Pith. sign in

REVIEW 3 major objections 6 minor 1 cited by

Interstitial Solute Segregation at Triple Junctions: Implications for the Hydrogen Storage Properties of Nanomaterials

T0 review · 3 major / 6 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read Triple junctions, not just grain boundaries, act as strong hydrogen traps that can triple hydrogen storage in nanocrystalline palladium.

desk verdict Solid extension of the spectral framework to interstitial TJ segregation, but the storage tripling rests on an unvalidated classifier and fits without error bars. read the letter →

arxiv 2411.18537 v1 pith:E3CHR5WG submitted 2024-11-27 cond-mat.mtrl-sci

Nutth Tuchinda , Malik Wagih , Christopher A. Schuh This is my paper · ORCID
classification cond-mat.mtrl-sci
keywords grainboundarysegregationtriplejunctionsinterstitialsolutehydrogenstoragenanocrystallinepalladiumenergyspectrathermodynamicsatomisticsimulation
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

The paper develops a spectral thermodynamic framework for interstitial solutes in nanocrystalline metals, demonstrated on palladium with dissolved hydrogen, that separates hydrogen segregation into three site populations: bulk octahedral and tetrahedral sites, grain-boundary sites, and triple-junction sites. Its central claim is that triple-junction sites are about 3 kJ/mol more attractive to hydrogen than grain-boundary sites, and that this extra trapping, together with the large intergranular site fractions at fine grain sizes, can triple the hydrogen density of nanocrystalline palladium relative to microcrystalline palladium. The paper also claims that the individual grain-boundary and triple-junction subspectra shift toward weaker segregation as grain size decreases, because lattice dilation strain makes the intergranular environments less hospitable, and that this intrinsic shift counteracts the extrinsic growth of triple-junction content so that the total spectrum appears nearly grain-size independent. A sympathetic reader would care because it gives a quantitative, mechanism-based route to engineer interfaces for hydrogen storage and for interstitial solute behavior in nanocrystalline alloys generally.

What carries the argument

The carrying object is the size-scaling spectral isotherm: a solute conservation equation that partitions total hydrogen among bulk octahedral and tetrahedral sites and two defect populations, each described by a weighted double-Gaussian segregation-energy distribution with fitted means, widths, and weights, and by site fractions that scale with grain size through a polynomial law. The defect site fractions are computed directly from self-similar atomistic polycrystals, and the segregation-energy distributions are obtained by molecular statics relaxation of hydrogen at every identified interstitial site. The mechanism that produces the intrinsic size dependence is lattice dilation strain: finer grains dilate, bulk interstitial cavities grow and become slightly more stable, and grain-boundary and triple-junction interstitial sites grow and become slightly less attractive, shifting the defect subspectra toward weaker segregation by about 1 kJ/mol. This dilational shift is what counteracts the extrinsic effect of rising triple-junction fractions, and it is the feature that distinguishes interstitial segregation from substitutional segregation in this work.

What would settle it

Recompute the grain-boundary and triple-junction subspectra and the hydrogen density prediction while varying the site-classification cutoff across a physically plausible range (for example 2.0 to 3.4 Å) at the smallest grain size; if the ~1 kJ/mol size shifts or the threefold storage enhancement depend strongly on this choice, the central mechanism is not robust. Alternatively, measure equilibrium hydrogen solubility of nanocrystalline palladium with controlled grain sizes between about 6 and 20 nm at 700 K and low hydrogen content; if the stored hydrogen density does not rise strongly toward three times the microcrystalline value as grain size decreases, the predicted triple-junction-driven enhancement is contradicted.

Watch

Extended reading notes

Core claim

The central discovery is a size-scaling segregation isotherm for interstitial alloys in which the grain-boundary and triple-junction contributions are resolved separately, applied to model Pd(H) polycrystals with grain sizes from 5.8 to 10.4 nm. Each defect population contributes a bimodal Gaussian segregation-energy spectrum, one mode for semi-octahedral and one for semi-tetrahedral interstitial sites, and the triple-junction spectra are consistently shifted roughly 3 kJ/mol more negative than the grain-boundary spectra at every grain size. The paper shows that the fraction of triple-junction interstitial sites rises steeply as grain size falls, so the extra binding at junctions should dominate the finest microstructures. In addition, unlike substitutional systems studied earlier, the individual spectra here are intrinsically size dependent: at the smallest grain sizes the grain-boundary and triple-junction sites drift about 1 kJ/mol toward weaker segregation, an effect attributed to grain-size-dependent lattice dilation and enlarged interstitial site volumes that destabilize solute occupation at defects while slightly stabilizing bulk sites. These two contributions oppose each other, leaving the overall intergranular spectrum nearly size invariant. Feeding the measured spectra through the dilute-limit isotherm at 700 K and a bulk octahedral occupancy near 10 at.% yields a total hydrogen density in the finest nanocrystalline structures that is more than three times the microcrystalline value, with triple junctions providing the dominant defect contribution.

Load-bearing premise

The load-bearing assumption is that every interstitial site can be assigned unambiguously to either the grain-boundary or triple-junction population using a fixed 2.7 Å distance cutoff, and that this same cutoff remains valid at every grain size studied; at the finest grain sizes the two assignment zones may overlap, which would distort the extracted subspectra and the predicted hydrogen density.

Editorial extensions

If this is right

  • Triple junctions should be treated as a distinct, stronger class of hydrogen traps in nanocrystalline Pd, with local hydrogen concentrations several atomic percent higher than at grain boundaries at moderate total solute content.
  • At grain sizes below about 10 nm the intergranular network, especially triple junctions, controls total hydrogen uptake, while above about 10 nm the triple-junction contribution to total stored hydrogen becomes negligible.
  • The apparent grain-size independence of the total segregation spectrum in Pd(H) is a cancellation between rising triple-junction fraction and weakening intrinsic binding, so coarse-grain segregation parameters cannot be extrapolated to nanocrystalline grain sizes without resolving both effects.
  • Interstitial segregation spectra in general cannot be assumed size-independent on the basis of substitutional behavior; computations should use sufficiently large grains (above about 8 nm) to avoid intrinsic size artifacts when extracting alloy design parameters.

Reading between the lines

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

  • If the 3 kJ/mol triple-junction advantage is real, local triple-junction hydrogen enrichment should matter for hydrogen embrittlement even at coarse grain sizes, where the triple-junction fraction is tiny but the local concentration at those sites rises sharply; this follows from the paper's Figure 7 but is not a claim the authors develop.
  • The cancellation between intrinsic and extrinsic size effects suggests a design handle: changing the host lattice's stiffness or adding solutes that alter lattice dilation should shift the grain size at which hydrogen capacity peaks, and could turn the triple-junction contribution into a monotonic rather than compensated size effect.
  • Because the same spectral framework is generic to interstitial solutes, the analogous decomposition for B, C, N, or O in FCC nanocrystals is a natural extension; the sign and size of the GB-TJ contrast may differ from Pd(H) because those solutes have different site preferences and strain fields, so the near-cancellation seen here may be particular to Pd(H).
  • An experimental test could compare low-concentration hydrogen solubility of well-annealed nanocrystalline and microcrystalline Pd at fixed temperature; a measured storage enhancement approaching the predicted factor of three would corroborate the intergranular trapping mechanism, though the paper's dilute-limit assumptions exclude hydride formation and solute-solute interactions.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 6 minor

Summary. The paper extends the spectral segregation framework, previously developed for substitutional solutes, to interstitial solutes in nanocrystalline Pd(H). The authors construct self-similar polycrystals, identify grain-boundary (GB) and triple-junction (TJ) interstitial sites, compute segregation-energy spectra via molecular statics, and fit bimodal Gaussian subspectra for each defect type as a function of grain size. They report that TJ sites are about 3 kJ/mol more attractive for hydrogen than GB sites, that the individual GB and TJ subspectra shift by about 1 kJ/mol with grain size due to lattice dilation, and that these intrinsic shifts counteract the extrinsic increase in TJ fraction at fine grain sizes. Using a dilute-limit isotherm with fitted spectral parameters, they predict that the total hydrogen density in nanocrystalline Pd can triple relative to microcrystalline Pd at the finest grain sizes.

Significance. If the quantitative claims hold, this is a meaningful advance: it is the first systematic treatment of interstitial segregation to triple junctions, it explains why the total intergranular segregation spectrum can be nearly grain-size-independent even when the GB/TJ composition changes, and it identifies lattice strain as an intrinsic size effect that counteracts the extrinsic TJ-fraction effect. The framework is general for interstitial alloys and the tabulated parameters (Table II, Table I) make the calculations reproducible. The hydrogen-storage implication is a falsifiable but strongly approximate prediction. The main weakness is that the central quantitative claims rest on a fixed site-classification cutoff and on fitted parameters without reported uncertainties, and the headline tripling prediction is made with a dilute-limit model at non-dilute conditions.

major comments (3)
  1. [Section III, Table II, Eq. (7), Eq. (11)] The TJ/GB classification uses a fixed 2.7 Å cutoff around Pd atoms identified as triple-junction atoms, and this cutoff is not revalidated at each grain size. With the fitted intergranular width α = 1.72 nm, the defect network at d = 5.8 nm is dense, and the reported GB–TJ energetic contrast is only ~2–3 kJ/mol while the claimed intrinsic size shifts are ~1 kJ/mol. A small systematic misassignment of sites at the finest grain sizes could therefore create or mask exactly the reported contrast and size dependence. Because the same classification sets f_GB and f_TJ in Eq. (7) and the density estimate in Eq. (11), the paper's central quantitative claims are sensitive to this choice. Please include a sensitivity analysis varying the 2.7 Å cutoff (e.g., 2.3, 2.5, 2.9 Å) and show how the subspectral parameters in Table II and the hydrogen density in Fig. 8 change.
  2. [Table II] No uncertainties are reported for the fitted spectral parameters. The claimed size shifts (~1 kJ/mol) and GB–TJ contrasts (~3 kJ/mol) are roughly an order of magnitude smaller than the fitted distribution widths (σ ≈ 4–6 kJ/mol), and double-Gaussian fitting is known to have significant parameter degeneracy. To establish that the trends in Table II are statistically significant, the authors should provide confidence intervals, for example via bootstrapping of the site energies or repeated fits to subsamples, or at least an explicit statement of the fitting residuals.
  3. [Abstract, Fig. 8, Section V] The hydrogen-density tripling is computed using the dilute-limit isotherm (Eq. (2) integrated in Eq. (6)) at X_oct = 10 at.% and T = 700 K, which is not in the dilute regime. The authors correctly acknowledge that solute–solute interactions, hydride formation, and boundary structural transitions are omitted, but these effects are precisely what would control hydrogen uptake at such concentrations. As written, the abstract's claim that 'hydrogen density can triple' is a statement about the dilute-limit model outside its regime of validity. Please either restrict the quantitative prediction to concentrations where the dilute-limit isotherm applies, or extend the isotherm to include repulsive interactions at high occupancy, and re-frame the abstract accordingly.
minor comments (6)
  1. [Section V] Equations (7) and (8) are referenced inconsistently: the isotherm calculations in Fig. 7 and the accompanying text should refer to Eq. (7) (the solute-conservation constraint), not Eq. (8) (the polynomial fit for defect fractions). The same error appears in the Fig. 7 caption.
  2. [Section IV and V] The text cites 'Table I' when referring to the segregation-energy subspectral parameters; the correct reference is Table II. This occurs in Section IV ('cf. segregation subspectral parameters in Table I') and Section V ('in Fig. 5 and Table I').
  3. [Section III] The phrase '~106 sites at d = 10.4 nm' should read '~10^6 sites'.
  4. [Section IV] The description of the bulk reference sites should specify that only bulk octahedral interstitial sites were used for the reference energy E_oct_ref^relax, since the text says only 'interstitial sites around 25 randomly selected Pd sites' without stating the site type.
  5. [Fig. 8 and concluding paragraph] The quantity plotted in Fig. 8 is a hydrogen density (kg H per m^3), but the concluding paragraph describes it as 'bulk solubility.' Please rephrase to 'hydrogen density' or 'hydrogen uptake capacity' to avoid confusion with thermodynamic solubility.
  6. [Section III] In Table I, the polynomial fit coefficients are listed without any goodness-of-fit metrics or uncertainties. A brief statement of the fitting quality, or a plot showing residuals, would help the reader assess the reliability of f_GB and f_TJ used in the isotherm.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the GB/TJ subspectra and hydrogen-density results are computed from atomistic energies and measured site fractions, not imposed by the model inputs.

full rationale

The derivation chain is self-contained. The authors generate self-similar polycrystals, identify interstitial sites by Voronoi analysis, compute segregation energies directly from relaxed atomistic configurations via Eq. (10), fit the bimodal Gaussian form of Eq. (1) to those computed energies (Table II), measure site fractions as a function of grain size (Fig. 2, Table I), and then evaluate the conservation isotherm of Eqs. (2)-(7) and hydrogen density via Eq. (11). The central new claims—the ~3 kJ/mol GB vs TJ energetic contrast and the ~1 kJ/mol intrinsic size shifts of the subspectra—are computed energetic quantities, not outputs imposed by the geometric cutoffs. The 2.7 Å interstitial-site classification radius and the 5.2 Å triple-junction identification cutoff are imported from the authors' prior work and are modeling choices; they affect which sites are labeled GB or TJ, but the segregation energies of those sites are calculated, not prescribed, so the contrast is not established by definition. The hydrogen-density tripling in Fig. 8 is a forward consequence of the fitted spectral parameters and site fractions, which is extrapolation from a model rather than a circular validation; no measured hydrogen-density data are used as an input. No uniqueness theorem is invoked, and the self-citations supply methods and thermodynamic models (the spectral isotherm, polycrystal construction, TJ identification) rather than a circular justification of the new GB/TJ result. The paper explicitly acknowledges limitations such as the dilute-limit isotherm, neglect of solute-solute interactions, and hydride formation, which are scope conditions rather than circular steps. Sensitivity of the classification at d = 5.8 nm is a legitimate robustness concern, but it does not reduce any claimed result to an input by construction.

Assumptions & free parameters 6 free parameters · 6 assumptions · 0 invented entities

The central prediction depends on fitted spectral distributions, fitted site-fraction scaling coefficients, and several stated domain assumptions. The paper itself acknowledges the most important simplifications: no hydride formation, no solute-solute interactions, and a dilute-limit isotherm applied at finite concentration. The classification cutoff is the least externally constrained input.

free parameters (6)
  • Spectral distribution parameters for GB and TJ (Table II) = w_tet 0.15-0.22, mu_tet 1.39-6.3 kJ/mol, sigma_tet 3.91-5.82 kJ/mol, w_oct 0.78-0.85, mu_oct -10.4 to -5.36 kJ/mol…
    Fit to segregation energy histograms via Eq. (1); used in the isotherm Eq. (7) and the hydrogen density estimate Eq. (11).
  • Site-fraction polynomial coefficients (Table I) = GB: A5=1.96, A2=-3.64, A3=1.68, alpha=1.72 nm; TJ: A5=-1.94, A2=2.94, A3=0
    Fit to computed interstitial site fractions as a function of grain size using Eq. (8); the fitted alpha is much larger than substitutional values.
  • TJ classification cutoff radius = 2.7 Å
    Chosen by hand from the prior IG-site definition [20]; used to separate TJ from GB sites, directly affecting the subspectra and all downstream predictions.
  • IG site definition radius = 2.7 Å
    Interstitial sites within 2.7 Å of an intergranular solvent atom are counted as intergranular; this follows [20] and is not revalidated at every grain size.
  • Site deduplication distance = 1 Å
    H atoms within 1 Å of one another are deleted to prevent duplicate site counts; this affects the site fractions and spectral normalization.
  • Bulk octahedral concentration for Fig. 8 = 10 at.%
    A chosen operating point used with T = 700 K in Eq. (11); the tripling prediction is sensitive to this choice.
assumptions (6)
  • domain assumption Bimodal Gaussian spectral form (Eq. 1) from [20] describes interstitial segregation energies at both GBs and TJs.
    All subspectra are fit to this form; if the true distributions are not a sum of two Gaussians, the fitted parameters and the isotherm outputs are biased.
  • domain assumption Dilute-limit Langmuir isotherm (Eq. 2) remains valid at the concentrations used, including up to 10 at.% bulk octahedral hydrogen and locally much higher TJ concentrations.
    The model ignores solute-solute interactions and hydride formation; the authors acknowledge this in Section V, but the abstract presents the tripling without this caveat.
  • domain assumption The EAM Pd-H interatomic potential [39] accurately captures energetics in strained nanocrystalline grains and at triple junctions.
    All segregation energies and the hydrogen density prediction depend on this potential; no DFT or experimental cross-check is presented.
  • domain assumption The polycrystalline structures are self-similar and representative across grain sizes 5.7-10.4 nm.
    The study uses a limited set of seed orientations and a modest grain size range; extrapolation to other grain sizes relies on the fitted polynomial Eq. (8).
  • domain assumption Vibrational entropy is neglected; molecular statics energies at 0 K are used in the 700 K isotherm.
    The isotherm includes configurational entropy but not vibrational contributions, which can shift segregation preferences at elevated temperature.
  • ad hoc to paper The fixed 2.7 Å TJ classification cutoff cleanly separates TJ from GB sites at all grain sizes.
    This cutoff is introduced in the present work and not validated by a convergence study; at the finest grain sizes it could create artificial size dependence in the subspectra.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Interstitial Solute Segregation at Triple Junctions: Implications for the Hydrogen Storage Properties of Nanomaterials." pith.science (2026). https://pith.science/paper/E3CHR5WG

@misc{pith2026241118537,
  author       = {Pith},
  title        = {Pith review of: Interstitial Solute Segregation at Triple Junctions: Implications for the Hydrogen Storage Properties of Nanomaterials},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/E3CHR5WG}},
  note         = {Machine review of arXiv:2411.18537}
}
read the original abstract

At very fine grain sizes, grain boundary segregation can deviate from conventional behavior due to triple junction effects. While this issue has been addressed in prior work for substitutional alloys, here we develop a framework that accounts for interstitial sites in the grains, grain boundaries, and triple junctions of model Pd(H) polycrystals. This approach allows computation of interstitial segregation spectra separately at both defect types, which permits an understanding of segregation at all grain sizes via a size-scaling spectral isotherm. The size dependencies of dilute Pd(H) are found to be influenced not only by the triple junction content, but also by grain size-dependent lattice strains; the latter effect is evidenced by size dependencies of individual grain boundary and junction subspectra. The framework proposed here is applicable to interstitial alloys in general, and may serve as a basis for interfacial engineering in interstitial nanocrystalline alloys. As an example, we show using the dilute limit isotherm that hydrogen density can triple in nanocrystalline vis-\`a-vis microcrystalline Pd due to hydrogen adsorption at intergranular defect sites.

Figures

Figures reproduced from arXiv: 2411.18537 by the authors.

Figure 3
Figure 3. FIG. 3 [PITH_FULL_IMAGE:figures/full_fig_p008_3.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Grain boundary interstitial segregation in substitutional binary alloys

    cond-mat.mtrl-sci 2025-01 conditional novelty 6.0 of 10

    Nickel atoms prefer to segregate into interstitial cavities within kite-like grain boundary structures in aluminum, and a new Voronoi-based method with machine learning can identify and predict these sites.

Reference graph

Works this paper leans on

4 extracted references · 4 canonical work pages · cited by 1 Pith paper

  1. [1]

    Palumbo, S

    G. Palumbo, S. J. Thorpe, and K. T. Aust, On the contribution of triple junctions to the structure and properties of nanocrystalline materials, Scr. Metall. Mater. 24, 1347 (1990). [2] K. Ishida, Effect of grain size on grain boundary segregation, J. Alloys Compd. 235, 244 (1996). [3] N. Tuchinda and C. A. Schuh, Triple Junction Excess Energy in Polycryst...

    work page 1990

  2. [23]

    K. Ito, Y. Tanaka, K. Tsutsui, and H. Sawada, Analysis of grain-boundary segregation of hydrogen in bcc-Fe polycrystals via a nano-polycrystalline grain-boundary model, Comput. Mater. Sci. 225, 112196 (2023). [24] M. A. Tschopp, K. N. Solanki, F. Gao, X. Sun, M. A. Khaleel, and M. F. Horstemeyer, Probing grain boundary sink strength at the nanoscale: Ener...

    work page 2023

  3. [46]

    bond-focused

    A. P. Thompson, H. M. Aktulga, R. Berger, D. S. Bolintineanu, W. M. Brown, P. S. Crozier, P. J. in ’t Veld, A. Kohlmeyer, S. G. Moore, T. D. Nguyen, et al., LAMMPS - a flexible simulation tool for particle-based materials modeling at the atomic, meso, and continuum scales, Comput. Phys. Commun. 271, 108171 (2022). [47] P. Hirel, Atomsk: a tool for manipul...

    work page 2022

  4. [70]

    Strain Functionals: A Complete and Symmetry-adapted Set of Descriptors to Characterize Atomistic Configurations

    N. Tuchinda and C. A. Schuh, Computed entropy spectra for grain boundary segregation in polycrystals, Npj Comput. Mater. 10, 72 (2024). [71] E. M. Kober, J. P. Tavenner, C. M. Adams, and N. Mathew, Strain Functionals: A Complete and Symmetry-Adapted Set of Descriptors to Characterize Atomistic Configurations, arXiv:2402.04191. [72] C. B. Owens, N. Mathew,...

    work page Pith review arXiv 2024

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.