Pith. sign in

REVIEW 3 major objections 4 minor 60 references

Classification and enumeration of solid-solid phase transition mechanisms

T0 review · 3 major / 4 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read This paper claims that all physically relevant crystal-structure matches between two phases can be exhaustively enumerated, and provides an algorithm that does it.

desk verdict The cell-independent formalism and finiteness theorems are genuinely good; the 'exhaustive enumeration' claim, however, rests on an unproven sampling bound and should not be taken as a certificate. read the letter →

arxiv 2506.05105 v1 pith:VT4Z5SGH submitted 2025-06-05 cond-mat.mtrl-sci physics.comp-ph

Fang-Cheng Wang , Qi-Jun Ye , Yu-Cheng Zhu , Xin-Zheng Li This is my paper
classification cond-mat.mtrl-sciphysics.comp-ph
keywords crystal-structurematchsolid-solidphasetransitionsublatticeintegermatrixrepresentationdeformationgradientshuffledistancemechanismsenumerationalgorithm
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 tries to prove that the set of physically relevant crystal-structure matches (CSMs) between two phases is finite and fully listable, and to provide an algorithm that lists them. A CSM is the atom-to-atom correspondence between an initial and a final crystal structure; it defines the mechanism of a solid-solid phase transition. The authors build a cell-independent classification where every CSM decomposes into a lattice deformation (an SLM) plus a shuffle, and they show that bounding multiplicity, strain, and shuffle distance leaves only finitely many noncongruent CSMs (Theorem 9). The crystmatch algorithm then enumerates these by representing them as integer matrices, reproducing known mechanisms in minutes and yielding thousands of previously unreported candidates. If correct, this gives a complete input list for barrier calculations, orientation-relationship analysis, and machine learning of transition mechanisms.

What carries the argument

The load-bearing object is the IMT (integer-matrix triplet) representation of a sublattice match, together with the PCT (permutation with class-wise translations) representation of the shuffle part. An IMT is a pair of Hermite normal form matrices encoding the sublattices plus a unimodular matrix Q encoding the deformation gradient; Theorem 5 makes this a bijection with SLMs of a given multiplicity, so enumeration reduces to scanning finite integer-matrix sets. The PCT representation turns every CSM with a fixed SLM into a permutation and lattice translations, which is what makes the shuffle-distance cutoff enumerable. The tree structure with multiplicity as depth and the finiteness theorem (Theorem 9) carry the exhaustiveness claim.

What would settle it

Run the enumeration twice with different sampling distributions or random seeds (or with an independent exact enumeration for small multiplicity µ) and check whether the two IMT lists agree; any IMT appearing in one list and missing from the other, or any measured CSM with strain and shuffle distance within the stated bounds that is absent from the database, would refute the exhaustiveness claim.

Watch

Extended reading notes

Core claim

The central discovery is that the CSM optimization problem — finding the best atom-to-atom correspondence under any geometric criterion — is solved once all CSMs are organized into a tree by multiplicity and sublattice match. With the physically motivated upper bounds µ ≤ µ_max, w(S) ≤ w_max, and d(J) ≤ d_max, the tree has finitely many noncongruent leaves (Theorem 9). The proof uses the IMT representation: any SLM is uniquely encoded by a pair of Hermite normal form matrices and a unimodular matrix Q, making the enumeration a finite integer-matrix search, and any shuffle is uniquely a permutation with class-wise translations. Algorithm 1 loops over trial deformation gradients, rounding to integer matrices and stopping when fresh IMTs cease to appear; Algorithms 2 and 3 enumerate shuffles up to the shuffle-distance cutoff. On B1–B2 and graphite-to-diamond, the method recovers the established Buerger, WTM, TKED, and TGS mechanisms and identifies new deformation gradients that reproduce observed orientation relationships without rotation.

Load-bearing premise

The exhaustive claim of Algorithm 1 rests on the assumption that its random sampling of trial deformation gradients is dense enough to hit the preimage of every relevant integer matrix; non-appearance of a new IMT for many consecutive draws is treated as proof that none remains.

Editorial extensions

If this is right

  • Every CSM allowed by the chosen bounds can be supplied as an input to solid-state NEB-like barrier calculations, so the lowest-barrier path no longer depends on a user-specified guess.
  • The database can be filtered by experimentally observed orientation relationships; the paper shows that only one deformation gradient realizes the WB OR exactly and two previously unreported ones realize the OFO OR.
  • For most transitions, reproducing all known deformation mechanisms and CSMs takes under 10 CPU minutes, so the enumeration is cheap enough for high-throughput screening.
  • The lattice-mismatch spectrum under pressure shows which deformation gradients become energetically easier as pressure increases, potentially identifying pressure-dependent mechanism changes.
  • The enumerated CSM set can serve as training data for machine-learning models of transition mechanisms.

Reading between the lines

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

  • If the enumeration is truly exhaustive, then any simulation study that reports a mechanism outside this database has either violated the stated bounds or is using a non-geometric selection rule; this gives a clean way to audit simulation outcomes.
  • The tree classification suggests a natural database schema, with multiplicity as a primary key and SLM/PCT as separate tables, so experimental orientation relationships could be matched to candidate mechanisms in milliseconds.
  • The finiteness theorem is conditional on the chosen cutoffs; a testable extension would be to relax the shuffle-distance or multiplicity bound for a specific material and map how the mechanism landscape grows with the bounds.
  • Because the formalism is cell-independent and species-preserving, the same enumeration machinery should transfer to heteroepitaxy, grain-boundary migration, and defect migration, as the authors note in passing.
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 / 4 minor

Summary. The paper proposes a cell-independent mathematical formalism for crystal-structure matches (CSMs) in solid-solid phase transitions, decomposing each CSM into a deformation gradient, a sublattice match (SLM), and a shuffle, with congruence relations under crystal symmetries. It represents SLMs by integer-matrix triplets (IMT) and shuffles by permutation-with-translations (PCT), proves that under bounds on multiplicity, strain energy, and shuffle distance the number of noncongruent CSMs is finite (Theorems 7-9), and presents three algorithms to enumerate SLMs and CSMs. Applications to B1-B2 and graphite-to-diamond reproduce several known mechanisms and report many new candidates.

Significance. The formal framework is well-structured, with carefully stated definitions, lemmas, and theorem proofs; the IMT/PCT encoding is elegant and cell-independent, and the open-source implementation plus timing data are valuable. If the enumeration algorithms were certified complete, this would indeed provide a practical resolution of the CSM enumeration problem and enable high-throughput barrier studies. However, the central exhaustive-enumeration claim currently rests on probabilistic sampling and heuristic optimization, and the thresholds bounding 'practical interest' are empirical; hence the significance is conditional on closing these gaps.

major comments (3)
  1. [Section IV A, Eq. (91) and Appendix B] The exhaustiveness of Algorithm 1 is probabilistic, not mathematically certified. The stopping rule is a hypothesis test that rejects incompleteness only if the lower bound η in Eq. (B2) is valid; that bound rests on the assertion that the volume of the relevant section of the rounding preimage is 'no less than 1', which is stated without proof. Consequently, the algorithm may terminate while some IMT congruence classes remain undiscovered, and statements such as 'essentially an exhaustive search algorithm' (Section IV A) and the abstract's 'rigorously proved ... exhaustively enumerate' are not supported. The later claim in Section V A that the WTM/TKED deformation gradient is 'the only one that can produce the WB OR' depends directly on this completeness.
  2. [Section IV B, Algorithms 2 and 3] Algorithm 2's enumeration of all bottom PCTs assumes OptimizePCT returns the global minimum of the shuffle distance for each constraint. The text states that basin-hopping is employed for the τ minimization; basin-hopping is a stochastic global-optimization heuristic with no convergence certificate in this setting. If OptimizePCT misses a bottom PCT, the corresponding permutation is never split and its PCTs are omitted from mypcts. No exact bound or deterministic alternative is supplied, so the 'exhaustive' claim for shuffles is also not certified.
  3. [Section III B 3, Eqs. (72)-(74) and Fig. 2] The finiteness theorem applies only after choosing empirical cutoffs μ_max, w_max, and d_max; the text explicitly calls Eq. (72) an 'empirical constraint inferred from Table II'. The abstract's phrase 'only a finite number of noncongruent CSMs are of practical interest' and the conclusion that the method solves the CSM optimization problem 'under any geometric criterion' therefore hold only relative to cutoffs that were selected to include previously known mechanisms. This does not invalidate the formalism, but the scope needs to be stated precisely and the transferability of the cutoffs justified or explicitly left as a user choice.
minor comments (4)
  1. [After Eq. (69) in Section III B 2] The sentence beginning 'Eq. 9 ensures...' appears to reference the wrong equation; it should refer to the well-definedness argument for the shuffle distance, likely Eq. (68) or a related definition.
  2. [Figures 9-14] Several axis labels in the submitted figures are garbled or unreadable (for example, the vertical axis label in Fig. 12 and the horizontal label in Fig. 14). Please regenerate the figures with properly embedded vector fonts.
  3. [Section IV B, Algorithm 3] The word 'pseodocode' should be 'pseudocode'.
  4. [Fig. 2 caption and Section V] The caption states that the bounds are 'empirical (see Table II)', but the default numerical values of μ_max, w_max, and d_max used in the applications are not collected in one place; please state them explicitly for each application.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the finiteness theorem and IMT/PCT representations are derived self-contained; the sampling-based exhaustiveness gap is a completeness risk, not a circular reduction.

full rationale

The paper's central mathematical derivation is self-contained. The finiteness of CSM*(A,B) (Theorem 9) follows from Theorem 7 (for fixed multiplicity µ, only finitely many IMTs survive a strain bound, using finiteness of HNF(k) and a singular-value bound on Q) and Theorem 8 (for a fixed SLM, only finitely many noncongruent shuffles survive a shuffle-distance bound, proved by a lower-bound argument on d). These proofs do not assume the existence of any particular CSM or the completeness of any database. The IMT and PCT representations are established by bijections proved in Theorems 5 and 6; the citation to Ref. [3] for the IMT name is not load-bearing because the mapping is re-proved here. The thresholds µ_max, w_max, and d_max are explicitly introduced as empirical and computational constraints, not as fitted parameters, and the reproduction of known mechanisms in Table II is a sanity check rather than a prediction derived from those thresholds; the two WYZLJ entries sourced from the authors' own Ref. [3] are benchmark entries that do not support the finiteness or enumeration theorems. The main caveat is Algorithm 1's probabilistic stopping rule (Eq. 91) and the unproved volume lower bound in Appendix B, which means the 'exhaustive' enumeration is not a deterministic certificate. That is a correctness/completeness limitation, not a circularity: the algorithm's output is not defined in terms of the desired complete set, and the completeness claim is an inference from a sampling hypothesis, not an identity or a fit. Because no load-bearing step reduces by construction to its own input, the paper has no significant circularity.

Assumptions & free parameters 4 free parameters · 5 assumptions · 3 invented entities

The enumeration is self-contained only after the user supplies material parameters and cutoffs, and after choosing a stochastic sampling distribution. All reported counts and new candidates are relative to those choices.

free parameters (4)
  • Multiplicity cutoff µ_max = 6 or 12 in the B1-B2 study; 6 in the graphite-diamond study
    User-selected bound on CSM intricacy, inferred from known mechanisms in Table II (multiplicity values 1 to 6) rather than derived. It directly determines the breadth of the enumeration tree.
  • Strain energy cutoff w_max = 13×10^8 J/m^3 for B1-B2; 5×10^10 J/m^3 for graphite
    Chosen large enough to include the Buerger and other known mechanisms. The strain energy estimate in Eq. (66) uses elastic constants and an ad hoc split of deformation between initial and final structures, so w(S) is not the true PES barrier density.
  • Shuffle distance cutoff d_max = 1.8 Å in the B1-B2 exhaustive enumeration
    Selected to include the TKED mechanism; the enumerated set is finite only after this empirical cutoff and congruence reduction.
  • Sampling completeness parameters ϵ and η (or i_th) = i_th = ln ϵ / ln(1-η), dynamically adjusted in practice
    Algorithm 1 declares the IMT set complete after i_th consecutive random trials produce nothing new. The probability estimate relies on a lower bound η that is not rigorously computed for each material; this is a probabilistic completeness assumption, not a proof of exhaustive coverage.
assumptions (5)
  • standard math Classical lattice theory, Hermite normal form uniqueness, elementary divisor theorem, SVD and polar decomposition.
    Invoked for the IMT bijection (Theorem 5), for the finiteness arguments, and for the strain energy estimate.
  • domain assumption Every CSM of interest has a finite multiplicity and a deformation gradient S ∈ GL(3,R).
    Section II C and Definition 9 declare CSMs without a deformation gradient to have multiplicity ∞ and exclude them from enumeration; all subsequent algorithms operate only on such decomposable CSMs.
  • domain assumption Low-barrier SSPT mechanisms satisfy the empirical bounds µ ≤ µ_max, w(S) ≤ w_max, d(J) ≤ d_max.
    Section III B, equations (72)-(74). These bounds are presented as empirical and inferred from Table II; all finiteness and enumeration results are relative to them.
  • ad hoc to paper The 'standard shuffle' and strain energy w(S) defined in Eqs. (65)-(67) faithfully rank mechanisms by energetic cost.
    The split deformation of A and B plus elastic stiffness tensors is an estimator. The paper uses it as a pruning criterion but does not benchmark it against actual energy barriers for the enumerated mechanisms.
  • ad hoc to paper The stochastic sampling distribution on SO(3) from Appendix B covers each IMT's rounding preimage with a known positive probability.
    Used to derive i_th. The lower bound depends on approximating the intersection as a thin 3D section and on σ_7,σ_8,σ_9 of φ; it is an estimate, not a certified geometric covering.
invented entities (3)
  • SLM (sublattice match)
    purpose: Classify CSMs into a deformation gradient S and a shuffle lattice pair, making multiplicity cell-independent.
    An internal mathematical structure, not a physical observable; its value is that congruence classes of CSMs can be enumerated through IMTs.
  • IMT and PCT representations
    purpose: Represent each SLM and each shuffle as integer matrices, permutations, and lattice translations for computer generation.
    Algorithmic encodings with no experimental handle by themselves; they enable enumeration but produce no directly measurable prediction beyond the geometric candidates.
  • Standard shuffle J̃
    purpose: Symmetrically split a CSM into a deformation and a translation, and define the shuffle distance d(J).
    A construction specific to this paper; the shuffle distance is used for pruning and ranking, but its physical relevance is justified only by reference to average displacement and mass weighting.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Classification and enumeration of solid-solid phase transition mechanisms." pith.science (2026). https://pith.science/paper/VT4Z5SGH

@misc{pith2026250605105,
  author       = {Pith},
  title        = {Pith review of: Classification and enumeration of solid-solid phase transition mechanisms},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/VT4Z5SGH}},
  note         = {Machine review of arXiv:2506.05105}
}
read the original abstract

Crystal-structure match (CSM), the atom-to-atom correspondence between two crystalline phases, is used extensively to describe solid-solid phase transition (SSPT) mechanisms. However, existing computational methods cannot account for all possible CSMs. Here, we propose a formalism to classify all CSMs into a tree structure, which is independent of the choices of unit cell and supercell. We rigorously proved that only a finite number of noncongruent CSMs are of practical interest. By representing CSMs as integer matrices, we introduce the crystmatch method to exhaustively enumerate them, which uncontroversially solves the CSM optimization problem under any geometric criterion. For most SSPTs, crystmatch can reproduce all known deformation mechanisms and CSMs within 10 CPU minutes, while also revealing thousands of new candidates. The resulting database can be further used for comparing experimental phenomena, high-throughput energy barrier calculations, or machine learning.

Figures

Figures reproduced from arXiv: 2506.05105 by the authors.

Figure 1
Figure 1. FIG. 1. A schematic diagram showing that the initial structure (orthorhombic, yellow) may evolve to the final structure [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. FIG. 2. Given the initial and final crystal structures, all [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 4
Figure 4. FIG. 4. (a) An example of [PITH_FULL_IMAGE:figures/full_fig_p004_4.png] view at source ↗
Figures from the paper (10 more)
Figure 5
Figure 5. Figure 5: FIG. 5. An example of [PITH_FULL_IMAGE:figures/full_fig_p005_5.png]
Figure 6
Figure 6. Figure 6: FIG. 6. Two PCTs representing [PITH_FULL_IMAGE:figures/full_fig_p009_6.png]
Figure 7
Figure 7. Figure 7: FIG. 7. A schematic diagram showing how Algorithm 1 works, [PITH_FULL_IMAGE:figures/full_fig_p012_7.png]
Figure 8
Figure 8. Figure 8: FIG. 8. A schematic diagram showing how Algorithm 2 works, [PITH_FULL_IMAGE:figures/full_fig_p013_8.png]
Figure 9
Figure 9. Figure 9: FIG. 9. All deformation gradients [PITH_FULL_IMAGE:figures/full_fig_p015_9.png]
Figure 10
Figure 10. Figure 10: FIG. 10. All deformation gradients [PITH_FULL_IMAGE:figures/full_fig_p016_10.png]
Figure 11
Figure 11. Figure 11: FIG. 11. How the shuffle distance [PITH_FULL_IMAGE:figures/full_fig_p016_11.png]
Figure 12
Figure 12. Figure 12: FIG. 12. All CSMs with [PITH_FULL_IMAGE:figures/full_fig_p017_12.png]
Figure 13
Figure 13. Figure 13: In fact, all CSMs exhibit this repetitive property [PITH_FULL_IMAGE:figures/full_fig_p017_13.png]
Figure 14
Figure 14. Figure 14: (b) extends leftward with increasing pressure, con￾sistent with previous experimental and computational studies. VI. SUMMARY We have introduced a cell-independent formalism that unequivocally defines the deformation gradient, SLM, and standard shuffle of a CSM, as wel…

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

60 extracted references · 59 canonical work pages

  1. [1]

    We say thatC(Z 3) = Ck k∈Z 3 is the lattice generated byC, andCis the base matrix of C(Z 3)

    Representing sublattice matches Definition 12 (Base Matrix).LetC∈R 3×3 be a non- singular matrix. We say thatC(Z 3) = Ck k∈Z 3 is the lattice generated byC, andCis the base matrix of C(Z 3). Lemma 2.Any lattice has a base matrix. Lemma 3.LetC, ˜C∈R 3×3 be nonsingular matrices. ˜C(Z 3) is a sublattice ofC(Z 3) if and only if there exists anM∈Z 3×3 such tha...

    work page

  2. [2]

    permutation with class-wise translations

    Representing shuffles Definition 16 (Motif ).LetAbe a crystal structure and ˜LA a sublattice ofA. We say that{a 1,a 2 · · ·,an} ⊂ Ais an ˜LA-motif ofAifa 1,· · ·,an are ˜LA-inequivalent and n=|A/ ˜LA|. Theorem 6.Let ( ˜LA, ˜LB, S) be an SLM with period˜Z, {a1,· · ·,a˜Z}an ˜LA-motif ofA, and{b 1,· · ·,b˜Z}an ˜LB- motif ofB. IfJ ∈CSM(A,B) has the SLM ( ˜LA,...

    work page

  3. [3]

    Strain energy density From an energetic perspective, the constraint on strain is necessary, since excessive deformation would lead to a large strain energy. We are only concerned with those CSMs whose deformation gradients satisfy w(S)≤w max,(63) wherew(S) denotes the strain energy density associated with the deformation gradientS, andw max is a trun- cat...

    work page

  4. [4]

    average distance

    Shuffle distance For each SLM, there are still infinitely many possible CSMs—even a single atoma∈ Ahas infinitely many choices of its counterpartJ(a). An intuitive constraint is thatJ(a) should not be too far froma. This is not an energetic consideration, but rather an empirical one, as discussed in Ref. 25, where the average distance traveled by the atom...

    work page

  5. [5]

    all CSMs

    Finiteness of candidate crystal-structure matches As shown in Fig. 2, the enumeration includes a CSM Jwith multiplicityµand deformation gradientSif µ≤µ max,(72) w(S)≤w max,(73) d(J)≤d max.(74) They have all been detailed earlier, except for Eq. (72), which can be regarded as an empirical constraint inferred from Table II—despite MD and MetaD simulations u...

    work page

  6. [6]

    Stevanovi´ c, R

    V. Stevanovi´ c, R. Trottier, C. Musgrave, F. Therrien, A. Holder, and P. Graf, Predicting kinetics of polymor- phic transformations from structure mapping and coor- dination analysis, Phys. Rev. Mater.2, 033802 (2018)

    work page 2018

  7. [7]

    Therrien, P

    F. Therrien, P. Graf, and V. Stevanovi´ c, Matching crystal structures atom-to-atom, J. Chem. Phys.152, 074106 (2020)

    work page 2020

  8. [8]

    Wang, Q.-J

    F.-C. Wang, Q.-J. Ye, Y.-C. Zhu, and X.-Z. Li, Crystal- structure matches in solid-solid phase transitions, Phys. Rev. Lett.132, 086101 (2024)

    work page 2024

Show all 60 references
  1. [9]

    D. Zahn, Y. Grin, and S. Leoni, Mechanism of the pressure-induced wurtzite to rocksalt transition of cdse, Phys. Rev. B72, 064110 (2005)

    2005

  2. [10]

    Shimojo, S

    F. Shimojo, S. Kodiyalam, I. Ebbsj¨ o, R. K. Kalia, A. Nakano, and P. Vashishta, Atomistic mechanisms for wurtzite-to-rocksalt structural transformation in cad- mium selenide under pressure, Phys. Rev. B70, 184111 (2004)

    2004

  3. [11]

    Cai and N

    J. Cai and N. Chen, First-principles study of the wurtzite-to-rocksalt phase transition in zinc oxide, J. Phys.: Condens. Matter19, 266207 (2007)

    2007

  4. [12]

    Zahn and S

    D. Zahn and S. Leoni, Nucleation and growth in pressure- induced phase transitions from molecular dynamics simu- lations: Mechanism of the reconstructive transformation of NaCl to the CsCl-type structure, Phys. Rev. Lett.92, 250201 (2004)

    2004

  5. [13]

    Badin and R

    M. Badin and R. Martoˇ n´ ak, Nucleating a different co- ordination in a crystal under pressure: A study of the B1-B2 transition in NaCl by metadynamics, Phys. Rev. Lett.127, 105701 (2021)

    2021

  6. [14]

    P. A. Santos-Florez, H. Yanxon, B. Kang, Y. Yao, and Q. Zhu, Size-dependent nucleation in crystal phase tran- sition from machine learning metadynamics, Phys. Rev. Lett.129, 185701 (2022)

    2022

  7. [15]

    R. Z. Khaliullin, H. Eshet, T. D. K¨ uhne, J. Behler, and M. Parrinello, Nucleation mechanism for the direct graphite-to-diamond phase transition, Nat. Mater.10, 693 (2011)

    2011

  8. [16]

    Eyring, The activated complex in chemical reactions, J

    H. Eyring, The activated complex in chemical reactions, J. Chem. Phys.3, 107 (1935)

    1935

  9. [17]

    Sheppard, P

    D. Sheppard, P. Xiao, W. Chemelewski, D. D. Johnson, and G. Henkelman, A generalized solid-state nudged elas- tic band method, J. Chem. Phys.136, 074103 (2012)

    2012

  10. [18]

    Therrien, E

    F. Therrien, E. B. Jones, and V. Stevanovi´ c, Metastable materials discovery in the age of large-scale computation, Appl. Phys. Rev.8, 031310 (2021)

    2021

  11. [19]

    L. Zhu, R. E. Cohen, and T. A. Strobel, Phase transi- tion pathway sampling via swarm intelligence and graph theory, J. Phys. Chem. Lett.10, 5019 (2019)

    2019

  12. [20]

    Shang, X.-J

    C. Shang, X.-J. Zhang, and Z.-P. Liu, Stochastic surface walking method for crystal structure and phase transition pathway prediction, Phys. Chem. Chem. Phys.16, 17845 (2014)

    2014

  13. [21]

    Guan, X.-J

    S.-H. Guan, X.-J. Zhang, and Z.-P. Liu, Energy land- scape of zirconia phase transitions, J. Am. Chem. Soc. 137, 8010 (2015)

    2015

  14. [22]

    Zhang and Z.-P

    X.-J. Zhang and Z.-P. Liu, Variable-cell double-ended surface walking method for fast transition state location of solid phase transitions, J. Chem. Theory Comput.11, 4885 (2015)

    2015

  15. [23]

    X. Chen, Y. Song, N. Tamura, and R. D. James, Deter- mination of the stretch tensor for structural transforma- tions, J. Mech. Phys. Solids93, 34 (2016)

    2016

  16. [24]

    Huang, Y

    Q.-S. Huang, Y. Zhang, P. F. Liu, H. G. Yang, X. Zhang, and S.-H. Wei, Interface-confined intermediate phase in tio2 enables efficient photocatalysis, Proc. Natl. Acad. 22 Sci. U.S.A.121, e2318341121 (2024)

    2024

  17. [25]

    M. J. Field,A practical introduction to the simulation of molecular systems(Cambridge University Press, 1999) pp. 153–156

    1999

  18. [26]

    Scandolo, M

    S. Scandolo, M. Bernasconi, G. L. Chiarotti, P. Focher, and E. Tosatti, Pressure-induced transformation path of graphite to diamond, Phys. Rev. Lett.74, 4015 (1995)

    1995

  19. [27]

    Laio and M

    A. Laio and M. Parrinello, Escaping free-energy minima, Proc. Natl. Acad. Sci. U.S.A.99, 12562 (2002)

    2002

  20. [28]

    Martoˇ n´ ak, A

    R. Martoˇ n´ ak, A. Laio, and M. Parrinello, Predicting crys- tal structures: the Parrinello-Rahman method revisited, Phys. Rev. Lett.90, 075503 (2003)

    2003

  21. [29]

    Li and Z.-P

    Y.-F. Li and Z.-P. Liu, Smallest stable Si/SiO 2 inter- face that suppresses quantum tunneling from machine- learning-based global search, Phys. Rev. Lett.128, 226102 (2022)

    2022

  22. [30]

    Therrien and V

    F. Therrien and V. Stevanovi´ c, Minimization of atomic displacements as a guiding principle of the martensitic phase transformation, Phys. Rev. Lett.125, 125502 (2020)

    2020

  23. [31]

    Capillas, J

    C. Capillas, J. Perez-Mato, and M. Aroyo, Maximal sym- metry transition paths for reconstructive phase transi- tions, J. Phys.: Condens. Matter19, 275203 (2007)

    2007

  24. [32]

    H. T. Stokes, D. M. Hatch, J. Dong, and J. P. Lewis, Mechanisms for the reconstructive phase transition be- tween the B 1 and B 2 structure types in NaCl and PbS, Phys. Rev. B69, 174111 (2004)

    2004

  25. [33]

    E. C. Bain and N. Dunkirk, The nature of martensite, Trans. AIME70, 25 (1924)

    1924

  26. [34]

    Sandoval, H

    L. Sandoval, H. M. Urbassek, and P. Entel, The Bain versus Nishiyama–Wassermann path in the martensitic transformation of Fe, New J. Phys.11, 103027 (2009)

    2009

  27. [35]

    Pitsch, The martensite transformation in thin foils of iron-nitrogen alloys, Philos

    W. Pitsch, The martensite transformation in thin foils of iron-nitrogen alloys, Philos. Mag.4, 577 (1959)

    1959

  28. [36]

    Nishiyama,Martensitic transformation(Elsevier, 2012)

    Z. Nishiyama,Martensitic transformation(Elsevier, 2012)

    2012

  29. [37]

    Kurdjumow and G

    G. Kurdjumow and G. Sachs, ¨Uber den mechanismus der stahlh¨ artung, Z. Phys.64, 325 (1930)

    1930

  30. [38]

    M. J. Buerger,Phase transformations in solids(John Wi- ley, 1951) pp. 183–211

    1951

  31. [39]

    Nga and C

    Y. Nga and C. Ong, Mechanism of pressure-induced poly- morphic transition in NaCl, Phys. Rev. B46, 10547 (1992)

    1992

  32. [40]

    Zahn, Molecular dynamics investigation of the pres- sure induced B1 to B2 phase transitions of RbBr, J

    D. Zahn, Molecular dynamics investigation of the pres- sure induced B1 to B2 phase transitions of RbBr, J. Solid State Chem.177, 3590 (2004)

    2004

  33. [41]

    Watanabe, M

    M. Watanabe, M. Tokonami, and N. Morimoto, The tran- sition mechanism between the CsCl-type and NaCl-type structures in CsCl, Acta Crystallogr. Sect. A33, 294 (1977)

    1977

  34. [42]

    Blaschko, G

    O. Blaschko, G. Ernst, G. Quittner, G. P´ epy, and M. Roth, Investigations of the NaCl-to-CsCl phase tran- sition in RbI by elastic diffuse and inelastic neutron scat- tering, Phys. Rev. B20, 1157 (1979)

    1979

  35. [43]

    Tol´ edano, K

    P. Tol´ edano, K. Knorr, L. Ehm, and W. Depmeier, Phe- nomenological theory of the reconstructive phase transi- tion between the NaCl and CsCl structure types, Phys. Rev. B67, 144106 (2003)

    2003

  36. [44]

    S. H. Tolbert and A. Alivisatos, The wurtzite to rock salt structural transformation in cdse nanocrystals under high pressure, J. Chem. Phys.102, 4642 (1995)

    1995

  37. [45]

    Sowa, On the transition from the wurtzite to the NaCl type, Acta Crystallogr

    H. Sowa, On the transition from the wurtzite to the NaCl type, Acta Crystallogr. Sect. A57, 176 (2001)

    2001

  38. [46]

    Wilson and P

    M. Wilson and P. A. Madden, Transformations between tetrahedrally and octahedrally coordinated crystals: the wurtzite→rocksalt and blende→rocksalt mechanisms, J. Phys.: Condens. Matter14, 4629 (2002)

    2002

  39. [47]

    M. S. Dresselhaus, G. Dresselhaus, and A. Jorio,Group theory: application to the physics of condensed matter (Springer Science & Business Media, 2007)

    2007

  40. [48]

    G. L. W. Hart and R. W. Forcade, Algorithm for gen- erating derivative structures, Phys. Rev. B77, 224115 (2008)

    2008

  41. [49]

    Axler,Linear algebra done right(Springer, 2015) pp

    S. Axler,Linear algebra done right(Springer, 2015) pp. 233–238

    2015

  42. [50]

    Jonker and A

    R. Jonker and A. Volgenant, A shortest augmenting path algorithm for dense and sparse linear assignment prob- lems, Computing38, 325 (1987)

    1987

  43. [51]

    D. F. Crouse, On implementing 2d rectangular assign- ment algorithms, IEEE Trans. Aerosp. Electron. Syst. 52, 1679 (2016)

    2016

  44. [52]

    D. J. Wales and J. P. Doye, Global optimization by basin- hopping and the lowest energy structures of lennard-jones clusters containing up to 110 atoms, J. Phys. Chem. A 101, 5111 (1997)

    1997

  45. [53]

    K. G. Murty, An algorithm for ranking all the assign- ments in order of increasing cost, Oper. Res.16, 682 (1968)

    1968

  46. [54]

    Seehttps://www.phy.pku.edu.cn/xzli/RESEARCH.htm, for the documentation and source code ofcrystmatch

  47. [55]

    K. S. Atsushi Togo and I. Tanaka, Spglib: a software library for crystal symmetry search, Sci. Technol. Adv. Mater., Meth.4, 2384822 (2024)

    2024

  48. [56]

    F. P. Bundy, H. P. Bovenkerk, H. M. Strong, and R. H. Wentorf Jr, Diamond-graphite equilibrium line from growth and graphitization of diamond, J. Chem. Phys.35, 383 (1961)

    1961

  49. [57]

    Irifune, A

    T. Irifune, A. Kurio, S. Sakamoto, T. Inoue, and H. Sumiya, Ultrahard polycrystalline diamond from graphite, Nature421, 599 (2003)

    2003

  50. [58]

    J. Wei, B. Feng, R. Ishikawa, T. Yokoi, K. Matsunaga, N. Shibata, and Y. Ikuhara, Direct imaging of atomistic grain boundary migration, Nat. Mater.20, 951 (2021)

    2021

  51. [59]

    Li, Y.-F

    J.-L. Li, Y.-F. Li, and Z.-P. Liu, Silicon orientations to grow semi-polar aln, Nat. Commun.16, 1 (2025)

    2025

  52. [60]

    Cohen,A course in computational algebraic num- ber theory, Graduate texts in mathematics, Vol

    H. Cohen,A course in computational algebraic num- ber theory, Graduate texts in mathematics, Vol. 130 (Springer, 2011)

    2011

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.