{"id":"5c8925fd-c330-4723-b750-12df7e70a556","arxiv_id":"2506.08339","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"A machine-learning interatomic potential for La-Si-P is developed via iterative training, reproducing DFT energetics and liquid structure, with melting temperatures 5 to 20 percent below experiment.","lead":"The authors trained a neural network interatomic potential for the La-Si-P ternary system using density functional theory data and molecular dynamics simulations, then used it to predict melting temperatures and crystallization. The potential reproduces many crystal and liquid properties, though melting temperatures are systematically underestimated by 5 to 20 percent, especially for binary and elemental phases.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The claimed melting-temperature accuracy is not an independent test: melting-path snapshots were added to the training set (Fig. 2d and Sec. 4), so the 5-10% agreement with experiment may be a fitting artifact and the \"not explicitly included\" statement is contradicted by the paper's own text.","rationale":"The reader's verdict is CONDITIONAL, and I agree with that verdict, so no verdict change is needed. However, the identity of the load-bearing concern differs from the reader's stated weakest assumption. The reader emphasized non-spin-polarized PBE for La 4f electrons; that is a reasonable concern in principle, but for La in La-Si-P compounds La is typically trivalent with an empty 4f shell, and non-spin-polarized PBE is a standard treatment for such systems. The more concrete and self-evident problem is the training-data circularity: the paper explicitly states that melting-path configurations were added to training and that this improved melting-temperature prediction, while simultaneously claiming melting temperatures were not explicitly included. Since the paper's own Fig. 2d shows that adding these configurations substantially shifted the predicted Tm, the melting-temperature comparison cannot be treated as an independent transferability test. This does not invalidate the potential's utility for qualitative MD studies, and the authors are transparent about the melting-data additions in Section 4, but it does undercut the strongest quantitative validation claim in the abstract. A single ablation experiment would settle whether the 5-10% agreement survives without the melting-path training data. The lack of released code/data and the unpublished nucleation comparison [57] also limit verifiability, but those are secondary to the internal inconsistency about training data.","tokens_in":11791,"tokens_out":5506,"duration_ms":72762,"concrete_test":"Perform an ablation: retrain the ANN-ML model from the same initial training set (distorted crystals + AIMD liquids) but exclude all solid-liquid transition snapshots added in the 4th and later iterations. Recompute coexistence Tm for Aea2 LaSiP3 and La2SiP4 with the ablated model. If Tm drops by a large amount (e.g., >100 K) relative to the final model, the reported 5-10% agreement is a training artifact and the \"not explicitly included\" claim fails. In parallel, inspect the training-data log to determine whether Aea2 LaSiP3 melting-path configurations are present; this resolves whether the claim is false for one or both ternary phases.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central evidence for transferability of the La-Si-P ANN-ML potential is its reported 5-10% agreement with experimental melting temperatures for ternary phases. But the paper's own description undermines that evidence. Section 2 (Fig. 2d) states that in the 4th training iteration, \"using the pretrained ANN-ML model to add to the training data set with additional structures collected during the crystal to liquid transition\" substantially increased the predicted melting temperature of La2SiP4. Section 4 repeats: \"we also add structures near the melting temperature to fine-tune the ANN-ML model... adding these new training data can improve the performance of the ANN-ML model significantly, especially for melting temperature prediction.\" Yet Section 3 concludes that the 5-10% agreement is \"remarkable\" because \"the melting temperatures are not explicitly included in training.\" This is internally inconsistent. If the final model was trained on solid-liquid coexistence configurations for La2SiP4 (explicitly) and possibly for Aea2 LaSiP3 (covered by the general statement), then the coexistence MD simulation is not an independent prediction of Tm but a test of how well the model interpolates on transition configurations whose DFT energies and forces were provided. The paper shows in Fig. 2d that adding these configurations changed Tm substantially, so the final Tm is sensitive to those data. The claim that the potential accurately predicts melting temperatures therefore rests on an undocumented assumption that the added melting-path data did not effectively encode Tm. Without an ablation or a clear list of which phases contributed melting-path training snapshots, the strongest validation of the potential is not established.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports the development of a DeePMD-based neural network machine learning interatomic potential for the La-Si-P ternary system. The authors construct a large training set of DFT energies and forces for distorted crystals and liquid snapshots across many compositions, use iterative retraining with adjusted data weights and added transition-state structures, and validate the potential against DFT energy-volume (E-V) curves and liquid pair-correlation functions. They then predict melting temperatures of ternary LaSiP3 and La2SiP4 via solid/liquid coexistence molecular dynamics (MD), compare with experimental values (including a new differential scanning calorimetry measurement), and simulate LaP nucleation from La-Si-P liquids. The paper claims that the potential is accurate and transferable, with a 5-10% underestimation of ternary melting temperatures and qualitative agreement with observed crystallization behavior.","tokens_in":12058,"tokens_out":5462,"duration_ms":62690,"significance":"If the developed potential is genuinely accurate and transferable, it would provide a valuable computational tool for guiding synthesis of ternary La-Si-P phases and for studying phase competition and growth kinetics in this materials family. The paper has notable strengths: a systematic iterative training workflow, broad compositional coverage of the training set, a new DSC measurement of the La2SiP4 melting temperature, and successful reproduction of DFT E-V curves and liquid structures. However, the load-bearing evidence for transferability--the melting-temperature predictions--is significantly weakened by the paper's own description of the training procedure, which added solid-liquid transition configurations to the training set. The agreement for ternary compounds is therefore partly a fitting result, and the large errors reported for LaSi and elemental La (about 20% and 19%, respectively) further temper the transferability claim. The choice of non-spin-polarized PBE for lanthanum-containing phases also raises a correctness risk that is not addressed in the manuscript.","major_comments":[{"comment":"The statement in Section 3 that 'the melting temperatures are not explicitly included in training' is internally inconsistent with the procedure described elsewhere. Section 2 states that in the 4th iteration, 'additional structures collected during the crystal to liquid transition' were added to the training data set, and Fig. 2(d) shows that this addition substantially increased the predicted melting temperature of La2SiP4. Section 4 repeats that adding structures near the melting temperature 'can improve the performance of the ANN-ML model significantly, especially for melting temperature prediction.' The predicted Tm values for La2SiP4 (and possibly LaSiP3, given the general statement) are therefore not independent predictions but tests of how well the model interpolates on transition-state configurations whose DFT energies and forces were explicitly provided during training. The authors should remove or substantially qualify the 'remarkable' claim and instead discuss the agreement as a validation that the potential describes the solid-liquid coexistence region, not as evidence that Tm was predicted from data that omitted melting-relevant configurations.","section":"Section 3, page 11; Section 2, page 6 (Fig. 2d); Section 4, page 15"},{"comment":"The potential underestimates the melting temperatures of LaSi and elemental La by about 20% and 19%, respectively, while the ternary compounds are within 5-10% of experiment. The paper attributes the systematic underestimation to the use of GGA-DFT reference data, but this explanation does not account for the large variation across phases. Given the demonstrated sensitivity of Tm to the addition of melting-path training data in Fig. 2(d), the difference between the ternary and elemental/binary results likely reflects a deficiency in transition-state training data for the latter phases rather than a purely systematic DFT offset. The claim of an 'accurate and transferable' potential should be tempered, or the authors should add melting-transition training data for LaSi and La to test whether the errors can be reduced, rather than attributing the full discrepancy to the DFT functional.","section":"Section 3, Fig. 8"},{"comment":"The exchange-correlation functional is non-spin-polarized PBE, and no justification or test of this choice is provided for lanthanum, a rare-earth element with 4f electrons. If the DFT reference is inaccurate for La-containing phases due to unpaired 4f spins or strong correlations, the machine learning potential will inherit these errors, undermining the claimed transferability across the La-Si-P phase space. The authors should either verify that spin-polarized (or DFT+U) calculations give negligible energy differences for representative elemental, binary, and ternary La-containing structures, or explicitly discuss this limitation as a correctness risk in the computed melting temperatures and nucleation behavior.","section":"Section 2, page 5 (DFT settings)"},{"comment":"The E-V curves are presented as evidence of transferability, but Section 2 indicates that the training set already includes the same stable elemental, binary, and ternary phases taken from the Materials Project, as well as the previously predicted ternary compounds. The E-V agreement is therefore partly a check on the training set, not an independent test. To substantiate the transferability claim, the authors should identify which phases in Fig. 4 were not included in the training data, report their E-V errors separately, and make clear whether the iterative training procedure was adjusted based on these same E-V results.","section":"Section 3, page 8 (E-V curves)"}],"minor_comments":[{"comment":"The description of training data generation states that '21 to 41 non-distorted structures' and '2100 to 4100 distorted structures' are generated, but it is unclear whether these numbers are per parent phase or in total; please clarify and verify that the stated total of 71,800 distorted crystal structures is consistent with the number of parent compounds.","section":"Section 2, page 4"},{"comment":"The caption says 'adding relevant training data from MD simulation of liquid to solid transitions,' while the main text describes 'structures collected during the crystal to liquid transition'; please unify the terminology.","section":"Fig. 2(d) caption"},{"comment":"The phase notation 'Aea2 LaSiP3' is used without explanation; please define the space group and, if needed, the structure type, so that readers unfamiliar with the polymorph can follow the discussion.","section":"Section 3, page 10"},{"comment":"The DSC sample is described as a mixture of LaP and La2SiP4, and the endothermic peak at ~1330 K is only 'likely' assigned to the melting of La2SiP4; please provide additional evidence for this assignment, such as a comparison with pure La2SiP4 or a discussion of the LaP-La2SiP4 phase diagram and possible eutectic effects.","section":"Section 3, page 11 (DSC of La2SiP4)"},{"comment":"The coexistence MD description does not specify the length of the NVE simulation over which the temperature (kinetic energy) is averaged; please provide the simulation time and any equilibration criteria used to determine when coexistence is reached.","section":"Section 3, page 10 (coexistence simulations)"},{"comment":"The nucleation and growth results are stated to be in agreement with experimental observations from reference [57], which is listed as 'to be published'; this comparison cannot currently be verified, so please provide the supporting data or cite a published source.","section":"Section 4 and reference [57]"},{"comment":"The name 'Behler and Perrinello' is a typo; it should read 'Behler and Parrinello.'","section":"Abstract and page 2"},{"comment":"The RMS errors reported in the caption are computed using 1320 randomly selected snapshots from the training data set; please clarify that these are training errors, not held-out test errors, and consider reporting errors on a separate validation set for a more meaningful accuracy assessment.","section":"Fig. 3"}],"recommendation":"major_revision","confidential_remarks":"The central issue is the internal inconsistency between the claim that melting temperatures are not explicitly included in training and the detailed description of adding solid-liquid transition configurations to the training set in Sections 2 and 4. This is a load-bearing point because the manuscript uses the 5-10% melting-temperature agreement as its principal evidence of transferability. The non-spin-polarized DFT treatment of La is a further correctness concern that the editor should ask the authors to address substantively. If these points are not satisfactorily resolved, the manuscript's central claims would need to be substantially weakened."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThis paper is worth a look if you work on MLIPs for complex ternaries or on La-Si-P chemistry specifically. The authors deliver the first ML interatomic potential for the La–Si–P system, trained with DeePMD-kit on a broad set of crystalline and liquid structures, and they show it reproduces DFT energy–volume curves and liquid pair-correlation functions well. The iterative training scheme, with usage-probability reweighting and targeted addition of transition-state structures, is a sensible incremental contribution. The nucleation-and-growth simulations of LaP from La-Si-P liquids look reasonable and match experimental synthesis observations.\n\nThe real soft spot is the melting-temperature validation. The paper says in Sections 2 and 4 that they added solid–liquid transition configurations to the training set and that this 'significantly' improved the predicted Tm, then Section 3 calls the 5-10% agreement with experiment 'remarkable' because melting temperatures were 'not explicitly included in training.' That is internally inconsistent. Configurations along the crystal-to-liquid path are exactly where the free-energy balance is decided, so adding them to the training set means the Tm agreement is partly a fit to the relevant region of the PES, not an independent prediction. This doesn't sink the paper—the potential still captured the right trend, and the systematic underestimation is honestly discussed—but the claim should be reframed. An ablation showing Tm behavior with and without those melting-path snapshots would settle the matter.\n\nOther issues: the DFT reference is non-spin-polarized PBE for a La system with 4f electrons, which could skew energetics and thus the potential; the authors don't discuss this. No code or data are released, just 'available upon request,' which limits reproducibility. And the ~20% Tm misses for LaSi and La show the potential's transferability is uneven, which the authors acknowledge.\n\nOverall, the paper is a useful, honest piece of engineering for a system that needed a potential. The central contribution is solid, and the validation, while partly circular, is not fatally so. I'd send it to a serious referee, and I'd ask for the Tm claim to be corrected and for a little more detail on the training data composition and the spin treatment.\n\nRecommendation: engage with it, but push for the revision.","headline":"A genuinely useful first MLIP for La–Si–P, but the melting-temperature validation is less independent than claimed because transition configurations were added to the training set.","tokens_in":12674,"tokens_out":2571,"would_cite":false,"duration_ms":30174,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper claims that a neural-network machine-learning interatomic potential can be trained to accurately describe crystalline and liquid La-Si-P structures and to predict melting temperatures and LaP nucleation in this ternary system.","keywords":["machine learning interatomic potential","neural network","molecular dynamics","La-Si-P ternary system","melting temperature","nucleation and growth","iterative training","phase stability"],"falsifier":"Compare the training energies and forces against spin-polarized DFT calculations with an explicit treatment of lanthanum's 4f electrons for a representative subset of La-containing crystals and liquids; if the reference data shift enough to reorder competing phases or move predicted melting temperatures by more than a few percent, the transferability claim inherits a systematic error. A second check is to measure the melting temperature of LaP at controlled phosphorus pressure, since the potential predicts roughly 2690 K while no direct measurement exists above 1373 K.","tokens_in":11564,"feed_emoji":"⚛️","tokens_out":8626,"duration_ms":96811,"temperature":0.7,"pith_summary":"The paper claims that a neural-network machine-learning interatomic potential can be built for the ternary La-Si-P system with enough accuracy and transferability for molecular dynamics simulations of phase stability and transformations. Training data come from distorted crystal structures, liquid snapshots, and crystal-liquid transition configurations, and the model is refined iteratively by up-weighting poorly learned phases and adding new melting data. The resulting potential reproduces energy-volume curves for known elemental, binary, and ternary phases and liquid pair-correlation functions when checked against density-functional theory. It predicts the melting temperatures of two ternary compounds within about 5 to 10 percent of experiment, and molecular dynamics runs show LaP nucleating and growing from La-Si-P liquids as observed in synthesis. If correct, this makes molecular dynamics a practical tool for guiding the synthesis of new ternary La-Si-P compounds.","feed_headline":"Neural-network potential predicts La-Si-P melting within 10 percent","feed_subtitle":"Iterative training on crystals, liquids, and melting states produces a force field to guide synthesis of La-Si-P compounds.","key_machinery":"The central machinery is a local-environment neural network potential whose total energy is the sum of per-atom energies, so any structure can be evaluated quickly once the network is trained. The argument rides on the iterative training loop: the model is first fit to a broad set of distorted crystals and ab initio liquid snapshots, then re-trained with adaptively increased weights for phases it fits poorly, and finally supplemented with snapshots collected at the solid-liquid transition so the potential learns barriers and molten states rather than only equilibrium structures. That last step is what rescues the melting-temperature predictions, as shown by the La2SiP4 case where adding transition configurations raises the predicted melting point substantially.","core_discovery":"The central claim is that an accurate and transferable artificial-neural-network machine-learning interatomic potential can be developed for the La-Si-P system. The potential represents the total energy as a sum of atomic energies, each obtained by passing a local environment descriptor, namely neighbor positions within a 7.0 Å cutoff, through a filter network and then through four hidden layers of 120 nodes; forces follow as derivatives of the total energy. The training set contains about 71,800 distorted crystal structures and 130,000 liquid snapshots, plus configurations added later from solid-liquid transition simulations. On the final model, the energy RMS error is about 12 meV per atom and the force error about 0.23 eV per angstrom. The model reproduces the energy-volume curves of all known La-Si-P crystalline phases and the pair-correlation functions of La-Si-P liquids at 2500 K against ab initio results. It yields melting temperatures of the Aea2 polymorph of LaSiP3 and of La2SiP4 that are 5.2 percent and 9.8 percent below the measured values, and it captures LaP nucleation and growth from undercooled liquids at 1400 K.","pith_inferences":["This suggests the same iterative data-generation strategy could be transferred to other rare-earth pnictide ternaries, where synthesis is often limited by unknown liquid-phase kinetics rather than by thermodynamic stability alone.","Because the reference data come from non-spin-polarized DFT for a lanthanum-containing system, the potential may carry a systematic bias from the treatment of 4f electrons; testing against spin-polarized reference calculations would reveal the size of that bias.","Given the systematic low bias in melting temperatures, relative stability rankings and crystallization trends from this potential are more trustworthy than absolute melting points, and a small set of experimental anchor melting temperatures could calibrate the model for quantitative use."],"forward_implications":["Molecular dynamics studies of the La-Si-P system can now be run at sizes and time scales that ab initio MD cannot reach, allowing phase competition and transformation kinetics to be studied directly.","The predicted melting temperatures of the ternary compounds are systematically low by 5 to 10 percent, so the potential is reliable for trends and for guiding synthesis conditions rather than for exact thermodynamic values.","The simulations reproduce LaP and Si-substituted LaP crystallization from undercooled La-Si-P liquids, consistent with the experimental observation that LaP nucleates readily during synthesis.","The iterative training recipe, adaptive re-weighting of existing data plus addition of crystal-liquid transition configurations, is what carries the transferability, and the same recipe can be applied to other ternary systems.","For binary and elemental phases, larger melting-temperature errors of up to about 20 percent indicate that those transitions need dedicated training data before the potential is used quantitatively there."],"supporting_citations":[{"why":"Establishes the neural-network representation of high-dimensional potential-energy surfaces that this potential is built on.","marker":"[3]"},{"why":"Provides the deep-learning molecular dynamics model architecture and training scheme used here.","marker":"[14]"},{"why":"Shows that adding rare-event configurations to the training set improves ML potentials, justifying the melting-transition data addition.","marker":"[44]"},{"why":"Supplies the known and predicted parent crystal structures from which distorted training configurations were generated.","marker":"[46]"},{"why":"Underlies the plane-wave DFT reference calculations used to label energies and forces.","marker":"[47]"},{"why":"Defines the non-spin-polarized GGA exchange-correlation functional whose energies are the training reference.","marker":"[49]"},{"why":"Provides the solid-liquid coexistence method used to extract predicted melting temperatures.","marker":"[53]"},{"why":"Reports the experimental melting temperature of the Aea2 LaSiP3 phase used to benchmark the prediction.","marker":"[33]"},{"why":"Documents the experimental synthesis observations against which the simulated LaP nucleation and growth are compared.","marker":"[57]"}],"fun_headline_variants":["La-Si-P simulations get a machine-learning interatomic potential","Neural network potential models La-Si-P: Melting and nucleation","Accurate ML interatomic potential for La-Si-P dynamics","Neural network force field for La-Si-P crystals and liquids"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the non-spin-polarized DFT calculations used to label the training data are accurate enough for lanthanum-containing phases; if those reference energies and forces are biased, the neural network faithfully learns the bias.","fun_headline_variants_meta":{"raw":{"variants":["La-Si-P simulations get a machine-learning interatomic potential","Neural network potential models La-Si-P: Melting and nucleation","Accurate ML interatomic potential for La-Si-P dynamics","Neural network force field for La-Si-P crystals and liquids"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000865,"raw_usage":{"total_tokens":3800,"prompt_tokens":1046,"completion_tokens":2754,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":662,"completion_tokens_details":{"reasoning_tokens":2684}},"tokens_in":662,"tokens_out":2754,"duration_ms":25058,"temperature":1.0,"reasoning_tokens":2684,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T05:13:15.309355+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compare the training energies and forces against spin-polarized DFT calculations with an explicit treatment of lanthanum's 4f electrons for a representative subset of La-containing crystals and liquids; if the reference data shift enough to reorder competing phases or move predicted melting temperatures by more than a few percent, the transferability claim inherits a systematic error. A second check is to measure the melting temperature of LaP at controlled phosphorus pressure, since the potential predicts roughly 2690 K while no direct measurement exists above 1373 K.","supporting_citations":[{"cited_title":"Kresse, J","cited_arxiv_id":null,"evidence_quote":"Underlies the plane-wave DFT reference calculations used to label energies and forces."}],"review_version":1}