Pith. sign in

REVIEW 3 major objections 5 minor 88 references

Discovery of Spin-Crossover Candidates with Equivariant Graph Neural Networks and Relevance-Based Classification

T0 review · 3 major / 5 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read This paper claims that a small equivariant graph neural network, trained on density functional theory spin-switching energies for 1,439 transition-metal complexes, can identify spin-crossover candidates with roughly four times the hit…

desk verdict Useful dataset, credible regression, but the four-fold enrichment claim is internally inconsistent. read the letter →

arxiv 2501.05341 v2 pith:2K43VCSZ submitted 2025-01-09 cond-mat.dis-nn cond-mat.mtrl-sci

Angel Albavera-Mata , Pawan Prakash , Jason B. Gibson , Eric Fonseca , Sijin Ren , Xiao-Guang Zhang , Hai-Ping Cheng , Michael Shatruk
show 2 more authors
S.B. Trickey Richard G. Hennig
This is my paper · ORCID
classification cond-mat.dis-nncond-mat.mtrl-sci
keywords spincrossoverequivariantgraphneuralnetworksmaterialsscreeningdensityfunctionaltheoryr2SCANrelevance-basedclassificationtransitionmetalcomplexes
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 claims that a small equivariant graph neural network can screen for spin-crossover candidates far more efficiently than random high-throughput searching. The authors assemble 1,439 mononuclear Cr, Mn, Fe, and Co complexes, compute their low-spin to high-spin energy differences with r2SCAN density functional theory, and train a 915-parameter graph network that predicts those energies with a test mean absolute error of 360 meV. They then convert the regression into a classifier using a relevance function centered on the candidate window $-200 \le \Delta E_{\mathrm{HL}} \le 500$ meV, obtaining a 56% posterior probability of finding a true candidate versus 13% for random picking, about 80% precision at 35% recall, and a 94% refusal probability that reduces redundant DFT calculations by roughly 17-fold. Applied to 11,356 additional complexes, the model predicts 1,076 potential spin-crossover materials, most of them iron-based.

What carries the argument

The load-bearing object is an equivariant graph neural network, a message-passing architecture equivariant to rotations, translations, and permutations, modified with an attention layer that feeds atom oxidation states and bond orders into the edges. A single convolutional layer keeps the model to 915 trainable parameters. The regression output is converted into a candidate classifier through a relevance function built from two sigmoids centered at the interval boundaries, following the precision-recall-for-regression method; Bayes' theorem then turns precision and recall into the reported posterior and refusal probabilities. A coordination-shell ablation, in which graphs are built from progressively larger shells centered on the transition metal versus on a random atom, serves as the diagnostic showing that the model learns the local dominance of the metal coordination environment.

What would settle it

Take a random sample of 50 to 100 of the 1,076 predicted candidates and recompute $\Delta E_{\mathrm{HL}}$ with a higher-level method such as CASPT2 or CCSD(T), or add zero-point corrections at the DFT level; if the fraction that still falls inside the $-200$ to $500$ meV window is far below the claimed about 80% precision, the screening gain is inflated. A smaller set of 20 to 30 synthesized and magnetically measured complexes would provide a direct test of the predictions.

Watch

Extended reading notes

Core claim

The paper's central claim is that a graph neural network with only 915 trainable parameters can learn the relationship between molecular structure and spin-switching energy well enough to act as a high-throughput pre-screen for spin crossover. Trained on r2SCAN DFT values of $\Delta E_{\mathrm{HL}}$ for 1,439 mononuclear complexes drawn from a crystallographic structure database, the network reaches a test mean absolute error of 360 meV and $R^2 = 0.88$. Converting the regression into a classification with a relevance function whose sigmoid thresholds sit at the candidate-window boundaries yields about 80% precision at 35% recall on the held-out set, a 56% posterior probability of finding a candidate versus 13% for random picking, and a refusal probability near 94% that corresponds to an approximate 17-fold reduction in redundant Kohn-Sham DFT calculations. On a broader set of 11,356 additional complexes, the model predicts 1,076 plausible spin-crossover candidates, with Fe-based species dominating the list.

Load-bearing premise

The load-bearing premise is that the r2SCAN density functional energies define ground truth: a complex counts as a spin-crossover candidate exactly when its $\Delta E_{\mathrm{HL}}$ falls between $-200$ and $500$ meV, and the paper itself concedes this window is debatable and that zero-point corrections were omitted. Any systematic bias in these labels is inherited by the 56% enrichment figure and the 1,076 predicted candidates.

Editorial extensions

If this is right

  • If the claimed enrichment holds, screening a large structure database can be reduced from tens of thousands of DFT calculations to a few hundred follow-up calculations on the predicted candidates.
  • The coordination-shell result implies that only the inner coordination environment, roughly five shells around the metal, carries the essential information, so future screening models could be built from smaller molecular fragments.
  • The dominance of Fe among the 1,076 predicted candidates matches the known prevalence of Fe(II) spin-crossover chemistry and suggests the model can rank ligand families for targeted synthesis.
  • The relevance threshold gives an adjustable operating point: tightening it trades recall for precision, so experimentalists can choose a stricter cutoff that yields a smaller, higher-confidence candidate list.

Reading between the lines

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

  • The 56% posterior is computed using the training-set prevalence as the prior; in a different chemical library with a different base rate of candidates, the posterior would shift even if precision and recall stayed the same.
  • Because the label window excludes zero-point and thermal contributions, the 1,076 candidates should be read as "worth a higher-level or experimental check" rather than confirmed spin-crossover materials; a prospective test on a subset would be the natural next step.
  • The fourfold gain is measured against random picking from the same database; in a real pipeline where chemists already bias their searches toward known spin-crossover ligand families, the practical advantage over that stronger baseline could be smaller.
  • The same equivariant architecture combined with relevance-based classification could transfer to other narrow-window material properties, such as singlet-triplet gaps or redox potentials, wherever DFT labels define a target interval.
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 / 5 minor

Summary. The paper constructs a data set of 1,439 mononuclear transition-metal complexes from the Cambridge Structural Database, computes r2SCAN spin-switching energies ΔEHL for each, and trains an equivariant graph neural network with 915 trainable parameters to predict ΔEHL. On a 10% holdout set the model achieves MAE 360 meV and R² 0.88. The regression is converted into a classifier through a relevance function built from sigmoids on the interval −200 ≤ ΔEHL ≤ 500 meV; at relevance 0.9 the authors report precision ≈80%, recall ≈35%, and false-positive rate 1.7%. They then use Bayes' theorem with the numbers 167/1296, 81/167, and 1065/1129 to claim a 56% posterior probability, a four-fold improvement over random picking, a 94% refusal probability, and a 17-fold reduction in redundant DFT calculations. The model is applied to 11,356 CSD molecules, yielding 1,076 predicted candidates, of which about 861 are expected to be good based on the 80% test precision. The paper concludes by acknowledging that the chosen energy window is debatable and depends on the functional and on omitted zero-point corrections.

Significance. The screening pipeline is potentially useful: the data set of DFT-computed spin-switching energies, the compact equivariant model, and the relevance-based classification scheme are all sensible components, and the coordination-shell analysis agrees qualitatively with the spectrochemical series. The use of a standard holdout split, the small parameter count, and the explicit statement of the label definition are strengths. However, the headline enrichment and reduction claims are currently not internally consistent, and the numerical basis for them is not reported with sufficient provenance. The practical significance of the work therefore cannot be assessed until the statistical claims are reconciled.

major comments (3)
  1. [Results and discussion, Figure 5(b) and the Bayes paragraph] The operating point at relevance 0.9 is incompatible with the contingency table used in the Bayes calculation. The text reports precision ≈80%, recall ≈35%, and FPR ≈1.7% at relevance 0.9. The Bayes calculation uses P(ΔÊ)=167/1296, P(ΔE|ΔÊ)=81/167, and P(ΔE|ΔÊ^c)=1−1065/1129. This table implies recall = 81/167 = 48.5%, FPR = 64/1129 = 5.7%, and precision = 81/(81+64) = 55.9%, none of which matches the stated operating point. In addition, counts such as 81, 167, 1065, and 1129 cannot be held-out test counts for a test set of roughly 144 molecules; their provenance (training set, cross-validation aggregate, or a different threshold) is not given. Because the four-fold enrichment, the 94% refusal probability, and the 17-fold reduction are all derived from this Bayes table, the central screening claims must be recomputed from one explicitly sourced confusion matrix at the selected relevance threshold.
  2. [Results and discussion, Bayes paragraph and large-scale screen] The random-picking baseline used for the four-fold claim is the training-set prevalence 167/1296 ≈ 12.9%. The large-scale screen is performed on a different pool of 11,356 CSD molecules, whose composition and prevalence of candidates need not match the training set. The enrichment factor is not well-defined for the actual discovery scenario unless the baseline prevalence is stated for the screened pool or an explicit assumption is made that the training prevalence transfers. Please clarify which baseline is being used and recompute the enrichment factor accordingly.
  3. [Results and discussion, large-scale screen paragraph] The statement that approximately 861 of the 1,076 predicted materials are expected to have ΔEHL within the range of interest is obtained by multiplying 1,076 by the 80% test precision. Since the 80% precision and the precision implied by the Bayes table (≈56%) are mutually inconsistent, this expectation is not currently supported. Recomputing with the Bayes-table precision gives roughly 603 instead of 861, a material difference. The expected number of true candidates must be derived from the same consistent confusion matrix and threshold used for the other screening claims.
minor comments (5)
  1. [Introduction, p. 2] The sentence beginning 'The choice of ligands The attractive feature...' is incomplete and should be rewritten.
  2. [Equation (3) and Figure 5(a)] The same sigmoid form is used for the lower and upper relevance thresholds, but the parameters c and s are not specified separately for each boundary; please give the explicit values used.
  3. [Results and discussion, Figure 5(b)] Because the holdout set is only about 10% of 1,439 molecules, the reported precision and recall values are subject to substantial sampling uncertainty; reporting exact test-set counts and confidence intervals would make the operating point more interpretable.
  4. [Results and discussion, p. 15] There is a typo: 'minimizing the the number of redundant calculations' contains a duplicated 'the'.
  5. [Concluding remarks] The abstract and conclusions state the improvement as a property of the model, but the improvement is relative to the computational label −200 ≤ ΔEHL ≤ 500 meV; this qualification should appear wherever the four-fold and 17-fold numbers are quoted.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the regression and classifier claims are evaluated on held-out data, and the label interval is an acknowledged, debatable definition rather than a fitted input.

full rationale

The derivation chain is self-contained. The regression target ΔE_HL is computed for each CSD structure with r2SCAN, the model is trained on 90% of the data, and performance is reported on a held-out 10% test set (test MAE 360 meV, R^2 = 0.88), so the regression claim is not a restatement of the training data. The relevance classifier converts the same regression target into a binary label using the explicit interval −200 to 500 meV, and the precision-recall curve and Bayes table report empirical frequencies of model predictions against those labels; no fitted physical constant is recycled into the claimed enrichment. The paper explicitly flags the interval choice as debatable and the omission of zero-point corrections as a limitation, which concerns label validity rather than circularity. The only adjacent concerns are that Refs. 60 and 62 supporting the r2SCAN label quality have overlapping authorship and that the Bayes numbers (167/1296, 81/167, 1065/1129) are not reconciled with the separately reported 80% precision and 1.7% false-positive rate; these are correctness and provenance issues, not a demonstrated reduction of a prediction to its inputs.

Assumptions & free parameters 5 free parameters · 5 assumptions · 0 invented entities

The paper introduces no new physical entities or forces. Its free parameters are the ML hyperparameters, the relevance threshold, and the hand-chosen energy window; the key axioms are the accuracy of r2SCAN DFT energies, the validity of the energy-window definition of spin crossover, and the quality of the graph descriptors.

free parameters (5)
  • Relevance threshold f(Y) = 0.9
    Chosen on the test set to maximize precision while retaining reasonable recall; affects the reported 80%/35% operating point and the 1,076-candidate count.
  • SCO energy window boundaries = -200 meV and 500 meV
    Hand-selected interval defining spin-crossover candidacy; the paper admits it is debatable and dependent on the DFT functional.
  • Sigmoid slope s = set by 1 meV energy resolution
    A modeling choice in the relevance function linking continuous ΔE_HL to class relevance; influences the shape of the precision-recall curve.
  • EGNN hyperparameters = learning rate 1e-2, weight decay 1e-6, early stopping threshold 1 meV, tanhshrink activation, single layer, 915…
    Chosen by grid search and manual design; the model's 915 trainable parameters are fit to the training data and are the core of the predictor.
  • Gradient boosting hyperparameters = learning rate 7e-2, Huber loss, 350 estimators, max depth 64, min samples leaf 4, min samples split 52
    Baseline model hyperparameters from grid search; used for comparison only and not load-bearing for the main claim.
assumptions (5)
  • domain assumption r2SCAN provides sufficiently accurate spin-switching energies for these transition-metal complexes
    Invoked in Methodology to justify using r2SCAN; supported by the same group's earlier benchmark (Ref. 60) but still an approximation whose accuracy is comparable to the energy differences of interest.
  • ad hoc to paper The -200 to 500 meV ΔE_HL window identifies spin-crossover candidates
    Used to define the target class for the relevance classifier and all screening metrics; the paper concedes the assumption is debatable and that zero-point energies are omitted.
  • domain assumption Oxidation states and bond orders computed from DFT electron densities are valid node and edge descriptors
    Used as input features; no validation of their accuracy is provided in the paper.
  • domain assumption Isolating the molecular unit from the experimental crystal structure preserves the relevant spin-crossover physics
    The paper switches from crystalline energy differences to molecular units for the learning step without a detailed protocol or validation.
  • standard math The equivariant graph network architecture (Satorras et al.) is appropriate and its equivariance guarantees hold for this data
    The architecture and its equivariance proof are taken from the cited E(n)-GNN paper; no formal re-verification is provided here.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Discovery of Spin-Crossover Candidates with Equivariant Graph Neural Networks and Relevance-Based Classification." pith.science (2026). https://pith.science/paper/2K43VCSZ

@misc{pith2026250105341,
  author       = {Pith},
  title        = {Pith review of: Discovery of Spin-Crossover Candidates with Equivariant Graph Neural Networks and Relevance-Based Classification},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/2K43VCSZ}},
  note         = {Machine review of arXiv:2501.05341}
}
read the original abstract

Swift discovery of spin-crossover materials for their potential application in quantum information devices requires techniques which enable efficient identification of suitably bistable candidates. To this end, we screened the Cambridge Structural Database to develop a specialized database of 1,439 materials and computed spin-switching energies from density functional theory for each material. The database was used to train an equivariant graph convolutional neural network to predict the magnitude of the spin-conversion energy. A test mean absolute error was 360 meV. For candidate identification, we equipped the system with a relevance-based classifier. This approach leads to a nearly four-fold improvement in identifying potential spin-crossover systems of interest as compared to conventional high-throughput screening.

Figures

Figures reproduced from arXiv: 2501.05341 by the authors.

Figure 1
Figure 1. Distribution of the calculated crystalline spin-crossover energies per molecule for [PITH_FULL_IMAGE:figures/full_fig_p005_1.png] view at source ↗
Figure 2
Figure 2. (a) Behavior of the mean signed error loss with a summation reduction as a [PITH_FULL_IMAGE:figures/full_fig_p009_2.png] view at source ↗
Figure 3
Figure 3. Behavior of the RMSE and R2 as a function of graphs built with nodes and edges for sequentially increased numbers of coordination shell approximations. Two reference types were chosen, namely, coordination shells starting at the metallic center or at a random atom of the molecule. For both cases, the reference atom was kept fixed once picked. From the implementation perspective, our goals include learning chemically… view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: Normalized compositions for Carbon, Nitrogen, Oxygen, Hydrogen and other [PITH_FULL_IMAGE:figures/full_fig_p012_4.png]
Figure 5
Figure 5. Figure 5: Classification of molecular metal complexes as potential candidates for spin [PITH_FULL_IMAGE:figures/full_fig_p014_5.png]
Figure 6
Figure 6. Figure 6: Predicted spin-crossover energies for mononuclear complexes containing (a) Cr, (b) [PITH_FULL_IMAGE:figures/full_fig_p017_6.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

88 extracted references · 75 canonical work pages

  1. [1]

    G \"u tlich, P.; H, A. G. Spin Crossover in Transition Metal Compounds I; 2004; pp 1--47

    2004

  2. [2]

    Spin Crossover in Transition Metal Compounds I; 2004; pp 49--58

    Hauser, A. Spin Crossover in Transition Metal Compounds I; 2004; pp 49--58

    2004

  3. [3]

    Thermodynamical aspects of the spin crossover phenomenon

    Nicolazzi, W.; Bousseksou, A. Thermodynamical aspects of the spin crossover phenomenon. C. R. Chimie 2016, 21, 1060--1074

    2016

  4. [4]

    K.; Kr \"o ber, J.; Jay, C

    J., O. K.; Kr \"o ber, J.; Jay, C. Spin Transition Molecular Materials for displays and data recording. Adv. Mater. 1992, 4, 718--728

    1992

  5. [5]

    Kahn, O.; Martinez, C. J. Spin-Transition Polymers: From Molecular Materials Toward Memory Devices. Science 1998, 279, 44--48

    1998

  6. [6]

    Spin Transition Molecular Materials: New Sensors

    Garcia, Y.; Ksenofontov, V.; G \"u tlich, P. Spin Transition Molecular Materials: New Sensors. Hyperfine Interact. 2002, 139, 543--551

    2002

  7. [7]

    N.; Vander, E

    Muller, R. N.; Vander, E. L.; Laurent, S. Spin Transition Molecular Materials: Intelligent Contrast Agents for Magnetic Resonance Imaging. J. Am. Chem. Soc. 2003, 125, 8405--8407

    2003

  8. [8]

    Spin Crossover in Transition Metal Compounds III; 2004; pp 221--249

    L \'e tard, J.-F.; Guionneau, P.; Goux-Capes, L. Spin Crossover in Transition Metal Compounds III; 2004; pp 221--249

    2004

Show all 88 references
  1. [9]

    Towards the Ultimate Size Limit of the Memory Effect in Spin-Crossover Solids

    Larionova, J.; Salmon, L.; Guari, Y.; Tokarev, A.; Molvinger, K.; Moln \'a r, G.; Bousseksou, A. Towards the Ultimate Size Limit of the Memory Effect in Spin-Crossover Solids. Angew. Chem. Int. Ed. 2008, 47, 8236--8240

    2008

  2. [10]

    Molecular spintronics

    Sanvito, S. Molecular spintronics. Chem. Soc. Rev. 2011, 40, 3336--3355

    2011

  3. [11]

    M.; F \'e lix, G.; Suleimanov, I.; Moln \'a r, C

    Quintero, C. M.; F \'e lix, G.; Suleimanov, I.; Moln \'a r, C. J. S. G.; Salmon, L.; Nicolazzi, W.; Bousseksou, A. Hybrid spin-crossover nanostructures. Beilstein J. Nanotechnol. 2014, 5, 2230--2239

    2014

  4. [12]

    Charge transport properties of spin crossover systems

    Ruiz, E. Charge transport properties of spin crossover systems. Phys. Chem. Chem. Phys. 2014, 16, 14--22

    2014

  5. [13]

    M.; Linares, J.; Rotaru, A.; Ritti, M

    Jureschi, C. M.; Linares, J.; Rotaru, A.; Ritti, M. H.; Parlier, M.; D \^i rtu, M. M.; Wolff, M.; Garcia, Y. Pressure Sensor via Optical Detection Based on a 1D Spin Transition Coordination Polymer. Sensors 2015, 15, 2388--2398

    2015

  6. [14]

    D.; Rat, S.; Salmon, L.; Moln \'a r, G.; Quintero, C

    Manrique-Ju \'a rez, M. D.; Rat, S.; Salmon, L.; Moln \'a r, G.; Quintero, C. M.; Nicu, L.; Shepherd, H. J.; Bousseksou, A. Switchable molecule-based materials for micro- and nanoscale actuating applications: Achievements and prospects. Coord. Chem. Rev. 2016, 308, 395--408

    2016

  7. [15]

    Mikolasek, M. et al. Complete Set of Elastic Moduli of a Spin-Crossover Solid: Spin-State Dependence and Mechanical Actuation. J. Am. Chem. Soc. 2018, 140, 8970--8979

    2018

  8. [16]

    Cruddas, J.; Powell, B. J. Multiple Coulomb phases with temperature-tunable ice rules in pyrochlore spin-crossover materials. Phys. Rev. B 2021, 104, 024433

    2021

  9. [17]

    Daku, L. M. L.; Aquilante, F.; Robinson, T. W.; Hauser, A. Accurate Spin-State Energetics of Transition Metal Complexes. 1. CCSD(T), CASPT2, and DFT Study of [M(NCH) _6 ] ^ 2+ (M = Fe, Co). J. Chem. Theory. Comput. 2012, 8, 4216--4231

    2012

  10. [18]

    Ground state of a spin-crossover molecule calculated by diffusion Monte Carlo

    Droghetti, A.; Alf \'e , D.; Sanvito, S. Ground state of a spin-crossover molecule calculated by diffusion Monte Carlo. Phys. Rev. B 2013, 87, 205114

    2013

  11. [19]

    Benchmarks and Reliable DFT Results for Spin Gaps of Small Ligand Fe(II) Complexes

    Song, S.; Kim, M.-C.; Sim, E.; Benali, A.; Heinonen, O.; Burke, K. Benchmarks and Reliable DFT Results for Spin Gaps of Small Ligand Fe(II) Complexes. J. Chem. Theory. Comput. 2018, 14, 2304--2311

    2018

  12. [20]

    M.; Feldt, M.; Harvey, J

    Phung, Q. M.; Feldt, M.; Harvey, J. N.; Pierloot, K. Toward Highly Accurate Spin State Energetics in First-Row Transition Metal Complexes: A Combined CASPT2/CC Approach. J. Chem. Theory. Comput. 2018, 14, 2446--2455

    2018

  13. [21]

    M.; Guo, Y.; Riplinger, C.; Tuczek, F.; Neese, F

    Fl \"o ser, B. M.; Guo, Y.; Riplinger, C.; Tuczek, F.; Neese, F. Detailed Pair Natural Orbital-Based Coupled Cluster Studies of Spin Crossover Energetics. J. Chem. Theory Comput. 2020, 16, 2224--2235

    2020

  14. [22]

    Zhang, D.; Truhlar, D. G. Spin Splitting Energy of Transition Metals: A New, More Affordable Wave Function Benchmark Method and Its Use to Test Density Functional Theory. J. Chem. Theory Comput. 2020, 16, 4416--4428

    2020

  15. [23]

    A.; Pantazis, D

    Drosou, M.; Mitsopoulou, C. A.; Pantazis, D. A. Spin-state energetics of manganese spin crossover complexes: Comparison of single-reference and multi-reference ab initio approaches. Polyhedron 2021, 208, 115399

    2021

  16. [24]

    A.; Pantazis, D

    Drosou, M.; Mitsopoulou, C. A.; Pantazis, D. A. Reconciling Local Coupled Cluster with Multireference Approaches for Transition Metal Spin-State Energetics. J. Chem. Theory Comput. 2022, 18, 3538--3548

    2022

  17. [25]

    Kepp, K. P. Consistent descriptions of metal–ligand bonds and spin-crossover in inorganic chemistry. Coord. Chem. Rev. 2013, 257, 196--209

    2013

  18. [26]

    Kepp, K. P. In Transition Metals in Coordination Environments: Computational Chemistry and Catalysis Viewpoints; Broclawik, E., Borowski, T., Rado \'n , M., Eds.; Springer International Publishing, 2019; pp 1--33

    2019

  19. [27]

    E.; Benmansour, S.; Marchivie, M.; Triki, S.; Conan, F.; Patinec, V.; Handel, H.; Dupouy, G.; G \'o mez-Garc \'i a, C

    Atmani, C.; Hajj, F. E.; Benmansour, S.; Marchivie, M.; Triki, S.; Conan, F.; Patinec, V.; Handel, H.; Dupouy, G.; G \'o mez-Garc \'i a, C. J. Guidelines to design new spin crossover materials. Coord. Chem. Rev. 2010, 254, 1559--1569

    2010

  20. [28]

    Molecular Spin Crossover Materials: Review of the Lattice Dynamical Properties

    Moln \'a r, G.; Mikolasek, M.; Ridier, K.; Fahs, A.; Nicolazzi, W.; Bousseksou, A. Molecular Spin Crossover Materials: Review of the Lattice Dynamical Properties. Annalen der Physik 2019, 531, 1900076

    2019

  21. [29]

    Theoretical investigations on the pressure effects in spin-crossover materials: Reentrant phase transitions and other behavior

    Babilotte, K.; Boukheddaden, K. Theoretical investigations on the pressure effects in spin-crossover materials: Reentrant phase transitions and other behavior. Phys. Rev. B 2020, 101, 174113

    2020

  22. [30]

    Spin crossover and valence tautomerism conductors

    Wang, M.; Li, Z.-Y.; Ishikawa, R.; Yamashita, M. Spin crossover and valence tautomerism conductors. Coord. Chem. Rev. 2021, 435, 213819

    2021

  23. [31]

    P.; ; Kulik, H

    Janet, J. P.; ; Kulik, H. J. Predicting electronic structure properties of transition metal complexes with neural networks. Chem. Sci. 2017, 8, 5137--5152

    2017

  24. [32]

    P.; Chan, L.; Kulik, H

    Janet, J. P.; Chan, L.; Kulik, H. J. Accelerating chemical discovery with machine learning: simulated evolution of spin crossover complexes with an artificial neural network. J. Phys. Chem. Lett. 2018, 9, 1064--1071

    2018

  25. [33]

    G.; Yang, T.; Lin, S.; Nandy, A.; Janet, J

    Taylor, M. G.; Yang, T.; Lin, S.; Nandy, A.; Janet, J. P.; Duan, C.; Kulik, H. J. Seeing Is Believing: Experimental Spin States from Machine Learning Model Structure Predictions. J. Phys. Chem. A 2020, 124, 3286--3299

    2020

  26. [34]

    G.; Liu, F.; Steeves, A

    Nandy, A.; Duan, C.; Taylor, M. G.; Liu, F.; Steeves, A. H.; Kulik, H. J. Computational discovery of transition-metal complexes: from high-throughput screening to machine learning. Chem. Rev. 2021, 121, 9927--10000

    2021

  27. [35]

    Is There a Need for New Models of the Spin Crossover? Monatshefte f \"u r Chemie 2003, 134, 199--216

    Bo c a, R.; Linert, W. Is There a Need for New Models of the Spin Crossover? Monatshefte f \"u r Chemie 2003, 134, 199--216

    2003

  28. [36]

    B.; Linert, W

    Koudriavtsev, A. B.; Linert, W. Spin crossover - an unusual chemical equilibrium. J. Struct. Chem. 2010, 51, 335--365

    2010

  29. [37]

    B.; Terrones, G.; Duan, C.; Kulik, H

    Vennelakanti, V.; Kilic, I. B.; Terrones, G.; Duan, C.; Kulik, H. J. Machine Learning Prediction of the Experimental Transition Temperature of Fe(II) Spin-Crossover Complexes. J. Phys. Chem. A 2024, 128, 204--216

    2024

  30. [38]

    G.; Bellamkonda, A.; Maher, T

    Kench, T.; Rahardjo, A.; Terrones, G. G.; Bellamkonda, A.; Maher, T. E.; Storch, M.; Kulik, H. J.; Vilar, R. A semi-automated, high-throughput approach for the synthesis and identification of highly photo-cytotoxic iridium complexes. Angew. Chem. Int. Ed. 2024, 63, e202401808

    2024

  31. [39]

    Paulsen, H.; Trautwein, A. X. Spin Crossover in Transition Metal Compounds III. Topics in Current Chemistry, vol 235; Springer Berlin Heidelberg, 2004; pp 197--219

    2004

  32. [40]

    Paulsen, H.; Sch \"u nemann, V.; Wolny, J. A. Progress in electronic structure calculations on spin-crossover complexes. Eur. J. Inorg. Chem. 2013, 2013, 628--641

    2013

  33. [41]

    A feasibility study on obtaining d-orbital populations from aspherical-atom refinements on three spin crossover compounds

    Dittrich, B.; Ruf, E.; Meller, T. A feasibility study on obtaining d-orbital populations from aspherical-atom refinements on three spin crossover compounds. Struct. Chem. 2017, 28, 1333--1342

    2017

  34. [42]

    Benchmarking Density Functional Methods for Calculation of State Energies of First Row Spin-Crossover Molecules

    Cirera, J.; Via-Nadal, M.; Ruiz, E. Benchmarking Density Functional Methods for Calculation of State Energies of First Row Spin-Crossover Molecules. Inorg. Chem. 2018, 57, 14097--14105

    2018

  35. [43]

    Computational Modeling of Transition Temperatures in Spin-Crossover Systems

    Cirera, J.; Ruiz, E. Computational Modeling of Transition Temperatures in Spin-Crossover Systems. Inorg. Chem. 2019, 39, 216--241

    2019

  36. [44]

    o mer, A.; Hasecke, L.; Bl \

    R \"o mer, A.; Hasecke, L.; Bl \"o chl, P.; Mata, R. A. A Review of Density Functional Models for the Description of Fe(II) Spin-Crossover Complexes. Molecules 2020, 25, 5176

    2020

  37. [45]

    Zheng, B.; Gu, G. X. Prediction of Graphene Oxide Functionalization Using Gradient Boosting: Implications for Material Chemical Composition Identification. ACS Appl. Nano Mater. 2021, 4, 3167--3174

    2021

  38. [46]

    A.; Joshi, R

    Moses, I. A.; Joshi, R. P.; Ozdemir, B.; Kumar, N.; Eickholt, J.; Barone, V. Machine Learning Screening of Metal-Ion Battery Electrode Materials. ACS Appl. Mater. & Interfaces 2021, 13, 53355--53362

    2021

  39. [47]

    Duan, C.; Nandy, A.; Kulik, H. J. Machine Learning for the Discovery, Design, and Engineering of Materials. Annu. Rev. Chem. Biomol. Eng. 2022, 13

    2022

  40. [48]

    H.; Sun, M.; Huang, B

    Chan, C. H.; Sun, M.; Huang, B. Application of machine learning for advanced material prediction and design. EcoMat 2022, e12194

    2022

  41. [49]

    N.; Welling, M

    Kipf, T. N.; Welling, M. Semi-Supervised Classification with Graph Convolutional Networks. arXiv:1609.02907 2017, v4, 1--14

    2017 arXiv

  42. [50]

    M.; Bruna, J.; LeCun, Y.; Szlam, A.; Vandergheynst, P

    Bronstein, M. M.; Bruna, J.; LeCun, Y.; Szlam, A.; Vandergheynst, P. Geometric Deep Learning: Going beyond Euclidean data. IEEE Signal Processing Magazine 2017, 34, 18--42

    2017

  43. [51]

    In Representation Learning for Natural Language Processing; Liu, Z., Lin, Y., Sun, M., Eds.; Springer Nature Singapore, 2023; pp 169--210

    Yang, C.; Lin, Y.; Liu, Z.; Sun, M. In Representation Learning for Natural Language Processing; Liu, Z., Lin, Y., Sun, M., Eds.; Springer Nature Singapore, 2023; pp 169--210

    2023

  44. [52]

    A review of graph neural networks: concepts, architectures, techniques, challenges, datasets, applications, and future directions

    Khemani, B.; Patil, S.; Kotecha, K.; Tanwar, S. A review of graph neural networks: concepts, architectures, techniques, challenges, datasets, applications, and future directions. J. Big Data 2024, 11, 18

    2024

  45. [53]

    Invariant and Equivariant Graph Networks

    Maron, H.; Ben-Hamu, H.; Shamir, N.; Lipman, Y. Invariant and Equivariant Graph Networks. arXiv:1812.09902 2018, v2, 1--14

    2018 arXiv

  46. [54]

    Tensor field networks: Rotation- and translation-equivariant neural networks for 3D point clouds

    Thomas, N.; Smidt, T.; Kearnes, S.; Yang, L.; Li, L.; Kohlhoff, K.; Riley, P. Tensor field networks: Rotation- and translation-equivariant neural networks for 3D point clouds. arXiv:1802.08219 2018, v3, 1--19

    2018 arXiv

  47. [55]

    Equivariant Flows: sampling configurations for multi-body systems with symmetric energies

    K \"o hler, J.; Klein, L.; No \'e , F. Equivariant Flows: sampling configurations for multi-body systems with symmetric energies. arXiv:1910.00753 2019, v1, 1--5

    1910 arXiv

  48. [56]

    G.; Welling, E

    Satorras, V. G.; Welling, E. H. M. E(n) Equivariant Graph Neural Networks. Proceedings of the 38th International Conference on Machine Learning. 2021; pp 9323--9332

    2021

  49. [57]

    G.; Welling, E

    Satorras, V. G.; Welling, E. H. M. E(n) Equivariant Graph Neural Networks. 2021; https://github.com/vgsatorras/egnn, accessed on 09 December 2023

    2021

  50. [58]

    R.; Bruno, I

    Groom, C. R.; Bruno, I. J.; Lightfoot, M. P.; Ward, S. C. The Cambridge structural database. Acta Crystallogr. B: Struct. Sci. Cryst. Eng. Mater. 2016, 72, 171--179, accessed on 26 July 2021

    2016

  51. [59]

    W.; Kaplan, A

    Furness, J. W.; Kaplan, A. D.; Ning, A.; Perdew, J. P.; Sun, J. Accurate and numerically efficient r2SCAN meta-generalized gradient approximation. J. Phys. Chem. Lett. 2020, 11, 8208--8215

    2020

  52. [60]

    Mej \'i a-Rodr \'i guez, D.; Trickey, S. B. Spin-crossover from a Well-behaved, Low-cost meta-GGA Density Functional. Journal of Physical Chemistry A 2020, 124, 9889

    2020

  53. [61]

    Efficient iterative schemes for ab initio total-energy calculations using a plane-wave basis set

    Kresse, G.; Furthm \"u ller, J. Efficient iterative schemes for ab initio total-energy calculations using a plane-wave basis set. Phys. Rev. B 1996, 56, 11169

    1996

  54. [62]

    G.; Trickey, S

    Mej \'i a-Rodr \'i guez, D.; Albavera-Mata, A.; Fonseca, E.; Chen, D.-T.; Cheng, H.-P.; Hennig, R. G.; Trickey, S. Barriers to predictive high-throughput screening for spin-crossover. Comput. Mater. Sci. 2022, 206, 111161

    2022

  55. [63]

    D.; Wilkinson, A

    McNaught, A. D.; Wilkinson, A. IUPAC Compendium of Chemical Terminology; International Union of Pure and Applied Chemistry, 2019

    2019

  56. [64]

    In Neural Networks: Tricks of the Trade; Orr, G

    Prechelt, L. In Neural Networks: Tricks of the Trade; Orr, G. B., M \"u ller, K.-B., Eds.; Springer: Berlin, Heidelberg, 1998; pp 55--69

    1998

  57. [65]

    P.; Ba, J

    Kingma, D. P.; Ba, J. Adam: A Method for Stochastic Optimization. arXiv:1412.698 2017, v9, 1412.6980

    2017 arXiv

  58. [66]

    Krizhevsky, A.; Sutskever, I.; Hinton, G. E. ImageNet Classification with Deep Convolutional Neural Networks. Commun. ACM 2017, 60, 84--90

    2017

  59. [67]

    Pedregosa, F. et al. Scikit-learn: Machine Learning in Python. J. Mach. Learn. Res. 2011, 12, 2825--2830

    2011

  60. [68]

    Absorption Spectra of Co-ordination Compounds

    Tsuchida, R. Absorption Spectra of Co-ordination Compounds. I. Bull. Chem. Soc. Jpn. 1938, 13, 388--400

    1938

  61. [69]

    Absorption Spectra of Co-ordination Compounds

    Tsuchida, R. Absorption Spectra of Co-ordination Compounds. II. Bull. Chem. Soc. Jpn. 1938, 13, 436--450

    1938

  62. [70]

    Absorption Spectra of Co-ordination Compounds

    Tsuchida, R.; Kobayasi, M. Absorption Spectra of Co-ordination Compounds. III. Special Bands of Chromium Complexes. Bull. Chem. Soc. Jpn. 1938, 13, 471--480

    1938

  63. [71]

    Absorption Spectra of Co-ordination Compounds

    H.Kuroya,; Tsuchida, R. Absorption Spectra of Co-ordination Compounds. IV. Ethylenediamine Cobaltic Complexes. Bull. Chem. Soc. Jpn. 1940, 15, 427--439

    1940

  64. [72]

    Absorption Spectra of Coordination Compounds

    Yamada, S.; Tsuchida, R. Absorption Spectra of Coordination Compounds. V. The Hyperchromic Series of Ligands. Bull. Chem. Soc. Jpn. 1953, 26, 15--19

    1953

  65. [73]

    Spin crossover iron(III) complexes

    Nihei, M.; Shiga, T.; Maeda, Y.; Oshio, H. Spin crossover iron(III) complexes. Coord. Chem. Rev. 2007, 251, 2606--2621

    2007

  66. [74]

    B.; Seredyuk, M

    Gaspar, A. B.; Seredyuk, M. Spin crossover in soft matter. Coord. Chem. Rev. 2014, 268, 41--58

    2014

  67. [75]

    A.; Suleimenova, A

    Shatruk, M.; Phan, H.; Chrisostomo, B. A.; Suleimenova, A. Symmetry-breaking structural phase transitions in spin crossover complexes. Coord. Chem. Rev. 2015, 289--290, 62--73

    2015

  68. [76]

    J.; Harding, P.; Phonsri, W

    Harding, D. J.; Harding, P.; Phonsri, W. Spin crossover in iron(III) complexes. Coord. Chem. Rev. 2016, 313, 38--61

    2016

  69. [77]

    Molecular magnetism, quo vadis? A historical perspective from a coordination chemist viewpointi

    Ferrando-Soria, J.; Vallejo, J.; Castellano, M.; Mart \'i nez-Lillo, J.; Pardo, E.; Cano, J.; Castro, I.; Lloret, F.; Ruiz-Garc \'i a, R.; Julve, M. Molecular magnetism, quo vadis? A historical perspective from a coordination chemist viewpointi. Coord. Chem. Rev. 2017, 339, 17--103

    2017

  70. [78]

    W.; Singh, S.; Brooker, S

    Hogue, R. W.; Singh, S.; Brooker, S. Spin crossover in discrete polynuclear iron(II) complexes. Chem. Soc. Rev. 2018, 47, 7303

    2018

  71. [79]

    Infrared studies of coordination compounds containing low-oxidation-state metals

    Saito, Y.; Takemoto, J.; Hutchinson, B.; Nakamoto, K. Infrared studies of coordination compounds containing low-oxidation-state metals. I. Tris (2,2'-bipyridine) and tris (1,10-phenanthroline) complexes. Inorg. Chem. 1972, 11, 2003--2011

    1972

  72. [80]

    Phonon coupled cooperative low-spin ^1 A ^1 high-spin ^5 T ^2 transition in [Fe(phen) _2 (NCS) _2 ] and [Fe(phen) _2 (NCSe) _2 ] crystals

    Sorai, M.; Seki, S. Phonon coupled cooperative low-spin ^1 A ^1 high-spin ^5 T ^2 transition in [Fe(phen) _2 (NCS) _2 ] and [Fe(phen) _2 (NCSe) _2 ] crystals. J. Phys. Chem. Solids 1974, 35, 555--570

    1974

  73. [81]

    K.; Sasikala, R.; K \"o nig, E

    Kulshreshtha, S. K.; Sasikala, R.; K \"o nig, E. Calorimetric investigations of the low-spin ( ^1 A ^1 ) high-spin ( ^5 T ^2 ) transition in solid dithiocyanatobis(2,2'-BI-2-thiazoline iron(III). Chem. Phys. Lett. 1986, 123, 215--217

    1986

  74. [82]

    Paulsen, H.; Duelund, L.; Winkler, H.; Toftlund, H.; Trautwein, A. X. Free energy of spin-crossover complexes calculated with density functional methods. Inorg. Chem. 2001, 40, 2201--2203

    2001

  75. [83]

    Spin equilibrium in solutions

    Toftlund, H. Spin equilibrium in solutions. Chem. Mon. 2001, 132, 1269--1277

    2001

  76. [84]

    Calorimetric investigation of phase transitions occurring in molecule-based magnets

    Sorai, M.; Nakano, M.; Miyazaki, Y. Calorimetric investigation of phase transitions occurring in molecule-based magnets. Chem. Rev. 2006, 106, 976--1031

    2006

  77. [85]

    R.; Kepp, K

    Mortensen, S. R.; Kepp, K. P. Spin propensities of octahedral complexes from density functional theory. J. Phys. Chem. A 2015, 119, 4041--4050

    2015

  78. [86]

    Cook, L. J. K.; Kulmaczewski, R.; Mohammed, R.; Dudley, S.; Barrett, S. A.; Little, M. A.; Deeth, R. J.; Halcrow, M. A. A unified treatment of the relationship between ligand substituents and spin state in a family of iron (II) complexes. Angew. Chem. Int. Ed. 2016, 55, 4237--4331

    2016

  79. [87]

    Precision and Recall for Regression

    Torgo, L.; Ribeiro, R. Precision and Recall for Regression. Discovery Science. Berlin, Heidelberg, 2009; pp 332--346

    2009

  80. [88]

    ?N i Cs79 ]4] > MĜ8 W= hڕ

    Murray, K. S. In Spin-Crossover Materials: Properties and Applications; Halcrow, M. A., Ed.; Wiley-Blackwell: Australia, 2013; pp 1--54 mcitethebibliography Albavera_Mata_et_al_Feb_05_2025.tex0000664000000000000000000012740714752247444015703 0ustar rootroot achemso amsfonts,am...

    2013

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.