Pith. sign in

REVIEW 3 major objections 8 minor 1 cited by

Data-Driven Design-Test-Make-Analyze Paradigm for Inorganic Crystals: Ultrafast Synthesis of Ternary Oxides

T0 review · 3 major / 8 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 An integrated screening workflow produces a new oxide, ZnVO3, in minutes

desk verdict A credible end-to-end discovery pipeline with a central new-phase claim that currently outruns the diffraction evidence; worth refereeing, not accepting as-is. read the letter →

arxiv 2506.18542 v1 pith:MBSWVIZC submitted 2025-06-23 cond-mat.mtrl-sci

Haiwen Dai , Matthew J. McDermott , Andy Paul Chen , Jose Recatala-Gomez , Wei Nong , Ruiming Zhu , Maung Thway , Samuel Morris
show 14 more authors
Christian Schürmann Shreyas Dinesh Pethe Chenguang Zhang Wuan Geok Saw Bich Ngoc Tran Pritish Mishra Fengxia Wei Albertus Denny Handoko Sabrine Hachmioune Haipei Shao Ming Lin Chong Wai Liew Kristin A. Persson Kedar Hippalgaonkar
This is my paper · ORCID
classification cond-mat.mtrl-sci
keywords data-drivenmaterialsdiscoverydesign-test-make-analyzesynthesizabilitypredictionoxidationstateprobabilityreactionnetworksynthesisplanningultrafastmetal-deficientspinelmicrocrystalelectrondiffraction
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 establish that a fully computational screening pipeline can select previously unsynthesized oxide compositions and that rapid resistive heating can realize them within minutes. The pipeline filters candidate $\mathrm{ABO_3}$ crystals in stages: a machine-learned synthesizability score, an oxidation-state probability check, reaction-network planning that ranks precursor sets by thermodynamic driving force and competition, and density-functional stability assessment. For the two surviving targets, the experiments produced a new metal-deficient spinel phase of $\mathrm{ZnVO_3}$ and, for the $\mathrm{YMoO_3}$ target, near-composition crystals later solved as $\mathrm{Y_4Mo_4O_{11}}$. If the workflow holds up, it closes a loop that usually breaks: computational predictions rarely reach experimental realization.

What carries the argument

The carrying mechanism is a staged filter chain whose output feeds an ultrafast synthesis-and-characterization loop. First, a machine-learned synthesizability score and an oxidation-state probability model remove candidates that are unlikely to form or cannot charge-balance. Second, reaction-network planning enumerates thousands of precursor reactions and ranks them by thermodynamic driving force and by competition scores that penalize routes where competing phases form. Third, density-functional stability checks, including a low-energy polymorph search, flag which surviving compositions are accessible. The making step heats a graphite-enclosed powder with a fast current ramp, reaching heating and cooling rates near $10^3$ K/s and completing synthesis in about two minutes. The analysis step combines full-pattern X-ray refinement with site-occupancy statistics, thermogravimetric oxygen content, electron-microscopy composition maps, and microcrystal electron diffraction for ab initio structure solution inside a powder mixture. The decisive ingredient is partial disorder: only after a disordered spinel supercell was included did $\mathrm{ZnVO_3}$ become hull-stable.

What would settle it

Collect neutron powder diffraction, or resonant X-ray diffraction at the Zn and V absorption edges, on the same five samples and refine the octahedral site occupancy. If the site converges to the fully occupied 25% Zn / 75% V pattern of $\mathrm{Zn_3V_3O_8}$ or to $\mathrm{ZnV_2O_4}$, rather than to an approximately 20% Zn / 68% V occupancy with vacancies, the new-phase claim is disproved.

Watch

Extended reading notes

Core claim

The paper's central claim is that $\mathrm{ZnVO_3}$, a composition never reported in the Zn–V–O system, can be made by the predicted precursor route and crystallizes as a metal-deficient spinel. Full-pattern X-ray refinement of five samples gives an average stoichiometry near $\mathrm{Zn_{2.74}V_{2.74}O_{7.95}}$, with the octahedral cation site partially occupied and the tetrahedral site nearly full; thermogravimetric oxygen analysis, electron microscopy, and X-ray fluorescence support the same composition. A density-functional supercell containing the observed partial disorder lies on the ternary convex hull, while ordered polymorphs of $\mathrm{ZnVO_3}$ sit above it, so the authors argue the disorder is what makes the phase stable. For the second target, $\mathrm{YMoO_3}$, the experiments produced rod-shaped crystals with elemental ratio close to 1:1:3, solved by microcrystal electron diffraction as $\mathrm{Y_4Mo_4O_{11}}$. The paper presents both outcomes as evidence that an end-to-end design–test–make–analyze loop can deliver new inorganic phases.

Load-bearing premise

The claim that the synthesized powder is a genuinely new metal-deficient $\mathrm{ZnVO_3}$ phase, rather than a known zinc vanadium spinel with subtly different occupancies, rests on small differences in X-ray refinement quality and site occupancies; the paper itself reports that the $\mathrm{ZnVO_3}$ model fit only marginally better than $\mathrm{ZnV_2O_4}$.

Editorial extensions

If this is right

  • A composition that database tags had never marked as experimentally known can be realized when synthesizability, charge balance, reaction selectivity, and thermodynamic stability are screened together rather than formation energy alone.
  • The metal-deficient spinel $\mathrm{ZnVO_3}$ adds a new stable point to the Zn–V–O ternary phase diagram, so future phase-equilibrium and convex-hull calculations in this system should include the disordered phase.
  • Ultrafast resistive heating reproduces the same phase as conventional furnace synthesis in about two minutes, which makes temperature-series screening across compositions practical at high throughput.
  • Microcrystal electron diffraction can solve the structure of an unknown crystal in a powder mixture, such as $\mathrm{Y_4Mo_4O_{11}}$, even when competing phases dominate the bulk pattern.
  • Ordered-only computational screening left $\mathrm{ZnVO_3}$ metastable; adding partial disorder changed the stability ranking, indicating that disorder-aware screening could rescue other dismissed predictions.

Reading between the lines

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

  • A testable extension the authors do not pursue: re-running the same filtering pipeline on previously rejected candidates with disorder-aware supercells could reveal that many 'too unstable' predictions actually have low-energy partially disordered forms.
  • The metal vacancies in spinel $\mathrm{ZnVO_3}$ may give it functional behavior, such as reversible zinc or lithium storage or catalytic activity, because octahedral-site vacancies are a known design handle in such oxides; this is an inference, not a result of the paper.
  • The $\mathrm{YMoO_3}$ near-miss implies that reaction products whose composition sits close to the target should be treated as discovery output rather than failure, and that a workflow explicitly searching the local composition neighborhood around each target could surface phases missed by standard convex-hull scans.
  • Because the oxidation-state filter uses only composition, the synthesizability-times-oxidation-probability ranking could be transferred to sulfide, selenide, or pnictide spaces without retraining the charge-balance component; the paper reports no such test.
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 / 8 minor

Summary. The paper presents an end-to-end 'Design–Test–Make–Analyze' (DTMA) workflow for inorganic materials discovery, combining machine-learning synthesizability scores, oxidation-state probability filtering, thermodynamic reaction-network synthesis planning, DFT stability assessment, ultrafast Joule-heating synthesis, and detailed structural/compositional characterization. The workflow is applied to the ABO3 composition space; two candidates, ZnVO3 and YMoO3, pass the computational filters and are taken to experiment. For ZnVO3, the authors claim successful synthesis of a previously unreported metal-deficient spinel phase, supported by XRD refinement, TGA, EDX/XRF, ESR, and post-hoc DFT convex-hull calculations. For YMoO3, synthesis instead yields YMoO3-x and Y4Mo4O11, with a rod-like crystal assigned a new non-centrosymmetric Pba2 polymorph by microcrystal electron diffraction (microED). The paper argues that the integrated pipeline can identify promising candidates and refine the local phase landscape around them, marking a step toward closed-loop inorganic materials design.

Significance. If the ZnVO3 phase claim holds, the paper would demonstrate a genuinely new ternary oxide phase discovered through a computational pipeline with experimental feedback, and the DTMA loop would be a useful template for closed-loop materials discovery. The work has notable strengths: the computational and experimental data are openly shared via Zenodo and GitHub; the synthesis is genuinely fast (heating/cooling rates ~10^3 K/s); microED is used to solve a structure from an impure powder mixture; and the authors honestly acknowledge the challenge of partial disorder, which affects a large fraction of real materials. The YMoO3 case, while not yielding the target phase, is valuable as an illustration of how competing phases dominate and how microED can identify products in mixtures. However, the significance is conditional on the ZnVO3 assignment: the Rietveld discrimination against the known Zn3V3O8 spinel is reported only as 'marginally better' with no statistical test, and the DFT 'validation' is post-hoc because the observed disordered structure was added to the convex hull after synthesis. The Y4Mo4O11 polymorph claim is also under-supported without refinement statistics.

major comments (3)
  1. [Case Study 1 / Figure S10] The central claim that ZnVO3 is a new metal-deficient spinel rests on Rietveld discrimination among Zn-V-O spinels, but the manuscript reports only that the refinement for ZnVO3 was 'marginally better' than that for ZnV2O4 and noticeably better than those for Zn3V3O8 and Zn2VO4, without giving Rwp, GoF, or a Hamilton test. Because Zn3V3O8 has the same Zn:V ratio, the XRF and EDX data cannot exclude it, and the refined octahedral occupancies are strongly correlated with thermal-displacement and peak-shape parameters in laboratory XRD. The paper's own Discussion states that existing diffraction and elemental analysis have 'limited accuracy especially for disordered materials,' which is precisely the regime at issue. Without a significance test separating the proposed ZnVO3 model from Zn3V3O8, the new-phase claim is not established.
  2. [Case Study 1, Fig. 4D-E; Methods: DFT calculations] The DFT stability argument for disordered ZnVO3 is post-hoc: the disordered supercell was constructed from the refined site occupancies of the experimental sample and added to the convex hull only after synthesis. The conclusion that the observed phase is 'thermodynamically stable ... as predicted by our in-silico approach' conflates consistency with prediction; the DFT calculation cannot serve as independent validation of the phase assignment. The manuscript should be revised to present this step as an energy assessment of an experimentally discovered structure, not as a validation of the screening pipeline.
  3. [Case Study 2 / Fig. 5G-I] The claim of a new non-centrosymmetric Y4Mo4O11 polymorph (Pba2) is under-supported: no refinement residuals, completeness, or data quality statistics are reported for the microED structure solution, no quantitative test is given for rejecting the centrosymmetric Pbam polymorph, and the composition mismatch (EDX gives a ratio close to Y:Mo:O = 1:1:3 rather than 4:4:11) is acknowledged but left unresolved. As written, the evidence does not rule out the known Pbam Y4Mo4O11 structure with minor atomic displacement, and the 'new polymorph' claim goes beyond what the data demonstrate.
minor comments (8)
  1. [Abstract] In the abstract, 'confirmed the successful synthesis ZnVO3' is missing the word 'of'; please correct.
  2. [Case Study 1 text] The sentence 'The refinement for ZnVO3 was marginally better than that for ZnV2O4 and noticeably better than those for Zn3V3O8 and ZnV2O4' appears to contain a typo: the second 'ZnV2O4' should likely be 'Zn2VO4'.
  3. [Figure 2 and caption] The main text refers to 'Figure 2A, lower plot' for the YMoO3 combination-reaction energy, but the relevant plot appears to be in Figure 2C; the figure-number references should be checked and corrected.
  4. [Table S5] In Table S5(C), the 'Standard deviation' row appears to replicate values from other rows and is inconsistent with the listed individual measurements; the statistics should be recomputed and corrected.
  5. [Figure S6 caption] The caption of Figure S6 says 'XRD analysis towards ZnVO3 synthesis' but the content concerns YMoO3; the caption should be corrected.
  6. [References] References 55 and 56 are duplicate entries for the same paper (Gehringer, Friák, Holec, Comput. Phys. Commun. 286, 108664, 2023); one should be removed.
  7. [Note S1 / Table 1] The OSP normalization procedure described in Note S1 (dividing by the highest probability in the same chemical system) is essential for interpreting the ranking in Table 1 and should be described in the main text where the OSP model is first used.
  8. [Introduction / Case Study 2] The statement that two 'previously unsynthesized target compositions' were considered is stronger than the results support, because stoichiometric YMoO3 was not obtained; the product was YMoO3-x / Y4Mo4O11. Please adjust the wording to distinguish target compositions from final products.

Circularity Check

1 steps flagged · score 3.0 of 10

Post-hoc DFT stabilization of the experimentally fitted disordered ZnVO3 is presented as an in-silico prediction, while the experimental core remains externally validated.

  1. fitted input called prediction [Case Study 1, Figure 4D-4E and Methods 'DFT calculations']
    "We further constructed supercells of partially disordered spinel ZnVO3 and Zn3V3O8 structures for further DFT calculations. ... ZnVO3, previously considered metastable in an ordered P21/c structure during preliminary DFT screening (Figure S2), now resides on the hull surface. ... confirming the successful synthesis of thermodynamically stable ZnVO3, crystallized in a metal-deficient spinel structure, as predicted by our in-silico approach."

    The 'disordered spinel ZnVO3' inserted into the convex hull is not an independent first-principles prediction: its cell and occupancies (20.38% Zn1, 68.52% V; Table S6) come from Rietveld refinement of the synthesized powder. Building the DFT supercell from those fitted occupancies and then reporting that the same phase 'now resides on the hull surface' is a self-consistency check, not an external validation. The initial in-silico screen had predicted only an ordered P21/c polymorph (Ehull = 54 meV/atom) and explicitly stated that disordered structures were not included in the original hull construction.

full rationale

The paper's central demonstration is experimental: ZnVO3 and YMoO3 were selected through a computational screen and subjected to ultrafast synthesis, with products analyzed by XRD, TGA, XRF/EDX, TEM, and microED. That pipeline is not circular: the SC and OSP models are trained on external Materials Project/ICSD data and are evaluated with conventional precision/recall metrics; the reaction-network planning uses independent thermodynamic data; and the synthesis outcomes are external facts not encoded in the filters. Self-citations to the authors' earlier models are load-bearing for tool choice but not for the success claim, so they do not by themselves raise the score. The one genuinely circular-looking move is the DFT 'validation' of disordered ZnVO3: the supercell is constructed from occupancies fitted in the Rietveld refinement, placed on the convex hull, and then described as 'predicted by our in-silico approach.' That step is post-hoc self-consistency, not an independent test, and should not be quoted as evidence that the spinel assignment is correct. The same post-hoc pattern appears in the Y-Mo-O case, where experimentally encountered Y2.5MoO6 and Y4Mo4O11 are added to the hull after the fact. The separate question of whether the Rietveld fit truly rules out Zn3V3O8 (the paper reports only a 'marginally better' refinement with no significance test) is a correctness and evidence risk, which is outside circularity under the review rules. Overall, the experimental core is externally anchored, so the paper is only mildly circular in its validation rhetoric.

Assumptions & free parameters 4 free parameters · 6 assumptions · 2 invented entities

The screening pipeline depends on several hand-picked thresholds and weights, on the prior models (SC, OSP) that are used as filters, and on the assumption that DFT formation energies and thermodynamic reaction rankings at 650 C sufficiently capture synthesis accessibility. The disordered-phase DFT validation rests on the assumption that a single quasirandom supercell represents the real disorder. No ad hoc force or particle is introduced; the new phases are experimental outputs.

free parameters (4)
  • SC threshold = 0.3 (score > 0.3)
    Chosen to balance recall/precision; the paper states 'minimizes the exclusion of viable candidates'.
  • OSP threshold = 0.2
    Chosen to yield high recall (92%); no first-principles justification.
  • Reaction cost weights = 0.1 (dG), 0.45 (C1), 0.45 (C2)
    Weights in the cost function are hand-picked in Methods; they determine which reaction pathways are ranked top.
  • Gibbs free energy evaluation temperature = 650 C
    Chosen for the ML Gibbs descriptor; affects reaction costs.
assumptions (6)
  • domain assumption Materials Project GGA+U formation energies correctly rank stability among ABO3 candidates.
    Used to compute Ehull and reaction energies; no independent experimental check for the specific systems.
  • domain assumption Synthesizability score is a valid proxy for experimental accessibility.
    The SC model is trained on ICSD labels; its transferability to novel compositions is assumed.
  • domain assumption Oxidation-state probability ranking is meaningful for charge neutrality screening.
    OSP from arXiv preprint (ref 16); the normalized probability is used as a filter.
  • domain assumption Reaction network energies at 650 C predict synthesis selectivity despite neglected kinetics.
    Acknowledged in Discussion: 'fall short when reaction kinetics are considered'.
  • domain assumption The quasirandom supercell generated by sqsgenerator faithfully represents the observed partial disorder.
    Used for the DFT stability claim in Figure 4D; no convergence test over multiple disorder realizations is shown.
  • domain assumption XRD Rietveld refinement with a constrained Zn:V ratio uniquely identifies the spinel phase.
    The discrimination against ZnV2O4 is 'marginally better', so this assumption is load-bearing and not fully satisfied.
invented entities (2)
  • Metal-deficient spinel ZnVO3 phase independent evidence
    purpose: Target oxide claimed as newly synthesized; its identification is the paper's central result.
    Characterized by XRD, TGA, EDX, XRF, ESR, and post-hoc DFT hull placement; falsifiable by independent synthesis or property measurements.
  • Y4Mo4O11 Pba2 non-centrosymmetric polymorph
    purpose: Structural assignment for rod-like YMoO3-like crystals; claimed to differ from known Pbam polymorph by loss of centrosymmetry.
    Based solely on microED solution without published R-factors or completeness; paper admits purification difficulties and that further characterization is desirable.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Data-Driven Design-Test-Make-Analyze Paradigm for Inorganic Crystals: Ultrafast Synthesis of Ternary Oxides." pith.science (2026). https://pith.science/paper/MBSWVIZC

@misc{pith2026250618542,
  author       = {Pith},
  title        = {Pith review of: Data-Driven Design-Test-Make-Analyze Paradigm for Inorganic Crystals: Ultrafast Synthesis of Ternary Oxides},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/MBSWVIZC}},
  note         = {Machine review of arXiv:2506.18542}
}
read the original abstract

Data-driven methodologies hold the promise of revolutionizing inorganic materials discovery, but they often face challenges due to discrepancies between theoretical predictions and experimental validation. In this work, we present an end-to-end discovery framework that leverages synthesizability, oxidation state probability, and reaction pathway calculations to guide the exploration of transition metal oxide spaces. Two previously unsynthesized target compositions, ZnVO3 and YMoO3, passed preliminary computational evaluation and were considered for ultrafast synthesis. Comprehensive structural and compositional analysis confirmed the successful synthesis ZnVO3 in a partially disordered spinel structure, validated via Density Functional Theory (DFT). Exploration of YMoO3 led to YMoO3-x with elemental composition close to 1:1:3; the structure was subsequently identified to be Y4Mo4O11 through micro-electron diffraction (microED) analysis. Our framework effectively integrates multi-aspect physics-based filtration with in-depth characterization, demonstrating the feasibility of designing, testing, synthesizing, and analyzing (DTMA) novel material candidates, marking a significant advancement towards inorganic materials by design.

Figures

Figures reproduced from arXiv: 2506.18542 by the authors.

Figure 5
Figure 5. XRD analysis of ultrafast synthesis products targeting YMoO3 via synthesis planned ‘Na2CO3 + YCl3 + MoO2’ reaction with detailed characterizations of rod-like Y4Mo4O11 with elemental ratio close to 1:1:3. (A) Temperature series of XRD patterns, (B) Phase evolution resolved via XRD screening, and (C) Ternary phase diagram overlayed with competing phases observed, acquired from the Materials Project; (D) HR-TEM with i… view at source ↗

Review history (2 revisions) →

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. Dis-GEN: Disordered crystal structure generation

    cond-mat.mtrl-sci 2025-07 conditional novelty 6.0 of 10

    Dis-GEN learns to represent and generate disordered inorganic crystal structures with partial atomic occupancy and vacancies while preserving crystallographic symmetry.

Reference graph

Works this paper leans on

63 extracted references · 37 canonical work pages · cited by 1 Pith paper

  1. [1]

    Zunger, A. (2018). Inverse design in search of materials with target functionalities. Nat Rev Chem 2, 0121. https://doi.org/10.1038/s41570-018-0121

    doi:10.1038/s41570-018-0121 2018

  2. [2]

    Zeni, C., Pinsler, R., Zü gner, D., Fowler, A., Horton, M., Fu, X., Wang, Z., Shysheya, A., Crabbé , J., Ueda, S., et al. (2025). A generative model for inorganic materials design. Nature 639, 624–632. https://doi.org/10.1038/s41586-025-08628-5

    doi:10.1038/s41586-025-08628-5 2025

  3. [3]

    Merchant, A., Batzner, S., Schoenholz, S.S., Aykol, M., Cheon, G., and Cubuk, E.D. (2023). Scaling deep learning for materials discovery. Nature 624, 80 –85. https://doi.org/10.1038/s41586-023-06735-9

    doi:10.1038/s41586-023-06735-9 2023

  4. [4]

    Zhu, R., Nong, W., Yamazaki, S., and Hippalgaonkar, K. (2024). WyCryst: Wyckoff inorganic crystal generator framework. Matter, S2590238524003059. https://doi.org/10.1016/j.matt.2024.05.042

    doi:10.1016/j.matt.2024.05.042 2024

  5. [5]

    Szymanski, N.J., Rendy, B., Fei, Y., Kumar, R.E., He, T., Milsted, D., McDermott, M.J., Gallant, M., Cubuk, E.D., Merchant, A., et al. (2023). An autonomous laboratory for the accelerated synthesis of novel materials. Nature 624, 86–91. https://doi.org/10.1038/s41586- 023-06734-w

    doi:10.1038/s41586- 2023

  6. [6]

    -C., et al

    Montoya, J.H., Grimley, C., Aykol, M., Ophus, C., Sternlicht, H., Savitzky, B.H., Minor, A.M., Torrisi, S.B., Goedjen, J., Chung, C. -C., et al. (2024). How the AI -assisted discovery and synthesis of a ternary oxide highlights capability gaps in material s science. Chem. Sci., 10.1039.D3SC04823C. https://doi.org/10.1039/D3SC04823C

    doi:10.1039/d3sc04823c 2024

  7. [7]

    Jang, J., Noh, J., Zhou, L., Gu, G.H., Gregoire, J.M., and Jung, Y. (2024). Synthesizability of materials stoichiometry using semi -supervised learning. Matter 7, 2294 –2312. https://doi.org/10.1016/j.matt.2024.05.002. Page 28 of 53

    doi:10.1016/j.matt.2024.05.002 2024

  8. [8]

    Leeman, J., Liu, Y., Stiles, J., Lee, S.B., Bhatt, P., Schoop, L.M., and Palgrave, R.G. (2024). Challenges in High -Throughput Inorganic Materials Prediction and Autonomous Synthesis. PRX Energy 3, 011002. https://doi.org/10.1103/PRXEnergy.3.011002

    doi:10.1103/prxenergy.3.011002 2024

Show all 63 references
  1. [9]

    Zhu, R., Tian, S.I.P., Ren, Z., Li, J., Buonassisi, T., and Hippalgaonkar, K. (2023). Predicting Synthesizability using Machine Learning on Databases of Existing Inorganic Materials. ACS Omega 8, 8210–8218. https://doi.org/10.1021/acsomega.2c04856

    2023 doi

  2. [10]

    Zeni, C., Pinsler, R., Zü gner, D., Fowler, A., Horton, M., Fu, X., Shysheya, S., Crabbé , J., Sun, L., Smith, J., et al. (2024). MatterGen: a generative model for inorganic materials design. Preprint at arXiv

    2024

  3. [11]

    Gu, G.H., Jang, J., Noh, J., Walsh, A., and Jung, Y. (2022). Perovskite synthesizability using graph neural networks. npj Comput Mater 8, 71. https://doi.org/10.1038/s41524-022-00757-z

    2022 doi

  4. [12]

    Aykol, M., Dwaraknath, S.S., Sun, W., and Persson, K.A. (2018). Thermodynamic limit for synthesis of metastable inorganic materials. Sci. Adv. 4, eaaq0148. https://doi.org/10.1126/sciadv.aaq0148

    2018 doi

  5. [13]

    Bartel, C.J., Trewartha, A., Wang, Q., Dunn, A., Jain, A., and Ceder, G. (2020). A critical examination of compound stability predictions from machine -learned formation energies. npj Comput Mater 6, 97. https://doi.org/10.1038/s41524-020-00362-y

    2020 doi

  6. [14]

    Zou, H., Zhao, H., Lu, M., Wang, J., Deng, Z., and Wang, J. (2025). Predicting thermodynamic stability of inorganic compounds using ensemble machine learning based on electron configuration. Nat Commun 16, 203. https://doi.org/10.1038/s41467-024-55525-y

    2025 doi

  7. [15]

    Davies, D.W., Butler, K.T., Isayev, O., and Walsh, A. (2018). Materials discovery by chemical analogy: role of oxidation states in structure prediction. Faraday Discuss. 211, 553 –568. https://doi.org/10.1039/C8FD00032H

    2018 doi

  8. [16]

    Thway, M., Chen, A.P., Dai, H., Recatala-Gomez, J., Tian, S.I.P., Zhu, R., Zhai, W., Wei, F., Repaka, D.V.M., Buonassisi, T., et al. (2024). Exploring material compositions for synthesis using oxidation states. Preprint at arXiv

    2024

  9. [17]

    Szymanski, N.J., Nevatia, P., Bartel, C.J., Zeng, Y., and Ceder, G. (2023). Autonomous and dynamic precursor selection for solid -state materials synthesis. Nat Commun 14, 6956. https://doi.org/10.1038/s41467-023-42329-9

    2023 doi

  10. [18]

    Karpovich, C., Jensen, Z., Venugopal, V., and Olivetti, E. (2021). Inorganic Synthesis Reaction Condition Prediction with Generative Machine Learning. Preprint at arXiv

    2021

  11. [19]

    Kim, E., Jensen, Z., van Grootel, A., Huang, K., Staib, M., Mysore, S., Chang, H.-S., Strubell, E., McCallum, A., Jegelka, S., et al. (2020). Inorganic Materials Synthesis Planning with Literature-Trained Neural Networks. J. Chem. Inf. Model. 60, 1194 –1201. https://doi.org/10...

    2020 doi

  12. [20]

    Kim, E., Huang, K., Jegelka, S., and Olivetti, E. (2017). Virtual screening of inorganic materials synthesis parameters with deep learning. npj Comput Mater 3, 53. https://doi.org/10.1038/s41524-017-0055-6

    2017 doi

  13. [21]

    He, T., Huo, H., Bartel, C.J., Wang, Z., Cruse, K., and Ceder, G. (2023). Precursor recommendation for inorganic synthesis by machine learning materials similarity from scientific literature. Sci. Adv. 9, eadg8180. https://doi.org/10.1126/sciadv.adg8180

    2023 doi

  14. [22]

    Wen, M., Spotte -Smith, E.W.C., Blau, S.M., McDermott, M.J., Krishnapriyan, A.S., and Persson, K.A. (2023). Chemical reaction networks and opportunities for machine learning. Nat Comput Sci 3, 12–24. https://doi.org/10.1038/s43588-022-00369-z

    2023 doi

  15. [23]

    McDermott, M.J., Dwaraknath, S.S., and Persson, K.A. (2021). A graph -based network for predicting chemical reaction pathways in solid -state materials synthesis. Nat Commun 12,

    2021

  16. [24]

    Todd, P.K., McDermott, M.J., Rom, C.L., Corrao, A.A., Denney, J.J., Dwaraknath, S.S., Khalifah, P.G., Persson, K.A., and Neilson, J.R. (2021). Selectivity in Yttrium Manganese Oxide Synthesis via Local Chemical Potentials in Hyperdimensional Phase Spac e. J. Am. Chem. Soc. 143...

    2021 doi

  17. [25]

    Neilson, J.R., McDermott, M.J., and Persson, K.A. (2023). Modernist materials synthesis: Finding thermodynamic shortcuts with hyperdimensional chemistry. Journal of Materials Research 38, 2885–2893. https://doi.org/10.1557/s43578-023-01037-2

    2023 doi

  18. [26]

    Dai, H., Dimitriadou, S., Sankara Rama Krishnan, P.S., Handoko, A.D., Recatala -Gomez, J., Wang, Y., Repaka, D.V.M., Thway, M., Zhang, C., Duchamp, M., et al. (2023). Sub -10 nm Mixing and Alloying of Cu–Ag and Cu–Ni via Accelerated Solid Diffusion. ACS Appl. Mater. Interfaces...

    2023 doi

  19. [27]

    Kitchen, H.J., Vallance, S.R., Kennedy, J.L., Tapia -Ruiz, N., Carassiti, L., Harrison, A., Whittaker, A.G., Drysdale, T.D., Kingman, S.W., and Gregory, D.H. (2014). Modern Microwave Methods in Solid -State Inorganic Materials Chemistry: From Fundamental s to Manufacturing. Ch...

    2014 doi

  20. [28]

    Wang, C., Ping, W., Bai, Q., Cui, H., Hensleigh, R., Wang, R., Brozena, A.H., Xu, Z., Dai, J., Pei, Y., et al. (2020). A general method to synthesize and sinter bulk ceramics in seconds. Science 368, 521–526. https://doi.org/10.1126/science.aaz7681

    2020 doi

  21. [29]

    Oliynyk, A.O., Antono, E., Sparks, T.D., Ghadbeigi, L., Gaultois, M.W., Meredig, B., and Mar, A. (2016). High-Throughput Machine-Learning-Driven Synthesis of Full-Heusler Compounds. Chem. Mater. 28, 7324–7331. https://doi.org/10.1021/acs.chemmater.6b02724

    2016 doi

  22. [30]

    Montoya, J., Grimley, C., Aykol, M., Ophus, C., Sternlicht, H., Savitzky, B.H., Minor, A.M., Torrisi, S., Goedjen, J., Chung, C.-C., et al. (2023). Computer-assisted discovery and rational synthesis of ternary oxides (Chemistry) https://doi.org/10.26434/chemrxiv-2023-n4pz9. Pa...

    2023 doi

  23. [31]

    Haynes, A.S., Stoumpos, C.C., Chen, H., Chica, D., and Kanatzidis, M.G. (2017). Panoramic Synthesis as an Effective Materials Discovery Tool: The System Cs/Sn/P/Se as a Test Case. J. Am. Chem. Soc. 139, 10814–10821. https://doi.org/10.1021/jacs.7b05423

    2017 doi

  24. [32]

    Yang, T., Willhammar, T., Xu, H., Zou, X., and Huang, Z. (2022). Single -crystal structure determination of nanosized metal –organic frameworks by three -dimensional electron diffraction. Nat Protoc 17, 2389–2413. https://doi.org/10.1038/s41596-022-00720-8

    2022 doi

  25. [33]

    Peñ a, M.A., and Fierro, J.L.G. (2001). Chemical Structures and Performance of Perovskite Oxides. Chem. Rev. 101, 1981–2018. https://doi.org/10.1021/cr980129f

    2001 doi

  26. [34]

    Li, Y., Zhu, R., Wang, Y., Feng, L., and Liu, Y. (2023). Center -environment deep transfer machine learning across crystal structures: from spinel oxides to perovskite oxides. npj Comput Mater 9, 109. https://doi.org/10.1038/s41524-023-01068-7

    2023 doi

  27. [35]

    Zhao, J., and Wang, X. (2022). Screening Perovskites from ABO3 Combinations Generated by Constraint Satisfaction Techniques Using Machine Learning. ACS Omega 7, 10483–10491. https://doi.org/10.1021/acsomega.2c00002

    2022 doi

  28. [36]

    Shetty, V., Shedthi B, S., and Kumaraswamy, J. (2022). Predicting the thermodynamic stability of perovskite oxides using multiple machine learning techniques. Materials Today: Proceedings 52, 457–461. https://doi.org/10.1016/j.matpr.2021.09.208

    2022 doi

  29. [37]

    Talapatra, A., Uberuaga, B.P., Stanek, C.R., and Pilania, G. (2021). A Machine Learning Approach for the Prediction of Formability and Thermodynamic Stability of Single and Double Perovskite Oxides. Chem. Mater. 33, 845 –858. https://doi.org/10.1021/acs.chemmater.0c03402

    2021 doi

  30. [38]

    Liu, H., Cheng, J., Dong, H., Feng, J., Pang, B., Tian, Z., Ma, S., Xia, F., Zhang, C., and Dong, L. (2020). Screening stable and metastable ABO 3 perovskites using machine learning and the materials project. Computational Materials Science 177, 109614. https://doi.org/10.1016...

    2020

  31. [39]

    Jain, A., Ong, S.P., Hautier, G., Chen, W., Richards, W.D., Dacek, S., Cholia, S., Gunter, D., Skinner, D., Ceder, G., et al. (2013). Commentary: The Materials Project: A materials genome approach to accelerating materials innovation. APL Materials 1, 011002. https://doi.org/1...

    2013 doi

  32. [40]

    Damewood, J., Karaguesian, J., Lunger, J.R., Tan, A.R., Xie, M., Peng, J., and Gó mez - Bombarelli, R. (2023). Representations of Materials for Machine Learning. Annu. Rev. Mater. Res. 53, 399–426. https://doi.org/10.1146/annurev-matsci-080921-085947

    2023 doi

  33. [41]

    Chen, J., Cross, S.R., Miara, L.J., Cho, J.-J., Wang, Y., and Sun, W. (2024). Navigating phase diagram complexity to guide robotic inorganic materials synthesis. Nat. Synth. https://doi.org/10.1038/s44160-024-00502-y

    2024 doi

  34. [42]

    McDermott, M.J., McBride, B.C., Regier, C.E., Tran, G.T., Chen, Y., Corrao, A.A., Gallant, M.C., Kamm, G.E., Bartel, C.J., Chapman, K.W., et al. (2023). Assessing Thermodynamic Page 31 of 53 Selectivity of Solid-State Reactions for the Predictive Synthesis of Inorganic Materia...

    2023 doi

  35. [43]

    Vitoux, L., Guignard, M., Penin, N., Carlier, D., Darriet, J., and Delmas, C. (2020). NaMoO 2: a Layered Oxide with Molybdenum Clusters. Inorg. Chem. 59, 4015 –4023. https://doi.org/10.1021/acs.inorgchem.9b03688

    2020 doi

  36. [44]

    Vitoux, L., Guignard, M., Suchomel, M.R., Pramudita, J.C., Sharma, N., and Delmas, C. (2017). The NaxMoO2 Phase Diagram (1/2≤x< 1): An Electrochemical Devil’s Staircase. Chem. Mater. 29, 7243–7254. https://doi.org/10.1021/acs.chemmater.7b01834

    2017 doi

  37. [45]

    Barker, J. (2003). Lithium insertion properties of the layered LiMoO2 (R3m) made by a novel carbothermal reduction method. Solid State Ionics 158, 261 –267. https://doi.org/10.1016/S0167-2738(02)00875-5

    2003 doi

  38. [46]

    Todd, P.K., and Neilson, J.R. (2019). Selective Formation of Yttrium Manganese Oxides through Kinetically Competent Assisted Metathesis Reactions. J. Am. Chem. Soc. 141, 1191–

    2019

  39. [47]

    Sun, W., Dacek, S.T., Ong, S.P., Hautier, G., Jain, A., Richards, W.D., Gamst, A.C., Persson, K.A., and Ceder, G. (2016). The thermodynamic scale of inorganic crystalline metastability. Sci. Adv. 2, e1600225. https://doi.org/10.1126/sciadv.1600225

    2016 doi

  40. [48]

    Yi, H., Zuo, C., Ren, H., Zhao, W., Wang, Y., Ding, S., Li, Y., Qin, R., Zhou, L., Yao, L., et al. (2021). Structure evolution and energy storage mechanism of Zn 3 V 3 O 8 spinel in aqueous zinc batteries. Nanoscale 13, 14408–14416. https://doi.org/10.1039/D1NR02347K

    2021 doi

  41. [49]

    Gall, P., Barrier, N., Gautier, R., and Gougeon, P. (2002). Synthesis, Structural Trends, and Physical and Electronic Properties of the Reduced Molybdenum Oxides R 4Mo4O11 (R = Nd−Tm and Y) Containing Infinite Chains of Trans -Edge-Shared Octahedral Clusters. Inorg. Chem. 41, ...

    2002 doi

  42. [50]

    Wang, Y., Lv, J., Zhu, L., and Ma, Y. (2012). CALYPSO: A method for crystal structure prediction. Computer Physics Communications 183, 2063 –2070. https://doi.org/10.1016/j.cpc.2012.05.008

    2012 doi

  43. [51]

    Kresse, G., and Furthmü ller, J. (1996). Efficient iterative schemes for ab initio total-energy calculations using a plane -wave basis set. Phys. Rev. B 54, 11169 –11186. https://doi.org/10.1103/PhysRevB.54.11169

    1996 doi

  44. [52]

    Kresse, G., and Joubert, D. (1999). From ultrasoft pseudopotentials to the projector augmented- wave method. Phys. Rev. B 59, 1758–1775. https://doi.org/10.1103/PhysRevB.59.1758

    1999 doi

  45. [53]

    Perdew, J.P., Burke, K., and Ernzerhof, M. (1996). Generalized Gradient Approximation Made Simple. Phys. Rev. Lett. 77, 3865–3868. https://doi.org/10.1103/PhysRevLett.77.3865

    1996 doi

  46. [54]

    Page 32 of 53

    BaseMPR2SCANRelaxSet https://github.com/materialsproject/atomate2/blob/main/src/atomate2/vasp/sets/BaseMPR2S CANRelaxSet.yaml. Page 32 of 53

  47. [56]

    Gehringer, D., Friá k, M., and Holec, D. (2023). Models of configurationally-complex alloys made simple. Computer Physics Communications 286, 108664. https://doi.org/10.1016/j.cpc.2023.108664

    2023

  48. [57]

    Bartel, C.J., Millican, S.L., Deml, A.M., Rumptz, J.R., Tumas, W., Weimer, A.W., Lany, S., Stevanović, V., Musgrave, C.B., and Holder, A.M. (2018). Physical descriptor for the Gibbs energy of inorganic crystalline solids and temperature -dependent mater ials chemistry. Nat Com...

    2018 doi

  49. [60]

    If the structure is not already relaxed, perform DFT relaxation to obtain a meaningful input

    Start with a CIF file of the candidate compound. If the structure is not already relaxed, perform DFT relaxation to obtain a meaningful input

  50. [61]

    One can use the StructureMatcher() function from the pymatgen package to check whether the structure exists in MP (v2022.10.28)

    Confirm the structure is not already in the Materials Project database (optional). One can use the StructureMatcher() function from the pymatgen package to check whether the structure exists in MP (v2022.10.28). If it does, we recommend using a leave -one-out approach retraini...

  51. [62]

    Input the relaxed CIF file into the pretrained synthesizability model

    Predict synthesizability (SC). Input the relaxed CIF file into the pretrained synthesizability model. The output is a synthesizability score ranging from 0 to 100%. Model access and usage instructions are available on GitHub: https://github.com/Kedar-Materials-by-Design-Lab/De...

  52. [63]

    Input the chemical composition into the OSP model script

    Predict oxidation state probability (OSP). Input the chemical composition into the OSP model script. The model returns a raw probability, which should be normalized by the highest probability compound within the same chemical system. For example, if AB ₂O₄ has a raw probabilit...

  53. [64]

    GG +U”, r d d sh d i “metaGGA and r2SCAN

    Rank candidates. Multiply SC × OSP to obtain a final score between 0 and 100%, which can be used to prioritize candidate materials similar to Table 1 of the manuscript. Page 35 of 53 Figure S 1. Data -driven synthesis planning via reaction network for YMoO 3 showing the thermo...

    2023

  54. [1195]

    https://doi.org/10.1021/jacs.8b10123

    doi

  55. [3097]

    https://doi.org/10.1038/s41467-021-23339-x

    doi

Pith tools

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