Pith. sign in

REVIEW 3 major objections 5 minor 36 references

Prediction of Mechanical Properties and Thermodynamic Stability of Ti-N system using MTP Interatomic Potential

T0 review · 3 major / 5 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read This paper develops a moment tensor potential that predicts formation energies, elastic constants, and thermodynamic stability across the entire Ti-N composition range, including previously uncharacterized intermediate structures.

desk verdict A fitted MTP for Ti-N that is accurate on known phases, but the claim that all intermediate stoichiometries are thermodynamically stable is an unvalidated extrapolation from the potential. read the letter →

arxiv 2507.18873 v1 pith:CRKTPKJR submitted 2025-07-25 cond-mat.mtrl-sci physics.comp-ph

Pradeep Kumar Rana , Atharva Vyawahare , Rohit Batra , Satyesh Kumar Yadav This is my paper · ORCID
classification cond-mat.mtrl-sciphysics.comp-ph
keywords Ti-Nsystemmomenttensorpotentialinteratomicformationenergyconvexhullelasticconstantsmachinelearningthermodynamicstability
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

Titanium nitride is not a single compound but a family: ordered phases such as Ti2N, Ti3N2, Ti4N3, and Ti6N5, together with nitrogen solid solutions in titanium, span a composition range from pure Ti to TiN. This paper tries to establish that one machine-learned moment tensor potential can capture the energy, mechanical response, and thermodynamic stability of the whole family, not just the handful of known compounds. The authors build a training set by exploiting the structural relationship among the phases — most are rock-salt TiN with ordered nitrogen vacancies, while the solid solutions are nitrogen in octahedral voids of hcp Ti. On held-out structures the potential reaches formation-energy errors of a few meV per atom and elastic-constant errors with a median near 5%. Using this potential, they predict that intermediate structures with N/Ti ratios from 0 to 1 can be thermodynamically stable, within about 10 meV/atom of the 0 K convex hull.

What carries the argument

The central object is the moment tensor potential, a machine-learned interatomic potential that represents each atom's local environment by moment tensors built from radial functions and tensor products of neighbor position vectors, with a cut-off radius of 7 Å and 2621 moment-tensor descriptors. The specific variant selected, MTP0.75-0.25-0.25, weights energy, force, and stress terms in the loss function as 0.75, 0.25, and 0.25; the paper finds this balanced weighting is what makes energy predictions accurate for structures outside the training set. The argument is carried by a training-set design that treats Ti3N2, Ti4N3, and Ti6N5 as ordered nitrogen-vacancy derivatives of rock-salt TiN and treats solid solutions as nitrogen occupying octahedral voids in hcp Ti, so the same local environments recur across compositions. A convex hull of formation energy versus composition is the device that turns the potential's predictions into a statement about thermodynamic stability.

What would settle it

Take a random sample of the intermediate structures generated in Section 3.7, for instance ten compositions along the N/Ti range covered by nitrogen insertion into pure Ti and nitrogen removal from Ti2N, and compute their formation energies with DFT at the same level used to train the potential. If more than a few of these energies deviate from the MTP predictions by more than about 10 meV/atom, the claim that structures across the full 0-to-1 range are thermodynamically stable would not survive.

Watch

Extended reading notes

Core claim

The central claim is that a single MTP interatomic potential, trained on Ti, Ti2N, Ti3N2, Ti4N3, TiN, and one solid solution (Ti-0.14N) and validated on Ti6N5 and Ti-0.2N, reliably describes formation energies, elastic constants, and thermodynamic stability across the full Ti-N system. The chosen potential, MTP0.75-0.25-0.25, reproduces DFT formation energies with a root-mean-square error of 2.1 meV/atom on training data and 6.8 meV/atom on the two test structures; the peak of the absolute-error distribution lies at 3.8 meV/atom for training-set systems and 7.6 meV/atom for unseen systems. Elastic constants predicted by the potential show a cumulative median error of 4.83% and an interquartile range of 6.12%, better than the two alternative MTP weightings considered. The authors then use the potential to generate intermediate structures by inserting nitrogen into octahedral voids and removing nitrogen from known phases, relaxing every candidate with the potential, and find that structures across the entire N/Ti range from 0 to 1 fall on or near the 0 K convex hull, with a maximum reported deviation of about 10 meV/atom.

Load-bearing premise

The whole map of stable intermediate phases rests on the assumption that a potential trained on a small set of known compounds and two solid solutions stays accurate for the many off-stoichiometric structures it was never tested against.

Editorial extensions

If this is right

  • The potential can compute formation energies of arbitrary Ti-N compositions at near-DFT accuracy, making full-composition convex-hull screening practical without new DFT runs.
  • Elastic constants across Ti-N compounds and solid solutions can be predicted with a median error around 5%, supporting estimates of bulk, shear, and Young's moduli from the same potential.
  • Molecular dynamics simulations of Ti-N compounds remain stable from 10 K to 300 K, so finite-temperature behavior can be studied with the same potential.
  • The success of including force and stress weights in training suggests that energy-only fitted potentials are insufficient for transferable predictions across a composition landscape.
  • New structures with N/Ti between 0 and 1 are predicted thermodynamically stable within 10 meV/atom of the 0 K convex hull, identifying specific intermediate stoichiometries as synthesis targets.

Reading between the lines

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

  • Going beyond the paper, the same iterative nitrogen insertion and removal protocol could map off-stoichiometric stability in other interstitial metal-nitride or metal-hydride systems.
  • If the predicted intermediate phases are confirmed by DFT or experiment, the potential could enable interface-scale simulations of Ti/TiN diffusion barriers, where composition gradients pass through these intermediate stoichiometries.
  • The validation set is small (two held-out structures), so a natural extension is to test the potential against a broader random sample of off-stoichiometric configurations before relying on the full predicted phase diagram.
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

3 major / 5 minor

Summary. The paper develops a moment tensor potential (MTP) for the Ti-N system, trained on DFT data for Ti, TiN, Ti2N, Ti3N2, Ti4N3, and the solid solution Ti-0.14N, with Ti6N5 and Ti-0.2N held out for testing. Three weighting schemes are compared; the authors select MTP0.75-0.25-0.25 as the most accurate based on formation-energy error distributions and elastic-constant errors. The selected potential is then used to relax and compute formation energies of structures generated by inserting N at octahedral voids and removing N from ordered compounds, leading to the central claim that structures with N/Ti ratios ranging from 0 to 1 can be thermodynamically stable, with a maximum deviation of 10 meV/atom from the 0 K convex hull.

Significance. If fully validated, this work would provide a fast and accurate interatomic potential for a technically important system and a broadly applicable approach to training-set selection for MTPs. The reported held-out formation-energy errors (RMSE 6.8 meV/atom; peak errors 5.9 and 7.6 meV/atom for the two test structures) and elastic-constant median error (~4.8%) are respectable for an MTP, and the systematic comparison against MEAM and M3GNet is useful. However, the headline conclusion about the thermodynamic stability of new intermediate phases rests entirely on MTP-extrapolated energies with no DFT confirmation for the predicted structures; as presented, that conclusion is not yet established.

major comments (3)
  1. [Section 3.7] The central claim that 'structures with N/Ti ratios ranging from 0 to 1 can be thermodynamically stable' is based solely on formation energies computed with the MTP0.75-0.25-0.25 potential for structures generated by arbitrary N insertion/removal. None of these predicted structures is checked against DFT. The only held-out evidence of transferability is Ti6N5 and Ti-0.2N (Section 3.4.2), which are chemically close to the training set (an ordered N-vacancy derivative of TiN and an interstitial hcp solid solution). The Section 3.7 generated structures, such as random N placement at farthest octahedral voids and iterative single-N removals, include local environments not represented in the test set. To support the hull claim, representative predicted structures at several intermediate compositions should be relaxed with DFT and their formation energies compared with the MTP values.
  2. [Section 3.7 / Abstract] The claimed 10 meV/atom maximum deviation from the convex hull is of the same order as the potential's held-out errors (peak absolute errors 5.9 and 7.6 meV/atom in Section 3.4.2). An energy error of this magnitude is sufficient to move a structure from on-hull to off-hull and to invert the relative stability of competing configurations. The paper should quantify the uncertainty of the predicted hull points, for example by propagating the observed test-set error distribution into the hull construction or by DFT-validating the near-hull structures.
  3. [Section 3.4.2, Tables 3 and 4] The KDE/FWHM values reported for MTP1-0-0 on Ti-0.14N (FWHM 2.091E+018, peak position 4.063E+018 meV/atom) and on Ti-0.2N (FWHM 1.8721E+018, peak position 8.2722E+018 meV/atom) are unphysical and indicate either numerical overflow, outliers in the error data, or a computational bug in the KDE implementation. These entries are reproduced in the text as 'orders of 10^18' without explanation. Such artifacts must be diagnosed and corrected, as they undermine confidence in the statistical analysis that is used to select the final potential.
minor comments (5)
  1. [Abstract and Section 4] The phrase 'maximum deviation of 10 meV/atom from the convex hull plot of formation energy 0K' is ambiguous: it should specify whether the deviation is measured from the DFT hull, from the MTP hull, or from the hull constructed using the predicted structures themselves.
  2. [Section 3.4.2, Eq. (2)-(4)] The bandwidth formula uses sigma, but the standard deviation of the error distribution is not defined in the text; please define sigma precisely.
  3. [Section 3.5] The sentence listing supercells ('Ti2N ,Ti3N2, Ti4N3, Ti6N5 supercell containing 36, 72, 42, 44 and 64 atoms') has five atom counts for five structures but names only four; please correct the list and clarify which structure corresponds to which count.
  4. [Throughout] There are numerous typographical and grammatical errors, including 'Ti-N material system have', 'solutions likeTi', 'MTP0.75−0.25−0.25' written without spaces, and 'Nose Hoover thermostat' for Nosé-Hoover. A careful proofreading pass is needed.
  5. [Section 2.1, Eq. (1)] The radial function f_mu appearing in the moment tensor definition is not defined; please specify its form and how the 421 basis functions and 2621 descriptors are derived from it.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the MTP is a fitted surrogate validated on held-out DFT structures, and the Section 3.7 phase-stability predictions are extrapolations rather than re-statements of training inputs.

full rationale

The paper develops a moment tensor potential by fitting to DFT energies, forces, and stresses of selected Ti-N phases and solid solutions, then uses the resulting potential to relax and evaluate newly generated intermediate structures. This is the standard surrogate-model workflow, not a circular derivation. The central claim in Section 3.7—that structures with N/Ti ratios from 0 to 1 can be thermodynamically stable—is obtained by generating configurations through nitrogen insertion/removal heuristics, relaxing them with MTP0.75-0.25-0.25, and computing formation energies with that same potential. Those structures are not part of the training set, and their energies are model outputs rather than fitted targets, so the claim is an extrapolation. The paper provides independent grounding on held-out structures: Section 3.4.2 tests Ti6N5 and Ti-0.2N, reporting peak absolute formation-energy errors of 5.9 and 7.6 meV/atom, and Section 3.6 explicitly acknowledges that overlap of the hull in Figure 6 is expected for trained compounds ('While this level of agreement is expected as we trained the potential with these structures'). That admission shows the authors do not present training-data agreement as a prediction. The self-citations ([7] and [31]) are used for background structural and solubility context, not as load-bearing uniqueness constraints or as substitutes for validation. The potential's transferability to the specific insertion/removal structures is a validation-coverage concern rather than a circularity: no equation in the paper defines the predicted formation energies as equal to the fit, and no fitted parameter is renamed as a prediction. Therefore the derivation chain is not circular, and the paper is suitably self-contained with respect to external DFT benchmarks.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The central claim depends on the MTP being a faithful surrogate of DFT, which requires trusting the DFT reference and the model capacity. No new physical entities are introduced. The hand-tuned hyperparameters (weights, tensor level, cut-offs) are the main fitting knobs.

free parameters (3)
  • MTP energy/force/stress weights = 0.75, 0.25, 0.25
    The weights in the loss function were chosen by testing three combinations and selecting the one with lowest test error; this is a hand-tuned hyperparameter that affects the resulting potential.
  • Maximum tensor level (levmax) = 22
    Chosen as a compromise between computational cost and accuracy; a higher level would change the model capacity.
  • Radial cut-off range = 2 Å to 7 Å
    The minimum and maximum cut-off distances are user-set parameters that define the local atomic environment.
assumptions (4)
  • domain assumption DFT-PBE formation energies are the ground-truth reference for training and validation.
    The entire potential is fitted to PBE-GGA DFT energies; any inaccuracy in the DFT functional propagates into the potential.
  • domain assumption The MTP functional form can represent the Ti-N potential energy surface.
    The paper assumes that moment tensor descriptors are sufficient to capture the relevant atomic interactions; Section 2.1.
  • domain assumption Structural similarity (ordered N-vacancy compounds) implies transferability to other Ti-N configurations.
    The training set selection is based on the premise that Ti3N2, Ti4N3, Ti6N5 are similar to TiN with ordered vacancies, so a model trained on some of them should generalize to the others.
  • standard math N2 gas energy is the correct reference for nitrogen in formation energy calculations.
    Equation (10) uses DFT total energy of N2 molecule as reference; this is standard practice but depends on the DFT treatment of the molecule.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Prediction of Mechanical Properties and Thermodynamic Stability of Ti-N system using MTP Interatomic Potential." pith.science (2026). https://pith.science/paper/CRKTPKJR

@misc{pith2026250718873,
  author       = {Pith},
  title        = {Pith review of: Prediction of Mechanical Properties and Thermodynamic Stability of Ti-N system using MTP Interatomic Potential},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/CRKTPKJR}},
  note         = {Machine review of arXiv:2507.18873}
}
read the original abstract

Ti-N material system have range of compounds with different stoichiometry like Ti2N, Ti3N2, Ti6N5, Ti4N3 alongwith Ti , TiN and solid solutions of N in Ti with a maximum of 23% solubility. In this work, we develop an interatomic potential based on moment tensor potential (MTP) that could reliably predict mechanical properties and thermodynamic stability of all Ti-N system. Taking into account the structural similarity and dissimilarity of various Ti-N system to choose training dataset was crucial for development of the potential. Root mean square error (RMSE) in prediction of formation energy using MTP potential compared to one calculated using density functional theory (DFT) for training dataset is 2.1 meV/atom and for testing dataset is 6.8 meV/atom. The frequency of absolute error in formation energy peaks at a maximum value of 3.8 meV/atom for system that was part of training dataset, while it peaks at 7.6 meV/atom for systems that are not part of the training dataset. Furthermore, the distribution and variability of elastic constants across compositions are systematically evaluated, revealing trends consistent with DFT benchmarks. The developed potential was used to predict energy of new phases in Ti-N system. We show that structures with N/Ti ratios ranging from 0 to 1 can be thermodynamically stable. A maximum deviation of 10 meV/atom from the convex hull plot of formation energy 0K was observed for a few system.

Figures

Figures reproduced from arXiv: 2507.18873 by the authors.

Figure 1
Figure 1. Comparison of Root Mean Square Error (RMSE) values to assess the performance of interatomic potential across (a) Training dataset (b) Test Structure 1 (c) Test Structure 2 To understand the efficacy of various interatomic potential, we calculate the RMSE in formation energy per atom with respect to DFT values for training dataset and also for structures which are not part of the training dataset ie.. the test datase… view at source ↗
Figure 2
Figure 2. Kernel Density Estimation (KDE) plots of the energy distributions for different structures of the training data in Ti-N system (a) TiN (b)Ti (c)Ti3N2 (d)Ti4N3 (e)Ti2N (f)Ti-0.14N. Each plot represent the absolute error distribution of different interatomic potentials w.r.t DFT calculated values 9 [PITH_FULL_IMAGE:figures/full_fig_p009_2.png] view at source ↗
Figure 3
Figure 3. Kernel Density Estimation (KDE) plots of the energy distributions for different structures of the test data in Ti-N system (a)Ti6N5 (b)Ti-0.2N (ii) MTP100−10−1 interatomic potential: The selection of this interatomic potential was guided by prior results from the work of Novoselov et al. [33]. The interatomic potential reported low RMSE values, suggesting high predictive accuracy. The energy weight was set to 100, w… view at source ↗
Figures from the paper (5 more)
Figure 4
Figure 4. Figure 4: Combined boxplot and swwarmplot analysis of percentage error distribution vs Elastic constants as predicted by various interatomic potentials. The visualization enables comparative analysis of predictive performance of potentials across different elastic constants 13 …
Figure 5
Figure 5. Figure 5: , the temperature vs time profiles has been plotted for different compounds which are part of the Ti-N system. The analysis of temperature-dependant fluctuations was used to assess the thermal behaviour and stability of the materials with NPT canonical ensemble. The in…
Figure 6
Figure 6. Figure 6: Convex hull plot generated from different interatomic potentials for Ti-N system. The hull plots illustrates the phase stability as well deviation of hull from DFT predicted Convex hull The formation energies of these are calculated with the most stable reference state…
Figure 6
Figure 6. Figure 6: While this level of agreement is expected as we trained the potential with these structures, it [PITH_FULL_IMAGE:figures/full_fig_p016_6.png]
Figure 7
Figure 7. Figure 7: Convex hull illustrating the stability of intermediate structures generated through insertion and removal of nitrogen atoms. Each intermediate points in the hull plot corresponds to specific composition derived from the most stable phase either by insertion or removal …

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

36 extracted references · 25 canonical work pages

  1. [1]

    Wróblewski, R

    P. Wróblewski, R. Rogólski, Experimental analysis of the influence of the application of tin, tialn, crn and dlc1 coatings on the friction losses in an aviation internal combustion engine intended for the propulsion of ultralight aircraft, Materials 14 (11 2021). doi:10.3390/ma14226839. 18

    work page doi:10.3390/ma14226839 2021

  2. [2]

    doi:10.1016/j.cja.2019.08.015

    Z.SUN,G.HE,Q.MENG,Y.LI,X.TIAN, Corrosion mechanisminvestigationoftin/ticoatingand tc4 alloy for aircraft compressor application, Chinese Journal of Aeronautics 33 (2020) 1824–1835. doi:10.1016/j.cja.2019.08.015

    work page doi:10.1016/j.cja.2019.08.015 2020

  3. [3]

    R. P. V. Hove, I. N. Sierevelt, B. J. V. Royen, P. A. Nolte, Titanium-nitride coating of orthopaedic implants: A review of the literature (2015). doi:10.1155/2015/485975

    work page doi:10.1155/2015/485975 2015

  4. [4]

    Oliva, F

    S. Oliva, F. Diomede, Y. D. Rocca, A. Mazzone, G. D. Marconi, J. Pizzicannella, O. Tru- biani, G. Murmura, Anti-tlr4 biological response to titanium nitride-coated dental implants: anti- inflammatory response and extracellular matrix synthesis, Frontiers in Bioengineering and Biotech- nology 11 (2023). doi:10.3389/fbioe.2023.1266799

    arXiv 2023

  5. [5]

    R. V. V. S. S. Krishna, M. A. S. F, Comparison of hss m35 and titanium-nitride coated tool in novel cnc drilling operation of aluminium alloy aa6351-t6 material on surface roughness, in: Materials Today: Proceedings, Vol. 69, Elsevier Ltd, 2022, pp. 1014–1018. doi:10.1016/j.matpr.2022.08.098

    work page doi:10.1016/j.matpr.2022.08.098 2022

  6. [6]

    Kizil, C

    H. Kizil, C. Steinbrüchel, Tin, tan and wxn as diffusion barriers for cu on sio 2: Capacitance-voltage, leakage current, and triangular-voltage-sweep tests after bias temperature stress, Thin Solid Films 449 (2004) 158–165. doi:10.1016/j.tsf.2003.10.111

    work page doi:10.1016/j.tsf.2003.10.111 2004

  7. [7]

    Gollapalli, J

    P. Gollapalli, J. Varalakshmi, P. S. Kishor, P. Oza, S. K. Yadav, Atomically chemically graded ti/tin interface, Applied Surface Science 597 (2022) 153637. doi:10.1016/j.apsusc.2022.153637. URLhttps://doi.org/10.1016/j.apsusc.2022.153637

    work page arXiv 2022

  8. [8]

    H. A. Wriedt, J. L. Murray, Title of the specific section or chapter, in: T. B. Massalski (Ed.), Binary Alloy Phase Diagrams, Vol. 3, pp. 2705–2710

    work page

Show all 36 references
  1. [9]

    S. Yu, Q. Zeng, A. R. Oganov, G. Frapper, L. Zhang, Phase stability, chemical bonding and mechanical properties of titanium nitrides: A first-principles study, Physical Chemistry Chemical Physics 17 (2015) 11763–11769. doi:10.1039/c5cp00156k

    2015 doi

  2. [10]

    Y. M. Kim, B. J. Lee, Modified embedded-atom method interatomic potentials for the ti-c and ti-n binary systems, Acta Materialia 56 (2008) 3481–3489. doi:10.1016/j.actamat.2008.03.027

    2008 doi

  3. [11]

    K. Wan, J. He, X. Shi, Construction of high accuracy machine learning interatomic potential for surface/interface of nanomaterials—a review (5 2024). doi:10.1002/adma.202305758

    2024 doi

  4. [12]

    N. Ran, L. Yin, W. Qiu, J. Liu, Recent advances in machine learning interatomic potentials for cross-scale computational simulation of materials (4 2024). doi:10.1007/s40843-023-2836-0

    2024 doi

  5. [13]

    Mueller, A

    T. Mueller, A. Hernandez, C. Wang, Machine learning for interatomic potential models (2 2020). doi:10.1063/1.5126336

    2020 doi

  6. [14]

    Z. Qiao, M. Welborn, A. Anandkumar, F. R. Manby, T. F. Miller, Orbnet: Deep learning for quantum chemistry using symmetry-adapted atomic-orbital features, Journal of Chemical Physics 153 (9 2020). doi:10.1063/5.0021955

    2020 doi

  7. [15]

    J. Wang, P. Liu, H. Zhu, M. Liu, H. Ma, Y. Chen, Y. Sun, X. Q. Chen, Efficient moment tensor machine-learning interatomic potential for accurate description of defects in ni-al alloys, Physical Review Materials 9 (5 2025). doi:10.1103/PhysRevMaterials.9.053805. 19

    2025 doi

  8. [16]

    Gubaev, V

    K. Gubaev, V. Zaverkin, P. Srinivasan, A. I. Duff, J. Kästner, B. Grabowski, Performance of two complementary machine-learned potentials in modelling chemically complex systems, npj Compu- tational Materials 9 (12 2023). doi:10.1038/s41524-023-01073-w

    2023 doi

  9. [17]

    A. V. Shapeev, Moment tensor potentials: a class of systematically improvable interatomic poten- tials (12 2015). doi:10.1137/15M1054183. URLhttp://arxiv.org/abs/1512.06054 http://dx.doi.org/10.1137/15M1054183

    2015 arXiv

  10. [18]

    J. Qi, Z. H. Aitken, Q. Pei, A. M. Z. Tan, Y. Zuo, M. H. Jhon, S. S. Quek, T. Wen, Z. Wu, S. P. Ong, Machine learning moment tensor potential for modeling dislocation and fracture in l10- tial and d019-ti3al alloys, Physical Review Materials 7 (10 2023). doi:10.1103/PhysRevMat...

    2023 doi

  11. [19]

    I. E. Gaa, P. H. Ecole, I. Xornzale Superzeure, France, , W. Konnt, Physical review, Tech. rep

  12. [20]

    W. Kohn, L. J. Sham, Physical review self-consistent equations including exchange and correlation effects*, Tech. rep

  13. [21]

    Kresse, J

    G. Kresse, J. F. B. ’, Efficiency of ab-initio total energy calculations for metals and semiconductors using a plane-wave basis set, Tech. rep. (1996)

    1996

  14. [22]

    Kresse, J

    G. Kresse, J. Furthmü, Efficient iterative schemes for ab initio total-energy calculations using a plane-wave basis set, Tech. rep., University of Vienna, preprint (1996)

    1996

  15. [23]

    J. P. Perdew, K. Burke, M. Ernzerhof, Generalized gradient approximation made simple, Tech. rep. (1996)

    1996

  16. [24]

    Joubert, From ultrasoft pseudopotentials to the projector augmented-wave method, Physical Review B - Condensed Matter and Materials Physics 59 (1999) 1758–1775

    D. Joubert, From ultrasoft pseudopotentials to the projector augmented-wave method, Physical Review B - Condensed Matter and Materials Physics 59 (1999) 1758–1775. doi:10.1103/PhysRevB.59.1758

    1999 doi

  17. [25]

    H. J. Monkhorst, J. D. Pack, Special points for brillonin-zone integrations*, Tech. rep. (1976)

    1976

  18. [26]

    URLhttps://www.vasp.at/wiki/index.php/LANGEVIN_GAMMA

    VASP Wiki contributors, Langevin, accessed: 2025-06-13 (2025). URLhttps://www.vasp.at/wiki/index.php/LANGEVIN_GAMMA

    2025

  19. [27]

    G. Wang, C. Wang, X. Zhang, Z. Li, J. Zhou, Z. Sun, Machine learning interatomic poten- tial: Bridge the gap between small-scale models and realistic device-scale simulations (5 2024). doi:10.1016/j.isci.2024.109673

    2024

  20. [28]

    Stoehr, C

    M. Stoehr, C. S. Shin, I. Petrov, J. E. Greene, Raman scattering from tinx (0.67≤x≤1.00) single crystals grown on mgo(001), Journal of Applied Physics 110 (10 2011). doi:10.1063/1.3651381

    2011 doi

  21. [29]

    K. Liu, X. L. Zhou, H. H. Chen, L. Y. Lu, Structural and elastic properties of tin under high pressure, Physica B: Condensed Matter 407 (2012) 3617–3621. doi:10.1016/j.physb.2012.05.038

    2012 doi

  22. [30]

    R. Yang, B. Chai, C. Zhu, Q. Wei, Z. Du, First-principles study on the structural, elastic and electronic properties of ti4n3 and ti6n5 under high pressure, Modern Physics Letters B 31 (2017) 1750349. doi:10.1142/S0217984917503493. 20

    2017 doi

  23. [31]

    Varalakshmi, S

    J. Varalakshmi, S. K. Yadav, Thermodynamics and kinetics of ti2n formation by n atom inter- calation in ti: A density functional theory study, Physica B: Condensed Matter 699 (2 2025). doi:10.1016/j.physb.2024.416829

    2025

  24. [32]

    Mikhalev, I

    A. Mikhalev, I. V. Oseledets, Rectangular maximum-volume submatrices and their applications, Linear Algebra and Its Applications 538 (2018) 187–211. doi:10.1016/j.laa.2017.10.014

    2018 doi

  25. [33]

    I. I. Novoselov, A. V. Yanilkin, A. V. Shapeev, E. V. Podryabinkin, Moment tensor potentials as a promising tool to study diffusion processes, Computational Materials Science 164 (2019) 46–56. doi:10.1016/j.commatsci.2019.03.049

    2019 doi

  26. [34]

    V. Wang, N. Xu, J. C. Liu, G. Tang, W. T. Geng, Vaspkit: A user-friendly interface facilitating high-throughput computing and analysis using vasp code, Computer Physics Communications 267 (10 2021). doi:10.1016/j.cpc.2021.108033

    2021

  27. [35]

    Y. T. Meharenna, T. L. Poulos, Using molecular dynamics to probe the structural basis for enhanced stability in thermal stable cytochromes p450, Biochemistry 49 (2010) 6680–6686. doi:10.1021/bi100929x

    2010 doi

  28. [36]

    C. Chen, S. P. Ong, A universal graph deep learning interatomic potential for the periodic table, Nature Computational Science 2 (2022) 718–728. doi:10.1038/s43588-022-00349-3. 21

    2022 doi

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.