the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 15 Apr 2025
Weathering without inorganic CDR revealed through cation tracing
Abstract. Enhanced Weathering using basalt rock dust is a scalable carbon dioxide removal (CDR) technique, but quantifying rock weathering and CDR rates poses a critical challenge. Here, we investigated inorganic CDR and weathering rates by treating mesocosms planted with corn with basalt (0, 10, 30, 50, 75, 100, 150 and 200 t ha⁻¹) and monitoring them for 101 days. Surprisingly, we observed no significant inorganic CDR, as leaching of dissolved inorganic carbon did not increase, and soil carbonate content even declined over time.
To gain insights into the weathering processes, we analyzed the mass balance of base cations, which can be linked with anions (including HCO3-) through charge balance. This mass balance showed that most base cation charges were retained as (hydr)oxides in the reducible pool of the top soil, while increases in the exchangeable pool were about a factor 10 smaller. Soil base cation scavenging exceeded plant scavenging by approximately two orders of magnitude. From the base cations in all pools (soil, soil water and plants), we quantified log weathering rates of -11 mol TA m-2 basalt s-1 and a maximum CO2 removal potential of the weathered base cations (i.e., CDR potential) of 18 kg CO2 t⁻¹ basalt.
For climate change mitigation, not only the amount of CDR potential is important, but also the timescale at which that CDR would be realized. Our data suggests that the lag time for realization of inorganic CDR may be larger than commonly assumed. In conclusion, we observed that inorganic CDR was not directly linked to rock weathering in the short-term. Still, the observed increases in secondary minerals and base cation exchange may provide valuable benefits for soil fertility and organic matter stabilization in the long-term.
- Preprint
(1265 KB) - Metadata XML
-
Supplement
(2344 KB) - BibTeX
- EndNote
Status: open (extended)
-
CC1: 'Comment on egusphere-2025-1667', Adam Wolf, 21 Apr 2025
reply
Thank you for reporting the results of this fascinating study. Ordinarily I would reach out directly, but I appreciate the EGU ethos to encourage open discussion over draft papers.
Q. Could you say more about the C-value of the BET results you obtained? 9.2 m2/g is a high value, particularly for a D50 of ~310um. We have been using the C-value as a QA indicator - "good" values were below 100, or perhaps up to 150, whereas bad values are >200. The highest BET value we obtained, 9.6m2/g, had a C value of 1500. We don't believe these high BET values are real.
Q. Could you comment more on the types of secondary minerals that were formed? I am curious the extent to which you believe these are Al-oxides (intrinsic to basalt) or Fe-oxides (more prominent from olivine). Or are these amorphous aluminosilicates?
Q. The soil has low CEC (3), low base sat (50%), and low carbonates (0.003%). Irrigation is with rainwater that is presumably slighly acidic. Why then does the pH in the control plot increase from 5.6 to 7.2?
Comment: The application rates are quite large. Farmers are reluctant to put on more than 10-20 t/ha out of concerns for compaction, but your smallest replicated rate is 50t/ha and you go up to 200 t/ha. I think it is plausible that such high application rates put the system into a regime that is unlikely to be encountered in nature - indeed it jumps from ca 5.5 to ca 6.5 in a month and 7.5 in two months. If the target for a crop is 6.5, then these rates are un-agronomic. The question that arises is: to what extent are your results, specifically around development of secondary minerals, shaped by the high application rates and attendant abrupt shifts in soil chemical equilibria?
Comment: This is the second paper I have read from your group recently that reports results for a relatively short timeframe, with a key takeaway being the retention of the reaction products within the soil column. Our (Eion's) work in the field suggests that there are two regimes: one where ET > Precip due to growing crops, and a subsequent post-harvest phase where Precip > ET and reaction products are flushed out. It appears in this experiment that ~25 L of water was applied and ~15 L of water was collected as leachate, so indeed there is some flushing in the present work. However, I would be curious for the experiment to be run out longer to understand the fate of the secondary minerals that were formed. Do they subsequently dissolve and release these cations as the soil re-acidifies? Are there field trials established contemporaneously in similar soils to address this question?
Comment: You cite Kanzaki et al (2024) four times, but so far as I know this work was not accepted for publication.
Citation: https://doi.org/10.5194/egusphere-2025-1667-CC1 -
AC1: 'Reply on CC1', Arthur Vienne, 21 May 2025
reply
Dear Adam,
Thank you for your interesting input. Apologies for our late reply, please find below my answers to your questions:Q. Could you say more about the C-value of the BET results you obtained? 9.2 m2/g is a high value, particularly for a D50 of ~310um. We have been using the C-value as a QA indicator - "good" values were below 100, or perhaps up to 150, whereas bad values are >200. The highest BET value we obtained, 9.6m2/g, had a C value of 1500. We don't believe these high BET values are real.
9 m²/g is indeed a high value, and although high, this value should still be realistic. In the work of Lewis et al. (2021) surface roughnesses (BET-SSA/ geometric SSA) have a broad range from only 13 to 2423. We find a geometric surface area of 0.0064 (=6/3/ 310) with 310 µm the p80 value and 3 the density of basalt, equation of Tester et al. (1994)) so that the roughness becomes 1434 (within the range observed with a variety of basalts by Lewis et al. (2021)).
Some more analytical details on the BET-SSA of 9 m²/g: All the C-values are positive, with a range from 57 to 63. This sample was measured using Krypton gas with 11 points. To be sure, we reanalyzed BET-SSA with N2 gas. For N2 gas, we get a value of 6.283 m²/g. SSA’s measured with N2 gas can be 50% lower than samples analyzed with Kr gas. We will add this additional measurement to the revised manuscript and comment and report both measurements.
Q. Could you comment more on the types of secondary minerals that were formed? I am curious the extent to which you believe these are Al-oxides (intrinsic to basalt) or Fe-oxides (more prominent from olivine). Or are these amorphous aluminosilicates?
This is an interesting question and we are also very curious to know the type of secondary minerals that are formed. Unfortunately these cannot be easily determined in the soil, where the minerals that are already present at the start obscure the signal. We did try to link stoichiometry of extracted pools with existing crystalline secondary clay minerals (see Fig S.22) but did not find a good match. There is existing evidence on the formation of short-range-order amorphous compounds that contain Al, Si, Fe in volcanic soils. We therefore expect that most of the secondary minerals are amorphous. The Al and Si could both be present as an amorphous (hydr)oxide or as an amorphous aluminosilicate.
Q. The soil has low CEC (3), low base sat (50%), and low carbonates (0.003%). Irrigation is with rainwater that is presumably slighly acidic. Why then does the pH in the control plot increase from 5.6 to 7.2?This is indeed an interesting observation. While rainwater is slightly acidic, the rainwater used for extra irrigation was stored in an underground tank that released some calcium, increasing the pH. Potentially the fertilization with Ca nitrate may have affected soil water pH as well. The Ca-leaching tank was not yet mentioned in the previous version of the manuscript but will be included.
Comment: The application rates are quite large. Farmers are reluctant to put on more than 10-20 t/ha out of concerns for compaction, but your smallest replicated rate is 50t/ha and you go up to 200 t/ha. I think it is plausible that such high application rates put the system into a regime that is unlikely to be encountered in nature - indeed it jumps from ca 5.5 to ca 6.5 in a month and 7.5 in two months. If the target for a crop is 6.5, then these rates are un-agronomic. The question that arises is: to what extent are your results, specifically around development of secondary minerals, shaped by the high application rates and attendant abrupt shifts in soil chemical equilibria?
We agree that the utilized application rates are high and not agronomic. Our experiment demonstrates that adding more and more basalt will increase the risk of oversaturation and secondary mineral formation. Keeping application rates low in practice is thus not only more feasible, but likely also more effective for inorganic C export than high application rates. Repeated annual application of high application of (50 t/ha) basalt has been suggested previously (e.g. https://www.pnas.org/doi/10.1073/pnas.2319436121), but our dose-response experiment demonstrates that this may not the best choice to make.
Comment: This is the second paper I have read from your group recently that reports results for a relatively short timeframe, with a key takeaway being the retention of the reaction products within the soil column. Our (Eion's) work in the field suggests that there are two regimes: one where ET > Precip due to growing crops, and a subsequent post-harvest phase where Precip > ET and reaction products are flushed out. It appears in this experiment that ~25 L of water was applied and ~15 L of water was collected as leachate, so indeed there is some flushing in the present work.
Thank you for raising this interesting point. We fully agree on the importance of the water balance for retention of reaction products as well as for secondary mineral formation.Indeed, in our experiment, precipitation exceeded ET and there was no runoff in the mesocosms. Over the course of the experiment (99 days), our mesocosms received approximately 450 mm of rainwater (partly via irrigation from rainwater in the tank and partly via direct natural rainfall) (see Fig S2 (a)). This is actually much higher than the typical amount of rain in Flanders (on average ~ 75 mm per month, which is equivalent to ~250 mm for a 100 day period). The summer of 2021 was one of the wettest summers ever in Belgium.
However, I would be curious for the experiment to be run out longer to understand the fate of the secondary minerals that were formed. Do they subsequently dissolve and release these cations as the soil re-acidifies?This is an interesting question that we are trying to answer in follow up experiments but that can unfortunately not be addressed with the current experiment that was destructively harvested after 100 days. You may be interested to read this preprint from our team, reporting results from a 15-month experiment: https://www.researchsquare.com/article/rs-5672251/v1.
Q: Are there field trials established contemporaneously in similar soils to address this question?
We are also executing field trials but with another soil and basalt type, thus far these results have not been published yet as it is work in progress.
Comment: You cite Kanzaki et al (2024) four times, but so far as I know this work was not accepted for publication.Indeed, this is still a pre-print, yet the valuable concept of TA retention is discussed in their preprint. To our knowledge, no published manuscripts address this point.
Hereby, I hope to have answered to all your questions and comments.
Best regards,
Arthur VienneCitation: https://doi.org/10.5194/egusphere-2025-1667-AC1
-
AC1: 'Reply on CC1', Arthur Vienne, 21 May 2025
reply
-
RC1: 'Comment on egusphere-2025-1667', Anonymous Referee #1, 08 Jul 2025
reply
Main points:
- “Solid soil pools”. Analysis of different soils fractions to assess potential for uptake of weathered base cations is a key focus of the study. This needs better introducing, better assessment and better quantification of uncertainties.
In the introduction, expand the paragraph starting on L78 to include a discussion of how the reducible and oxidisable fractions take up base cations (specifically Mg, Ca, K and Na) and provide key references from the literature that provide evidence for this (and presumably motivated the approach used in this study).
Chemical leaching techniques are notoriously unreproducible and may attack other phases in addition to the target phases. The reproducibility of leaching needs to be reported- it is not clear how the error bars shown in for example Fig. 5 were generated. Multiple (n=??) separate analyses of the same soils and/or multiple experiments (at least for the control and 50 t applications)? Was uncertainty associated with control soil and basalt feedstock propagated? There also needs to be assurance as to what phases were released- how do the authors know that the target phases were indeed those released? The authors seem to doubt this themselves in L473-480. Basalt and soil will likely respond very differently to chemical leaching, yet it is assumed they respond the same.
Related, the authors then assess whether the soil reducible pool corresponds to clays (L473)- is the reducible fraction is Fe- and Mn-(oxyhydr)oxides (as stated in L172, or not? The paragraph L473-480 undermines the rest of the paper- saying more research is needed is not adequate (it is something I expect to read in an undergraduate dissertation, not a manuscript submitted for peer-review publication).
- Terminology. “Inorganic CDR” is used widely (even in the title!) but is not defined. Similarly, “CDR potential” should be explicitly defined and these terms should be compared with those in the literature (expanded under Introduction below). There is massive confusion around the definition of CDR and differences in quantification approaches used by the EW community, and this paper as it stands only adds to that. It’s really important to sort this out, to avoid accusations of green-washing and worse.
Title: Avoid using acronyms in titles. Title is not accessible and need to be revised- few will have any idea what “inorganic CDR” is, or “cation tracing”
Abstract: Needs substantial revision to improve clarity and accessibility. The final paragraph especially is vague and qualitative.
L24: Do not make subjective comments, e.g., delete “surprisingly and “even” (L25). If the term “inorganic CDR” is to be used, it needs to be defined. have expertise in the field but I am not familiar with this terminology.
See comments below on soil pools.
L31: What is CDR potential? Total dissolution of applied rock? Or potential over a time period? The abstract needs to stand alone (many will only read this), so it is essential to use accessible and inclusive language.
L34: refers to time, so “larger” should be replaced with “longer”. Please add timeframes- how long do your data suggest? And what timescale is “commonly assumed”? I’m unaware that as yet there are any commonly assumed timescales.
Introduction:
L43: Need to sort out terminology. TA is not a proxy for DIC; these are two distinct variables. TA is not “the sum of base cation charges”, linked to this, what is delta in Eq. 1?
L63: changes in DIC during soil water transport have been known about for years, not recently as indicated here. Be sure to include relevant refs from soils literature.
L70: temporally? Is this correct? Or do you mean temporarily?
L90: This paragraph needs revising. DIC and TA are being used interchangeably; the first part refers to DIC then this transforms into “scavenged TA”. As above, DIC and TA are not the same thing: make sure they are being used correctly.
L94: CDR potential is then CDR predicted from measured weathering of base cations from the applied feedstock. It would be helpful to express it as such, and to note also that it is not the same thing as CDR estimated using a TiCAT approach, because although TiCAT includes cations held in the exchangeable fraction, it does not include CDR associated with uptake of base cations on carbonates, reducible, oxidisable or secondary mineral pools, since these are retained in the soil. It should be noted also that CDR potential as defined here is not the same as CDR potential defined in Beerling et al. (2024). Making such comparisons is essential as differences in CDR approaches and terminology are leading to a lot of confusion even within the EW community, and even more beyond it.
Material and Methods:
L122 and elsewhere: missing subscripts/superscripts
L227 and elsewhere, make sure all acronyms are spelt out on first use (here, AIC)
L153 and elsewhere: please report accuracy and precision of all measurements, and how these were determined.
L262 - 270: stating that cations in the feedstock rock were “already weathered initially” is confusing- you just mean that you are correcting for the amount of cations initially present in that fraction in the applied feedstock- so it would be better to say this. Additionally it is not correct to say that these corrections have not been applied before (L266), this was already pointed out by Power et al. (2025).
L274 (& Table 4): how n changes during transport through the soil profile, into rivers & into the ocean is not tested in this study, hence I suggest the endmembers should be n=1 and n=0.5.
L294-5: Sentence doesn’t make sense. Note also when log SIc>0 minerals have potential (not “tendency”) to precipitate: it is well documented that for example that in rivers, calcite generally does not precipitate until log Sic>1 due to ion inhibition, e.g., by phosphate. L296: delete “perfect”
L326: not clear what “gave an even larger (signal)” means- do you mean there was higher accumulation of Mg in the reducible pool compared to the exchangeable? Why is this “borderline significant”- because the variability was very high?
L329: You don’t know what phase(s) Al is in- you just know that Al is being found in association with the oxidisable or reduced fraction, so please change to say this.
L334: Delete “even”. Why would concs of Na, Fe and Mg decrease in the oxidisable fraction? It is stated elsewhere (e.g., L100) that cations bound in these phases are unlikely to be released. Presumably the TOC content of the 20-30 cm layer did not decrease significantly? To me this hints at artefacts from the leaching procedure.
L348: Does this also include Na, K and Ca in the soil pore waters?
Fig. 6: Some of the text is not legible. For the 0-20cm fraction, would be informative to also show the quantity of cations residing in the soil waters.
Discussion:
L415: delete “shake flask”; L416: delete first “that”
L417: Is the reader expected to know what CDR export via TA was in Amann et al. (2020)? Please make sure all parts of the discussion are clear and transparent.
L424: was carbonate present in the initial soil and/or basalt? I.e., was it there to dissolve?
L427: Be aware that the cited study neglected to account for application of lime: there is no way that the increased SIC could have come from basalt weathering as it was way too large relative to the basalt application rate.
L433: word missing “exchangeable and pools”?
L435-437: Sentence not clear, seems to be incorrect punctuation and/or words missing
L432-443: There have been lots of studies on basalt dissolution, yet only one is briefly mentioned at the end of this paragraph. Expand this discussion to put your data in the context of existing work from the soils literature, which is far more extensive than EW literature.
L446 and elsewhere: re-phrase “shake flask” or at the least explain what on earth this means
L447: why should base cations be “irreversibly dissolved”? Because there is no uptake on exchange sites and conditions are far from equilibrium? I can believe the latter but it is hard to see how the former can be true.
L449: delete “only”- again, subjective comments have no place in scientific writing!
L457: add “over the timeframe of our experiments (111 days)” to the end of this sentence
L465: I think also Li isotopes, see Pogge von Strandmann papers
L467: change to “such as kaolinite that do not sequester base cations”.
L469: let’s hope EW doesn’t lead to chrysotile formation! The implications of this for human health would swamp the CDR effect……
L473-480: see comment above. I don’t think this paragraph is helpful, given that clays are not expected to be in the reducible fraction.
L483: what are “the latter clays”? Not clear to me.
L491: Delete last sentence of this paragraph.
L503: Avoid referencing non-peer reviewed articles. Here, would be better to quote e.g., Clarkson et al. 2024
L511: I think erosion is accounted for, that is the reason for normalising to Ti?
Conclusions:
L545: Change to: “….may not immediately lead to inorganic CDR benefits.”
L548: change “proving” to “demonstrating”
L549: “reducible bases”???? assuming the bases are Ca, Mg, Na and K these are not reduced in terrestrial environments. Correct the phrasing.
L551: Well maybe but in practise basalt application rates will be at the lower end of your experiments, i.e. <30 t/ha. Please link this to reality.
Suppl info:
I assume all data will be supplied?
Fig. S6, S11 etc: Wherever standard error is reported, please give n
Citation: https://doi.org/10.5194/egusphere-2025-1667-RC1
Data sets
Corn mesocosm experiment 2021 UAntwerpen Arthur Vienne https://doi.org/10.5281/zenodo.15129984
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 597 | 119 | 22 | 738 | 38 | 19 | 34 |
- HTML: 597
- PDF: 119
- XML: 22
- Total: 738
- Supplement: 38
- BibTeX: 19
- EndNote: 34
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1