{"id":"cb866a2a-9daf-4e1a-b76e-4a5e3e1c46a1","arxiv_id":"2501.01607","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"A DFT high-throughput screen of 49 ferromagnetic Fe-N compounds finds 15 permanent-magnet candidates with anisotropy above 0.4 MJ/m3 and 41 potential magnetocaloric alloys.","lead":"A high-throughput DFT screening of Fe-N alloys predicts 15 potential rare-earth-free permanent magnets and 41 potential magnetocaloric candidates. The best candidate, triclinic Fe16N3, shows a calculated magnetic anisotropy of 1.75 MJ/m3, comparable to known MnAl and FePd magnets.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Abstract claims 15 compounds with MAE > 1 MJ/m3, but Table A1 shows only 4; headline number contradicts its own data.","rationale":"The reader's weakest assumption focuses on the metastability of the USPEX-derived structures. That is a legitimate scientific concern for a screening paper, but it is not the most immediately load-bearing issue: metastable phases are routinely synthesized (e.g., α''-Fe8N is itself metastable), and the paper's validation against known compounds suggests the screening protocol can find real phases. The more unambiguous problem is that the paper's own headline number is contradicted by its own table. The abstract says 15 compounds have MAE > 1 MJ/m3, while Table A1, explicitly defined as MAE > 0.4 MJ/m3, contains only four such values. This is a factual, verifiable inconsistency that directly affects how the central claim is received: a reader relying on the abstract would overstate the number of high-anisotropy candidates by nearly fourfold. Since the reader's verdict is already CONDITIONAL, this finding does not change the verdict but reinforces the need for correction and careful verification before the results can be accepted. I agree partially with the reader: the metastability concern is real, but the internal inconsistency is the most concrete and actionable issue that should be settled first.","tokens_in":11704,"tokens_out":3177,"duration_ms":31776,"concrete_test":"Count the entries in Table A1 with MAE > 1 MJ/m3 (using either the MJ/m3 or meV/Fe columns) and compare to the abstract's claim of '15 compounds ... more than 1 MJ/m3'. Also count the entries in Table A2 with ΣM > 1.5% and compare with the numbers 40 and 41 reported in Section 3.2. If the counts do not match the stated figures (4 vs 15, and either 40 or 41 vs the table count), the headline numbers must be corrected and the abstract/conclusion aligned with the data.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's headline claim is internally inconsistent with its own data. The abstract states '15 compounds are potential permanent magnets with a magneto-crystalline anisotropy energy more than 1 MJ/m3', but Table A1, which is described as containing compounds with MAE larger than 0.4 MJ/m3, lists only four compounds exceeding 1 MJ/m3 (1.751, 1.384, 1.300, and 1.047 MJ/m3). The remaining eleven entries have MAE between 0.42 and 0.94 MJ/m3. The conclusion correctly states 'larger than 0.4 MJ/m3'. Thus the advertised number of high-MAE candidates is overstated by a factor of about 3.75 in the abstract. This matters because the application-level claim of 'filling the gap' relies on the 0.4 MJ/m3 threshold, while readers may reasonably take the abstract's 1 MJ/m3 figure as the headline result. The MCE section suffers from a similar counting inconsistency: it reports '40 newly potential' candidates, then 'Overall, we have found 41 compounds with a magnetic deformation > 1.5%', while the conclusion says 40, and Table A2 includes known α-Fe, Fe3N, and Fe4N among the listed entries. These errors do not necessarily invalidate the underlying DFT screening, but they make the central claim as presented unreliable and must be resolved before the results can be assessed.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports a high-throughput density functional theory screening of FexN1-x compounds, using 49 ferromagnetic structures previously predicted by the authors via USPEX. For each candidate the authors compute the magnetocrystalline anisotropy energy (MAE), saturation magnetization, and a magnetic-deformation proxy for the magnetocaloric effect (MCE). They identify 15 compounds with MAE larger than 0.4 MJ/m3 as potential rare-earth-free permanent magnets, with Fe16N3 having the largest out-of-plane MAE of 1.751 MJ/m3, and they report 40 (or 41) potential MCE candidates with magnetic deformation greater than 1.5%. The results are compared with known α''-Fe8N and ε-Fe3N data as validation.","tokens_in":1595,"tokens_out":1502,"duration_ms":34059,"significance":"If the predictions are reliable, the work would expand the family of rare-earth-free permanent-magnet candidates and identify new magnetocaloric materials in a chemically simple alloy system. The authors correctly use established computational proxies (MAE from total-energy differences with spin-orbit coupling, and the magnetic-deformation criterion introduced by Bocarsly et al.) and they validate two known phases against experiment and prior theory. The screening is systematic and the data tables are extensive. However, the abstract overstates the number of compounds with MAE above 1 MJ/m3, the MCE candidate counts are inconsistent, and the MAE values lack convergence tests or error estimates. These issues currently prevent the central claims from being assessed as stated.","major_comments":[{"comment":"The abstract states that '15 compounds are potential permanent magnets with a magneto-crystalline anisotropy energy more than 1 MJ/m3', but Table A1, which lists compounds with MAE larger than 0.4 MJ/m3, contains only four entries above 1 MJ/m3 (1.751, 1.384, 1.300, and 1.047 MJ/m3). The remaining eleven entries have MAE values between 0.42 and 0.94 MJ/m3. The conclusion correctly uses the 0.4 MJ/m3 threshold, so the abstract's headline number is inconsistent with the paper's own data and should be corrected before publication.","section":"Abstract and Table A1"},{"comment":"The MCE candidate counts are inconsistent: Section 3.2 first says 'We have found 40 newly potential MCE candidates', then 'Overall, we have found 41 compounds with a magnetic deformation ΣM > 1.5%', while the Conclusions state '40 newly potential giant magnetocaloric effect candidates'. Table A2 includes α-Fe, ε-Fe3N, and γ'-Fe4N among the listed entries, so it is unclear whether the counts refer to new candidates, all candidates, or compounds excluding known references. The text should state the exact counting convention and reconcile the 40/41 discrepancy.","section":"Section 3.2, Table A2, and Conclusions"},{"comment":"The MAE values are presented without convergence tests or error estimates, which is a serious concern because many of the key values (0.4-1.7 MJ/m3) are small energy differences obtained from total energies with spin-orbit coupling. The authors should provide at least a convergence check of the MAE with respect to k-point density and plane-wave cutoff for representative compounds (e.g., Fe16N3, Fe2N, and α''-Fe8N), and ideally estimate numerical uncertainties or compare results obtained with different exchange-correlation functionals.","section":"Section 2 and Section 3.1"},{"comment":"All 49 ferromagnetic structures are taken from the authors' own USPEX prediction (Refs. [27,28]) and are filtered by a convex-hull cutoff of 0.075 eV/at. Many candidates in Table A1 have hull energies of 0.02-0.066 eV/at (e.g., Fe16N3: 0.0464, Fe2N: 0.0511, Fe4N: 0.0519), meaning they are metastable with respect to decomposition. The manuscript should explicitly discuss the synthesizability and kinetic stability of these phases, since the application-relevant claims depend on the candidates being experimentally accessible.","section":"Section 2 and Table A1"}],"minor_comments":[{"comment":"In the introduction to Section 3.1, the text refers to 'an enhanced MAE in the order of 700-800 eV/at' for Fe-Co systems; this appears to be a typo, as the reported MAE values for tetragonal Fe-Co are on the order of 0.7-0.8 meV/atom, not hundreds of eV/atom.","section":"Section 3.1"},{"comment":"References [27] and [28] refer to the same manuscript (one arXiv, one SSRN). The authors should cite the published version or, if neither is peer-reviewed, state clearly that the parent structure prediction is a preprint and describe its availability in the main text.","section":"References [27] and [28]"},{"comment":"The tables use 'TRI', 'HEX', 'ORT', etc. as structure-type abbreviations, but the caption defines them only in a list. This is adequate, but the MAE units in Table A1 (MJ/m3 and meV/Fe) should be explicitly separated in the header, as the current format '1.7509 (0.1389)' may be ambiguous without the long caption.","section":"Table A1"},{"comment":"The figure caption states 'The filled blue diamond and red circle symbols represent the data sets of well-known permanent magnets and the predicted FexN1-x alloys, respectively', but it does not identify which symbol corresponds to which data set in the figure itself. Please ensure the legend is readable in the final version.","section":"Figure 1"}],"recommendation":"major_revision","confidential_remarks":"The manuscript's central screening results may be sound, but the abstract's misstatement of the MAE threshold and the inconsistent MCE counts could mislead readers. The requirement for convergence tests on MAE values is standard for this type of claim; without them, the reliability of the small energy differences is unverified. The reliance on the authors' own unpublished structure predictions is a legitimate sequential workflow, not circularity, but the editors may wish to confirm that Ref. [27] has undergone independent scrutiny before the present paper is accepted."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: this is a legitimate high-throughput DFT screening that produces a useful data table of MAE and magnetic deformation for 49 Fe-N compounds, with one standout candidate (Fe16N3, 1.75 MJ/m3 out-of-plane MAE, 1603 emu/cm3 MS). It is not conceptually new—the screening framework, thresholds, and the MCE proxy are all borrowed from prior work, including the authors' own USPEX structure search—but the specific results are new and, on the surface, calculated cleanly.\n\nWhat it does well: the validation against alpha''-Fe8N and epsilon-Fe3N is appropriate, the MS and MAE values are first-principles outputs not fits, and the paper is transparent about the metastability of candidates (Delta_EH listed in Tables A1/A2). The hardness classification (kappa) gives a useful metric. The application-spectrum plot (Fig. 1) is a nice presentation.\n\nThe soft spots, in order of real weight:\n\n1. The abstract says \"15 compounds are potential permanent magnets with a magneto-crystalline anisotropy energy more than 1 MJ/m3\"—that is wrong. Table A1 lists 15 compounds, but only four exceed 1 MJ/m3 (1.75, 1.38, 1.30, 1.05). The conclusion correctly says 0.4 MJ/m3. This is an internal contradiction in the headline claim, and it will mislead anyone who reads only the abstract. It is easy to fix, but right now it is the first thing a referee will hit.\n\n2. The computational details never name the exchange-correlation functional. VASP, PAW, cutoff energy, and k-point density are given, but no PBE/LDA label. That is a reproducibility gap and should have been caught.\n\n3. The MCE claim rests entirely on the magnetic-deformation proxy from Bocarsly et al., not on entropy change. The authors are open about this (\"proxy\") and they validate against epsilon-Fe3N and gamma'-Fe4N, so the screening is defensible, but the phrase \"potential giant magnetocaloric effect candidates\" overstates what the proxy can certify. There is also a counting slip: 40 vs 41 in Section 3.2.\n\n4. No convergence tests or error bars for the MAE values, some of which are close to the threshold (e.g., 0.42 MJ/m3). For screening that is acceptable, but it should be stated.\n\nThe central argument—that Fe-N phases offer a family of semi-hard/hard magnets and MCE candidates worth experimental attention—holds up. The flaws are presentation and missing details, not load-bearing. This paper deserves a serious referee; the proposed fixes are minor but necessary.","headline":"Solid DFT screening of 49 Fe-N compounds with one standout candidate, but the abstract's '15 compounds above 1 MJ/m3' overstates the data (only 4) and needs fixing before it can be taken at face value.","tokens_in":12541,"tokens_out":2775,"would_cite":true,"duration_ms":26499,"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":"High-throughput DFT screening of FexN1-x identifies 15 rare-earth-free permanent-magnet candidates and 40 magnetocaloric-effect candidates.","keywords":["permanent magnet","magnetocrystalline anisotropy energy","saturation magnetization","magnetocaloric effect","iron nitride","high-throughput density functional theory","rare-earth-free magnets","Fe-N alloys"],"falsifier":"Synthesize phase-pure triclinic $\\mathrm{Fe}_{16}\\mathrm{N}_{3}$ or orthorhombic $\\mathrm{Fe}_{2}\\mathrm{N}$ by low-temperature nitriding and measure the easy-axis anisotropy and magnetization; if the measured MAE comes out below 0.4 MJ/m³, or the phase decomposes into $\\mathrm{Fe}_{4}\\mathrm{N}$/$\\mathrm{Fe}_{3}\\mathrm{N}$ and Fe on the timescale of the measurement, the screening's headline prediction is refuted.","tokens_in":11468,"feed_emoji":"🧲","tokens_out":9948,"duration_ms":88126,"temperature":0.7,"pith_summary":"These calculations argue that the Fe-N alloy system, built from abundant iron and nitrogen, can supply both rare-earth-free permanent magnets and giant magnetocaloric materials. High-throughput density functional theory on 49 ferromagnetic $\\mathrm{Fe}_x\\mathrm{N}_{1-x}$ structures finds 15 compounds with at least one magnetocrystalline anisotropy energy above 0.4 MJ/m³, four of them above 1 MJ/m³, which the paper says fills the application gap between expensive high-performance magnets and cheap ferrites. The best candidate, triclinic $\\mathrm{Fe}_{16}\\mathrm{N}_{3}$, shows an out-of-plane MAE of 1.751 MJ/m³ and saturation magnetization of 1603 emu/cm³, approaching Nd-Fe-B in magnetization. The same list yields 40 compounds with magnetic deformation above 1.5%, the proxy for giant magnetocaloric response, and four of the strongest permanent-magnet candidates also appear there. A sympathetic reading is that Fe-N is a promising, earth-abundant playground for both applications, though many candidates are energetically metastable.","feed_headline":"DFT screen finds 15 Fe-N compounds for rare-earth-free magnets","feed_subtitle":"Top candidate Fe16N3 shows 1.75 MJ/m3 anisotropy and a magnetization near Nd-Fe-B's.","key_machinery":"The load-bearing machinery is the magnetocrystalline anisotropy energy (MAE) computed as the total-energy difference between magnetization orientations, $K_{\\hat{n}_1-\\hat{n}_2}=E_{\\hat{n}_1}-E_{\\hat{n}_2}$, evaluated for the three crystal directions [001], [010], and [100]; the paper's headline quantity is the largest absolute value $|K_{\\max}|$ among these. Candidate selection combines three filters: thermodynamic stability (negative formation energy and hull energy below 0.075 eV/at), mechanical and dynamical stability from the prior structure search, and an anisotropy threshold of 0.4 MJ/m³. For magnetocaloric screening it uses the magnetic deformation proxy $\\Sigma_M=\\frac{1}{3}\\sqrt{\\eta_1^2+\\eta_2^2+\\eta_3^2}\\times 100$, with $\\boldsymbol{\\eta}=\\frac{1}{2}(\\boldsymbol{P}^T\\boldsymbol{P}-\\boldsymbol{I})$ the Lagrangian finite strain between non-magnetic and magnetic states, and the 1.5% threshold from the literature. Magnets are then labelled by the dimensionless hardness parameter $\\kappa=\\sqrt{K_1/(\\mu_0 M_S)^2}$, with $\\kappa>1$ hard and $0.1<\\kappa<1$ semi-hard.","core_discovery":"On the paper's terms, the central discovery is that nitrogenation of iron yields a family of usable anisotropic magnets rather than a single accidental phase. From 49 ferromagnetic $\\mathrm{Fe}_x\\mathrm{N}_{1-x}$ structures, 15 have at least one computed magnetocrystalline anisotropy energy above 0.4 MJ/m³, with four above 1 MJ/m³; $\\mathrm{Fe}_{2}\\mathrm{N}$ is classified as a hard magnet ($\\kappa>1$) and the other 14 as semi-hard ($0.1<\\kappa<1$). The largest out-of-plane anisotropy, 1.751 MJ/m³ for $\\mathrm{Fe}_{16}\\mathrm{N}_{3}$ (#2111), is more than twice that of $\\alpha''$-$\\mathrm{Fe}_{8}\\mathrm{N}$ and comparable to L1₀ MnAl and FePd, while its saturation magnetization of 1603 emu/cm³ approaches $\\mathrm{Nd}_{2}\\mathrm{Fe}_{14}\\mathrm{B}$. The same screening, using the magnetic deformation proxy, flags 40 compounds as potential giant magnetocaloric materials, with $\\mathrm{Fe}_{7}\\mathrm{N}$ showing a 9.36% deformation, and places the set in the application gap between high-performance and widely used permanent magnets.","pith_inferences":["An implicit extension the authors do not develop: because most candidates sit 0.02–0.066 eV/at above the hull, the realistic synthesis routes are epitaxial growth or low-temperature nitriding, not bulk metallurgy; the list functions as a target library.","MAE is an intrinsic upper bound, not actual coercivity; if microstructural reversal mechanisms dominate, the hard/semi-hard labels should be read as intrinsic-crystal classifications, not device guarantees.","The magnetic-deformation proxy selects magnetoelastic coupling, but a working magnetocaloric refrigerator also needs the transition near the operating temperature; Curie temperatures for the new candidates are not computed here, so the 40-name list is a starting set.","A cheap falsification route would be to grow $\\mathrm{Fe}_{16}\\mathrm{N}_{3}$ on a lattice-matched substrate and measure ferromagnetic resonance; agreement with 1.75 MJ/m³ would validate the entire chain."],"forward_implications":["Fifteen Fe-N compounds become concrete synthesis targets for rare-earth-free permanent magnets aimed at the gap between ferrites/AlNiCo and Nd-Fe-B/Sm-Co.","$\\mathrm{Fe}_{2}\\mathrm{N}$ is predicted to be a hard magnet ($\\kappa>1$), meaning its coercivity is less shape-dependent than the semi-hard candidates.","$\\mathrm{Fe}_{16}\\mathrm{N}_{3}$ offers a computed MAE (1.751 MJ/m³) more than twice that of $\\alpha''$-$\\mathrm{Fe}_{8}\\mathrm{N}$ and a saturation magnetization close to $\\mathrm{Nd}_{2}\\mathrm{Fe}_{14}\\mathrm{B}$, making it the flagship candidate.","Forty Fe-N compounds pass the magnetocaloric proxy ($\\Sigma_M>1.5\\%$), and the four strongest permanent-magnet candidates also appear on the MCE list, suggesting dual-use materials.","Saturation magnetization rises approximately linearly with Fe content, giving a simple composition rule for tuning $M_S$ in $\\mathrm{Fe}_x\\mathrm{N}_{1-x}$."],"supporting_citations":[{"why":"supplies the 49 ferromagnetic Fe-N crystal structures and the stability filters (negative formation energy, hull energy <0.075 eV/at, mechanical and dynamical stability) on which all MAE and MCE screening is based.","marker":"[27]"},{"why":"defines the magnetic deformation proxy $\\Sigma_M$ and the 1.5% threshold used to label the 40 magnetocaloric candidates.","marker":"[26]"},{"why":"provides the experimental/computed $\\alpha''$-$\\mathrm{Fe}_{8}\\mathrm{N}$ MAE (0.775 meV/f.c.) and saturation magnetization (1750 emu/cm³) used to validate the DFT workflow.","marker":"[24]"},{"why":"supplies the hard/semi-hard/soft classification by $\\kappa$ used to call $\\mathrm{Fe}_{2}\\mathrm{N}$ a hard magnet and the other 14 candidates semi-hard.","marker":"[36]"},{"why":"defines the dimensionless figure of merit $\\kappa = \\sqrt{K_1/(\\mu_0 M_S)^2}$ used to classify the predicted magnets.","marker":"[43]"},{"why":"frames the application-spectrum gap between high-performance rare-earth magnets and cheap ferrites that the 15 candidates are claimed to fill.","marker":"[7]"}],"fun_headline_variants":["Fe-N screening yields 15 anisotropic magnets for rare-earth-free tech","Fe16N3: 1.75 MJ/m3 anisotropy, near-Nd-Fe-B magnetization","Nitrogenated iron unlocks 15 rare-earth-free magnet candidates","49 ferromagnetic Fe-N phases: 15 magnets, 40 magnetocaloric","Fe-N: 15 anisotropic magnets, no rare earths needed"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The candidate list assumes that the 49 Fe-N structures inherited from the earlier structure search are real, synthesizable phases, even though many sit slightly above the energy of competing phase mixtures; if any of them cannot be made, its computed anisotropy and magnetization values cease to matter.","fun_headline_variants_meta":{"raw":{"variants":["Fe-N screening yields 15 anisotropic magnets for rare-earth-free tech","Fe16N3: 1.75 MJ/m3 anisotropy, near-Nd-Fe-B magnetization","Nitrogenated iron unlocks 15 rare-earth-free magnet candidates","49 ferromagnetic Fe-N phases: 15 magnets, 40 magnetocaloric","Fe-N: 15 anisotropic magnets, no rare earths needed"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001336,"raw_usage":{"total_tokens":5446,"prompt_tokens":971,"completion_tokens":4475,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":587,"completion_tokens_details":{"reasoning_tokens":4374}},"tokens_in":587,"tokens_out":4475,"duration_ms":30373,"temperature":1.0,"reasoning_tokens":4374,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T22:24:05.747397+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Synthesize phase-pure triclinic $\\mathrm{Fe}_{16}\\mathrm{N}_{3}$ or orthorhombic $\\mathrm{Fe}_{2}\\mathrm{N}$ by low-temperature nitriding and measure the easy-axis anisotropy and magnetization; if the measured MAE comes out below 0.4 MJ/m³, or the phase decomposes into $\\mathrm{Fe}_{4}\\mathrm{N}$/$\\mathrm{Fe}_{3}\\mathrm{N}$ and Fe on the timescale of the measurement, the screening's headline prediction is refuted.","supporting_citations":[{"cited_title":"Bocarsly, E.E","cited_arxiv_id":null,"evidence_quote":"defines the magnetic deformation proxy $\\Sigma_M$ and the 1.5% threshold used to label the 40 magnetocaloric candidates."},{"cited_title":"Zhang, I","cited_arxiv_id":null,"evidence_quote":"provides the experimental/computed $\\alpha''$-$\\mathrm{Fe}_{8}\\mathrm{N}$ MAE (0.775 meV/f.c.) and saturation magnetization (1750 emu/cm³) used to validate the DFT workflow."},{"cited_title":"Mohapatra, X","cited_arxiv_id":null,"evidence_quote":"supplies the hard/semi-hard/soft classification by $\\kappa$ used to call $\\mathrm{Fe}_{2}\\mathrm{N}$ a hard magnet and the other 14 candidates semi-hard."},{"cited_title":"Coey, New permanent magnets; manganese compounds, J","cited_arxiv_id":null,"evidence_quote":"defines the dimensionless figure of merit $\\kappa = \\sqrt{K_1/(\\mu_0 M_S)^2}$ used to classify the predicted magnets."}],"review_version":1}