Pith. sign in

REVIEW 2 major objections 5 minor 6 references

Nanoscale Storage of Incompatible Elements at Olivine Grain Boundaries in Natural Basalts

T0 review · 2 major / 5 minor · reviewed 2026-08-01 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read The ~1 nm-wide crystalline boundaries between olivine crystals in natural basalts selectively store Ca, Al, P, Na, and Ti, forming a third trace-element reservoir alongside crystal interiors and residual melt.

desk verdict Qualitatively solid nanoscale characterization of olivine grain boundaries as trace-element reservoirs; the quantitative Ca isotherm comparison is more fragile than the paper suggests because the APT reconstruction was not plane-spacing calibrated. read the letter →

arxiv 2607.26329 v1 pith:ATCGZW5F submitted 2026-07-28 astro-ph.EP cond-mat.mtrl-sciphysics.geo-ph

Wenhao Zhao , Reid Cooper , Stephen Parman , Austin Akey , Greg Hirth This is my paper · ORCID
classification astro-ph.EPcond-mat.mtrl-sciphysics.geo-ph
keywords olivinegrainboundariesatomprobetomographytrace-elementsegregationincompatibleelementsbasaltinterfacialexcessnanoscalereservoir
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper sets out to show that grain boundaries in olivine from natural, undeformed basalts are not chemically inert seams but nanoscale reservoirs for incompatible elements. Combining transmission electron microscopy with atom probe tomography on boundaries from Piton de la Fournaise and Mauna Loa, the authors find that Ca, Al, P, Na, and Ti concentrate at a ~1 nm-wide crystalline interface while the major cations Mg, Fe, Ni, and Mn remain homogeneous. The measured calcium interfacial excess — about a third of a monolayer — sits near the published 1523 K equilibrium segregation isotherm for mantle olivine, which the paper reads as evidence that the boundaries formed near magmatic crystallization temperatures. The structural core is decoupled from a broader 6–15 nm chemical halo, a pattern the authors argue excludes a trapped melt film and instead records genuine grain-boundary segregation. If correct, olivine-rich rocks carry a third trace-element reservoir that should be included in mass balances used to interpret apparent partition coefficients and lunar magma-ocean evolution.

What carries the argument

The argument is carried by correlating transmission electron microscopy (TEM) with atom probe tomography (APT) on the same boundaries: TEM resolves the crystalline ~1 nm core; APT supplies three-dimensional, sub-nanometer chemical maps of which ions sit at the boundary. The quantitative anchor is the interfacial excess of calcium — the integrated Ca excess across one-dimensional compositional profiles, normalized by the olivine M-site density — compared against the published 1523 K equilibrium segregation isotherm for mantle olivine to yield a model-equivalent segregation temperature. Element-specific chemical widths come from a normalized cumulative-fraction regression of the profiles, quan

What would settle it

Analyze a population of boundaries — dozens of grains spanning misorientation angles and both localities — with atom-probe reconstructions calibrated against crystallographic plane spacings, and check whether the 6–15 nm halos persist and enrichment ratios stay well above 1. If the halo collapses toward the ~1 nm core under corrected reconstruction, or enrichment scatters across boundaries, the nanoscale-reservoir claim would be an artifact or a boundary-specific quirk rather than a general feature.

Watch

Extended reading notes

Core claim

The paper claims that olivine grain boundaries in undeformed basaltic olivine are crystalline (~1 nm wide by TEM) with no glassy film, and selectively enriched in Ca, Al, P, Na, and Ti (ratios ~1.4–6), while Mg, Fe, Ni, and Mn stay flat. The calcium interfacial excess is 2.94–2.97 atoms per square nanometer (0.327–0.330 monolayers), landing near the upper published 1523 K equilibrium segregation isotherm; the boundaries are read as forming during crystal growth at near-magmatic temperature, with synneusis plus hot residence an indistinguishable alternative. Chemical enrichment extends 6–15 nm beyond the ~1 nm core, decoupling structure from chemistry. The results are framed as case studies r

Load-bearing premise

Everything quantitative rests on just one grain boundary per rock, measured by a technique whose three-dimensional reconstruction was checked only against microscope images of the needle — if that reconstruction is distorted or those two boundaries are unusual, the enrichment ratios, the halo widths, and the calcium coverage that matches the high-temperature curve could all be misleading.

Editorial extensions

If this is right

  • Trace-element mass balances of olivine-rich aggregates should include a grain-boundary reservoir alongside crystal interiors and residual melt; apparent olivine–melt partition coefficients could be biased if boundary storage is attributed to the crystal.
  • The boundary's share of the inventory scales with grain-boundary area per unit volume, so the reservoir matters most in fine-grained, crystal-rich, or evolved cumulates where residual melts are enriched in incompatible elements.
  • The match between measured Ca coverage and the 1523 K equilibrium isotherm implies these boundaries formed near magmatic temperature during crystal growth; a synneusis origin followed by hot residence cannot be excluded.
  • Undeformed boundaries preserve broader chemical halos (locally ~10–15 nm) than published atom-probe data on deformed olivine, suggesting deformation narrows or redistributes segregation — a contrast the authors flag as suggestive, not established.
  • No rare-earth-element enrichment is claimed: several boundary-enriched mass windows coincide with nominal REE positions, but unresolved isobars and missing isotope envelopes leave the carriers unidentified.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • If interfacial segregation is a general property rather than an olivine quirk, the same three-reservoir logic should apply to other cumulus minerals and to phase boundaries in planetary rocks; the paper studies only olivine–olivine interfaces in two basalts.
  • The paper's own equations, combined with element-specific halo widths and grain-boundary area per volume, yield a directly computable correction to apparent partition coefficients — a generalization the authors describe conceptually but do not apply numerically to any existing dataset.
  • If boundary occupancy locks in at magmatic temperature, then the segregation state of a boundary may record cooling history; testing boundaries from samples with known, different emplacement rates could turn the isotherm match into a nanoscale record of thermal path.
  • A decisive discriminant between interfacial storage and trapped melt — which the single-boundary datasets cannot provide — would be comparing boundary element ratios with coexisting melt, as the authors note; extending that to the lunar case would quantify how much 'trapped liquid' in magma-ocean models could actually reside at grain boundaries.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

2 major / 5 minor

Summary. This paper presents a correlated EBSD/TEM/APT study of two olivine grain boundaries, one each from Piton de la Fournaise (Réunion) and Mauna Loa (Hawaii), both from undeformed basaltic aggregates. The central claims are that the boundaries are crystalline, ~1 nm wide, and selectively enriched in Ca, Al, P, Na, and Ti (enrichment ratios ~1.4–6) while Mg, Fe, Ni, and Mn remain homogeneous; that the Ca interfacial excess (2.94–2.97 atoms/nm²) lies near the upper 1523 K equilibrium segregation isotherm of Hiraga et al. (2004), implying near-magmatic boundary formation; and that the ~1 nm structural core is decoupled from a 6–15 nm chemical halo, so that grain boundaries constitute a distinct nanoscale reservoir for trace-element mass balances. The paper explicitly does not claim REE detection and excludes H from interpretation. It repeatedly acknowledges the single-boundary-per-locality limitation and the suggestive nature of the deformed-vs-undeformed width contrast.

Significance. If the central results hold, the paper provides the first direct nanoscale characterization of grain-boundary chemistry in undeformed natural basaltic olivine, demonstrating that incompatible-element storage at these interfaces is measurable and potentially relevant to trace-element budgets. The two independent specimens, the careful screening against REE misidentification, the exclusion of hydrogen artifacts, the provision of data and analysis code for the Ca-excess calculation, and the transparent acknowledgement of limitations are clear strengths. The qualitative enrichment pattern is well supported, but the quantitative isotherm comparison depends on unquantified reconstruction-scale and compositional-bias uncertainties that need to be addressed before the magmatic-temperature inference can be considered robust.

major comments (2)
  1. [§2.3.2 and §2.4, Fig. 7] The reported Ca interfacial excesses (2.94–2.97 atoms/nm²) and their placement relative to the Hiraga et al. (2004) isotherms depend directly on the absolute spatial scaling of the APT reconstruction. The paper states that the reconstruction was calibrated only against SEM needle images, with no independent plane-spacing calibration, and that systematic APT bias is excluded from the error bars. A 20–30% systematic error in the depth scale would shift Γ_Ca by a comparable factor, moving the points from near the 1523 K curve to below the 1473 K or above the 1523 K curves and weakening the inference of near-magmatic boundary formation. Because the abstract and conclusions elevate this agreement to a central finding, the authors should provide a sensitivity analysis varying reconstruction parameters (e.g., image compression factor, field factor) or use any available crystallographic features
  2. [§2.4, Fig. 7] The abscissa X_Ca^GM is the APT-measured grain-interior Ca M-site occupancy. Systematic APT compositional bias—from oxygen loss, Fe–Si isobaric overlaps, or preferential evaporation—is not included in the reported uncertainties for X_Ca^GM. Because the Hiraga isotherms are steep at the relevant X_Ca range, an unrecognized bias in this quantity would move the points horizontally and could alter the inferred model-equivalent segregation temperature. Please validate the absolute Ca concentration against an independent method (e.g., EPMA on the same grains) or at least quantify the sensitivity of the isotherm comparison to plausible compositional bias in X_Ca^GM.
minor comments (5)
  1. [§4.1, Fig. 6] The deformed-vs-undeformed width contrast, while explicitly hedged in the text, is presented graphically with a visual force that may overstate the robustness of the difference. Consider adding a sentence to the Figure 6 caption reiterating that the comparison is hypothesis-generating and limited by cross-laboratory, cross-instrument, and single-boundary-per-locality issues.
  2. [Abstract] The abstract states the isotherm agreement without explicitly noting that it is based on one boundary per locality. A short phrase such as “for the two analyzed boundaries” would help calibrate reader expectations.
  3. [§3.1] The Réunion EBSD metadata do not retain the rotation-axis/plane information, so a full twin-law exclusion is not possible. This is clearly stated in §3.1; a brief reminder in the abstract or conclusions would prevent overinterpretation of the boundary character.
  4. [Fig. 4] The cumulative-fraction regression method for chemical thickness (fits y1, y2, y3 and their intersections) is described only in the figure caption. Moving a one-sentence description of the method into §2.4 would improve readability and reproducibility.
  5. [§2.4] The choice of |d| ≥ 8 nm for the matrix baseline and ±2 nm window for the boundary mean are reasonable, and the sensitivity of Γ_Ca to these choices is reported. Good practice; no change needed.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: the central Ca-interfacial-excess result is measured, not fitted, and benchmarked against the independent Hiraga et al. (2004) experimental isotherms; self-citations are hedged and non-load-bearing.

full rationale

The derivation chain is self-contained against external benchmarks. ΓCa is obtained by direct signed integration of the APT 1-D profile (XCa = Ca/(Mg+Fe+Ca+Mn+Ni+Co), matrix baseline at |d|≥8 nm, N_M = 27.9 nm−3), i.e., it is a measured quantity, not a parameter fitted to any target. The comparison that carries the paper's central inference — ΓCa ≈ 2.94–2.97 atoms/nm² near the upper 1523 K isotherm — uses published experimental curves from Hiraga et al. (2004), whose authors are distinct from the present authors and whose parameter sets were determined from annealed synthetic aggregates, not from this work; the agreement is therefore a genuine external benchmark rather than a self-consistency check. The paper explicitly declines to treat the two nearly identical points as independent validation ('we do not treat the nearly identical two study points as independent validation') and repeatedly hedges the single-boundary-per-locality basis. The deformed-olivine comparison (Fig. 6) cites Cukjati et al. (2019), which shares authors with the present paper, but the paper twice labels that contrast 'suggestive rather than established' and does not use it to support the temperature or reservoir claims; it is therefore a non-load-bearing self-citation. The Figure 8 elastic curve is 'anchored to Ca' and 'drawn as a guide,' so the statement that Ca lies on it is partly by construction; however, the qualitative divalent size-misfit statement rests on the measured near-unity ratios of Ni, Mg, Fe, Mn and the large measured Ca ratio, and the aliovalent departures are explicitly left unconstrained. The lunar section self-limits: 'We deliberately stop short of a numerical lunar mass balance' and 'K_GB/melt_i is not constrained by our data.' The skeptic's calibration concern (reconstruction scaled only against SEM needle images, with no independent plane-spacing calibration) is a measurement-accuracy and error-budget issue that could shift ΓCa, but it does not make the isotherm comparison equivalent to its inputs — the Hiraga et al. (2004) isotherms are independent of the APT reconstruction. Accordingly, no step reduces a prediction to its own inputs, and the circularity score is low.

Assumptions & free parameters 0 free parameters · 4 assumptions · 0 invented entities

No new particles, phases, or forces are introduced. The candidate REE-related mass windows are explicitly not claimed as elemental detections. The only 'model' inputs are standard olivine density and published segregation parameters. The main unstated inputs are representativeness and spatial-fidelity assumptions of APT.

assumptions (4)
  • domain assumption APT reconstructions of silicate specimens preserve the true relative spatial distribution of detected ions.
    Central to mapping grain-boundary enrichment; reconstruction calibrated only against SEM needle images, without independent plane-spacing calibration (Section 2.3.2, Figure S10).
  • domain assumption The single analyzed grain boundary per locality is representative of olivine boundaries in these basalts.
    Paper states conclusions rest on one boundary per locality and describes the results as case studies (Section 4.1, Conclusions).
  • domain assumption The observed enrichment is not dominated by a continuous trapped melt film.
    Supported by near-constant Si and Mg+Fe across the boundary, but the authors acknowledge that a former melt film or melt-fed halo cannot be excluded (Section 4.1).
  • domain assumption The Hiraga et al. (2004) equilibrium segregation isotherms and the olivine M-site density values apply to these natural boundaries.
    Used to convert Ca excess to equivalent monolayers and to compare with published isotherms (Section 2.4, Figure 7).

how reviews work

0 comments
Cite this review

Pith. "Pith review of Nanoscale Storage of Incompatible Elements at Olivine Grain Boundaries in Natural Basalts." pith.science (2026). https://pith.science/paper/ATCGZW5F

@misc{pith2026260726329,
  author       = {Pith},
  title        = {Pith review of: Nanoscale Storage of Incompatible Elements at Olivine Grain Boundaries in Natural Basalts},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/ATCGZW5F}},
  note         = {Machine review of arXiv:2607.26329}
}
read the original abstract

Grain boundaries are pervasive in polycrystalline olivine, yet their structure and trace-element chemistry remain poorly constrained. We characterize boundaries in undeformed olivine aggregates from Piton de la Fournaise (La Reunion) and Mauna Loa (Hawaii) by correlating electron backscatter diffraction, transmission electron microscopy, and atom probe tomography. The boundaries are crystalline, with a structural width of about 1 nm and no continuous glassy film resolved. Ca, Al, P, Na, and Ti are selectively enriched, whereas Mg, Fe, Ni, and Mn remain homogeneous. Several enriched mass windows coincide with nominal rare-earth-element ion positions, but unresolved isobars and the lack of a diagnostic isotope envelope preclude secure assignments. Divalent-cation segregation follows an ionic-size-misfit trend, while aliovalent departures require unconstrained contributions. The Ca interfacial excess is 2.94-2.97 atoms per square nanometer, equivalent to 0.327-0.330 monolayers, and lies near the upper published 1523 K equilibrium segregation isotherm of Hiraga et al. (2004), within the range of basaltic olivine crystallization temperatures. This supports boundary formation during crystal growth, although synneusis followed by high-temperature residence could produce a similar signature. Ca chemical widths are about 6-9 nm, and those of the most strongly enriched elements reach about 10-15 nm, demonstrating decoupling between the narrow structural core and broader chemical segregation. Compared with published atom-probe data on deformed olivine, these boundaries preserve broader halos, although analytical differences make the contrast tentative. Olivine grain boundaries therefore constitute a distinct nanoscale reservoir that should be included alongside crystal interiors and residual melt in trace-element mass balances of olivine-rich aggregates.

Figures

Figures reproduced from arXiv: 2607.26329 by the authors.

Figure 1
Figure 1. (d) shows the clot, which consists of three euhedral olivine crystals, reflecting a largely unaltered crystallization history. Thermal cracks extend across the crystals, likely from cooling-induced fracturing, and the crystals host abundant composite inclusions—glass–chromite assemblages, gas bubbles, and discrete chromite. The analyzed grain boundary is outlined with a green rectangle under reflected light in [PIT… view at source ↗
Figure 2
Figure 2. TEM of the Réunion olivine grain boundary. (a) Low-magnification bright-field image showing the boundary (GB) separating two olivine grains (Ol). (b) Higher-magnification bright-field view of the boundary. (c) High-resolution TEM (HR-TEM) image across the boundary; inset: selected-area electron diffraction (SAED) pattern of the adjacent grain, and the orange box marks the analyzed region. (d) Intensity profile acros… view at source ↗
Figure 3
Figure 3. Atom probe tomography (APT) ion maps of the Réunion olivine needle (an ∼500 nm section around the grain boundary, cropped from the full ∼1.16 µm reconstruction), viewed face-on to the grain boundary so that the boundary plane is seen edge-on. The top-left panel shows all ranged species together (one color per ion type). The remaining top-row maps (Mg+ , FeO+ , Si2+, O+ , Cr+ , Ni+ , Co+ ) are homogeneous across the … view at source ↗
Figures from the paper (8 more)
Figure 4
Figure 4. Figure 4: Calcium (Ca) interfacial enrichment at the Réunion olivine grain boundary and demonstration of the grain-boundary chemical-thickness measurement (APT). (a) Ca iso-concentration surface at 0.7 at.%, shown as two orthogonal projections of the reconstruction; the Ca enric…
Figure 5
Figure 5. Figure 5: One-dimensional APT profile across the Réunion boundary, regenerated directly from reconstruc￾tion R13_05409-v01, Cube ID 6, Y-axis (63.4 nm at 0.10 nm spacing), along the green arrow in [PITH_FULL_IMAGE:figures/full_fig_p015_5.png]
Figure 6
Figure 6. Figure 6: GB chemical widths for selected elements and the nominal (SmO)2+ mass channel in Réunion and Hawaiian olivine (this study), compared with deformed wehrlite (Cukjati et al., 2019) and deformed mylonite (Tacchetto et al., 2021). Points are not connected because the horiz…
Figure 7
Figure 7. Figure 7: Grain-boundary Ca coverage versus grain-interior Ca M-site occupancy. Curves are calculated from equations 1–2 and the three published parameter sets (1373, 1473 and 1523 K) of Hiraga et al. (2004); the previously plotted 1300 and 1573 K extrapolations have been remove…
Figure 8
Figure 8. Figure 8: Grain-boundary enrichment ratio (CGB/Cmatrix, measured from the 1-D profiles as the mean of the two samples; error bars span Réunion and Mauna Loa) versus six-fold ionic radius. Ca has sample ratios of 3.71 and 5.98 and the plotted two-sample mean is 4.84. The solid cu…
Figure 9
Figure 9. Figure 9: Schematic of an olivine–olivine grain boundary between two misoriented olivine grains (honeycomb lattices), illustrating the decoupling of its structural and chemical dimensions. The crystallographic core resolved by TEM is extremely narrow (∼1 nm; central hatched band…
Figure 10
Figure 10. Figure 10: This application is also motivated by recent petrogenetic studies of Chang’e-5 and Chang’e-6 lunar basalts. In the Chang’e-5 Nature study, the modeled mantle source was described as an 86% crystallized LMO cumulate with 2% trapped instantaneous residual liquid (TIRL),…
Figure 10
Figure 10. Figure 10: Highly schematic summary of the hypothesis (to be tested) that the grain- plus phase-boundary (GB + PB) network is a distributed incompatible-element reservoir in the lunar magma-ocean (LMO) cumulate pile. (a) Stratigraphic column of LMO cumulates, from the most mafic…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

6 extracted references · 3 canonical work pages

  1. [1]

    & Nespolo, M

    Azevedo, S. & Nespolo, M. (2017) Twinning in olivine group revisited. European Journal of Mineralogy 29, 213–226.https://doi.org/10.1127/ejm/2017/0029-2598 Bamber, E. C. et al. (2025) 3D quantification of nanolites using X-ray ptychography reveals syn-eruptive nanocrystallisation impacts magma rheology. Nature Communications 16,

    work page doi:10.1127/ejm/2017/0029-2598 2017

  2. [2]

    (b) Higher-magnification bright-field view

    (a) Low-magnification bright-field image of the boundary (GB) separating two olivine grains (Ol). (b) Higher-magnification bright-field view. (c) HR-TEM image across the boundary; inset: SAED pattern of the adjacent grain, with the orange box marking the analyzed region. (d) Intensity profile across the boundary (boundary region shaded), giving a structur...

  3. [89]

    & Li, Q.-L

    https://doi.org/10.1007/ s00445-025-01874-5 Zhao, W., Wang, B.-W., Gu, L.-X., Tang, X., Liu, Y ., Awais, M., Liu, X.-C., Li, X.-H., Hu, R. & Li, Q.-L. (2023) Nanoscale evidence of zircon–coffinite–xenotime solid solution observed by atom probe tomography. Chemical Geology 638, 121697.https://doi.org/10.1016/j.chemgeo.2023.121697 Zhou, Q. et al. (2025) Ult...

    arXiv 2023

  4. [136]

    Springer, Berlin Heidelberg.https://doi.org/10.1007/978-3-642-12505-8 Mancini, L. et al. (2014) Composition of wide bandgap semiconductor materials and nanostructures measured by atom probe tomography and its dependence on the surface electric field. The Journal of Physical Chemistry C 118, 24136–24151.https://doi.org/10.1021/jp5071264 Marquardt, K., Rohr...

    arXiv 2014

  5. [160]

    & Nativel, P

    Springer, New York.https://doi.org/10.1007/978-1-4614-3436-8 Gillot, P.-Y . & Nativel, P. (1989) Eruptive history of the Piton de la Fournaise volcano, Réunion Island, Indian Ocean. Journal of V olcanology and Geothermal Research 36, 53–65.https://doi.org/10.1016/ 0377-0273(89)90005-X Goldschmidt, V . M. (1937) The principles of distribution of chemical e...

    work page doi:10.1007/978-1-4614-3436-8 1989

  6. [7083]

    1038/s41467-025-62444-z Blundy, J

    https://doi.org/10. 1038/s41467-025-62444-z Blundy, J. & Wood, B. (1994) Prediction of crystal–melt partition coefficients from elastic moduli. Nature 372, 452–454.https://doi.org/10.1038/372452a0 Blundy, J. & Wood, B. (2003) Partitioning of trace elements between crystals and melts. Earth and Planetary Science Letters 210, 383–397.https://doi.org/10.1016...

    work page doi:10.1038/372452a0 1994

Pith tools

Reviewed August 1, 2026 · model on record in the stance chip above.