness?

Line 97: It would be useful to show field pictures of the sampled lava surfaces (either in Figure 1, or in a new figure). Field photographs would be very useful to show the vegetation cover.

Line 102-103: Are these mm-olivine phenocrysts or xenoliths?

Table 1: Why don't you provide thickness for samples 8, 10 and 11?

Line 119: Along Cox et al 2022, quote here the first article that identified this issue of helium adsorption (Protin et al., 2016).

Line 121: 1200 °C is well below the melting point of Mg-rich olivine (that is above 1500°C). Are you sure of 1) the accuracy of the furnace T control, and 2) are you really reaching fusion (isn't it an He extraction by complete diffusion)?

Line 135: Please provide the blank levels for each T step, as you did in the previous section for 1200°C.

Line 136: It is suprising that the contribution of the 1400°C blank is lower than the one of the 1200°C blank.

Line 153: It would be useful to give the value of this shielding correction here.

Table 2: I think this is not necessary to provide 3He and 4He with 2 different units (cc/g and at/g). This would be much more useful and informative to add a column with the computed cosmogenic 3He concentration corrected for the magmatic 3He for each sample.

Line 172-173: "mantle component": I guess you mean “magmatic 3He/4He ratios”, not magmatic helium concentrations, that need to be variable to ensure the isochron method working. Maybe rephrase to avoid confusion.

Line 176: Given that this isochron is built from samples having different exposure histories, I think you should use this number with a huge caution.

Figure 3: For a better readability, you could consider removing the vertical lines between each T extraction step.

Line 209: Replace "u" by "mu", the greek m.

Line 218-219: Burnard et al., 2015 also provided experimental evidence for the presence of a significant “reservoir” of mantle helium at the grain boundaries.

Figure 4: It would be useful to add an enlarged zoom centered on these "trail" of fluid inclusions.

Line 237 and 240: It would be useful to add these relative increases in cosmogenic 3He in Tables 2 and 3.

Line 240: “34 and 39%”: Why are these 2 numbers different from the 36-42% range mentioned line 237?

Line 241: What do you mean by "detailed step heat"? Do you mean an initial heating at lower tempertature?

Line 253: It is strange to compute a mean using a set of samples with heterogeneous ages.

Line 266: “7.7 +- 4 ka”. I think there are too many significant numbers.

Line 269 to 271: You should rather compare VM1 and 3 with VM2, 6 and 8, not with the isochron that is built from a dataset that breaks the required assumption of an homogeneous dataset.

Line 279: I disagree with this conclusion. You could conclude that if the same 3 methods were applied on the same samples. Here you applied different methods on different surface samples. The most plausible conclusions from your data is that some samples (9, 10, 11) have experienced less exposure at the surface than others (1, 2, 3, 6 and 8). Regarding the building of the isochron, you did not respect the necessary assumption of using aliquots with homogeneous cosmogenic 3He concentrations, so you obtain an underestimate average concentration. Nothing else can be concluded from that.

Line 291-292: As already mentioned before, this conclusion is a false overstatement. You cannot provide such a general conclusion about the isochron and the crush-fusion methods since you did not apply them to the same samples.