{"id":"241b698c-7b45-4592-b8b4-3d0ea6c54f2c","arxiv_id":"2607.21092","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"DFT calculations predict a ~300% tunnel magnetoresistance in Cr-doped RuO2(110)/TiO2(110) magnetic tunnel junctions.","lead":"Using first-principles simulations, this paper predicts a tunnel magnetoresistance of about 300% in magnetic tunnel junctions made from chromium-doped RuO2(110), an antiferromagnet with a special 'altermagnetic' spin texture. The result suggests that antiferromagnets, not just ferromagnets, could be used in spintronic memory devices.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"TMR prediction rests on assumed altermagnetic order in Cr-doped RuO2, which is neither verified nor tested for stability.","rationale":"The reader's weakest_assumption correctly identifies that the entire calculation presupposes the altermagnetic ground state. This is the most load-bearing concern because the TMR ratio, the spin polarization, and all transport results are computed from electronic states that only exist if the collinear antiferromagnetic order is realized. The paper explicitly acknowledges the controversy over RuO2 magnetism but does not perform a stability check for the doped (110) system. VCA further weakens the connection to real Cr-doped RuO2, as it cannot capture the local correlation effects that are claimed to stabilize the order. A concrete test using explicit supercells with DFT+U would either support or falsify this assumption. Since the paper is a prediction and the authors are transparent about the ongoing debate, the conditional verdict is appropriate; however, without this stability test, the ~300% TMR claim should not be treated as a robust physical prediction. Therefore, I agree with the reader's assessment and do not see the need to change the verdict.","tokens_in":13515,"tokens_out":4451,"duration_ms":44220,"concrete_test":"Compute DFT+U total energies for explicit Cr-substituted supercells (e.g., Ru1−xCrxO2 with x=0.45, 4 Cr atoms per 16 Ru sites) for nonmagnetic, ferromagnetic, and the assumed collinear altermagnetic orders, scanning U from 0 to 4 eV on Ru/Cr d-states. If the altermagnetic state is not the lowest energy (or not even locally stable), the predicted TMR ratio is not physically supported. Also compare with VCA results at the same U to test the alloy approximation.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim of a ~300% TMR in Ru1−xCrxO2(110)/TiO2(110)/Ru1−xCrxO2(110) depends entirely on the existence of a collinear altermagnetic ground state. The paper itself states in the Introduction that the magnetism of RuO2 is under debate (Refs [44–48] report nonmagnetic behavior), yet Section II imposes the altermagnetic structure without a magnetic-stability calculation. No total-energy comparison between the assumed AFM order and a nonmagnetic (or other) state is presented for the doped (110) system. Compounding this, Cr substitution is described as stabilizing altermagnetism via enhanced correlation, but the calculation uses the virtual crystal approximation (VCA) with PBE, which cannot capture the local correlation enhancement or the disorder of Cr on Ru sites. If the real ground state is nonmagnetic or has a different magnetic order, the spin-split bands and the resultant TMR vanish. The paper provides no evidence that the assumed order is even metastable in the VCA calculation, let alone the true ground state.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports first-principles DFT and Landauer–Büttiker transport calculations for Ru1−xCrxO2(110)/TiO2(110)/Ru1−xCrxO2(110) magnetic tunnel junctions with altermagnetic electrodes. The authors find a finite bulk transport spin polarization in Ru1−xCrxO2(110), and compute a TMR ratio that reaches about 300% at x = 0.45 for a four-monolayer TiO2(110) barrier, larger than their earlier result for the (001) orientation. They attribute the TMR to momentum-dependent spin splitting, interfacial magnetic configurations, and bulk spin-polarized current, and discuss barrier-thickness oscillations. The central assumption is that the collinear altermagnetic state, which is imposed at the outset, is the actual ground state of the doped system.","tokens_in":13808,"tokens_out":4631,"duration_ms":40918,"significance":"If the predicted altermagnetic order and the computed TMR are robust, this paper would provide a concrete materials realization of antiferromagnetic TMR in a (110)-oriented rutile oxide, with a ratio exceeding that of the (001) junction. The calculations are parameter-free in the sense that no TMR is fitted to experiment, and the computational setup is standard and well documented. The physical interpretation based on interfacial magnetic configurations and momentum-dependent spin splitting is insightful. However, the predictive value is contingent on the altermagnetic ground state being thermodynamically stable in Cr-doped RuO2, which the paper does not establish. Given the ongoing debate about the magnetic ground state of RuO2, this conditionality is central to the paper's central claim.","major_comments":[{"comment":"The entire TMR prediction assumes a collinear altermagnetic order with two inequivalent Ru/Cr sites. This is imposed as the input electronic structure, with no total-energy comparison to a nonmagnetic or other magnetic state. The Introduction itself acknowledges Refs. [44–48] reporting nonmagnetic RuO2. A magnetic-stability calculation (e.g., total energy vs. magnetic moment or spin-constrained calculations) for at least the representative compositions x = 0.3, 0.45, 0.5 is needed to support the claim that Cr doping stabilizes the altermagnetic phase. Without this, the ~300% TMR is a prediction conditional on an unverified premise.","section":"Section II (System and method)"},{"comment":"The virtual crystal approximation (VCA) with PBE, together with fixed RuO2 lattice constants for all x, cannot capture the local Coulomb enhancement or the real disorder of Cr substitution. The paper states that Cr substitution reinforces electron correlation and supports altermagnetism, but this mechanism is not verified by the VCA calculation. The quantitative TMR values (including the peak at x = 0.45) may be sensitive to these approximations. The authors should either perform explicit supercell calculations for at least x = 0.25 or 0.5, or provide a sensitivity analysis showing that the VCA and rigid-lattice approximations do not qualitatively change the result.","section":"Section II (DFT details)"},{"comment":"The statement that 'p_tot shows a monotonic decrease when the chemical potential increases' is contradicted by Fig. 2(e), which shows ptot increasing with energy (the text there says the magnitude increases with chemical potential, reaching ~35%). This discrepancy should be corrected and the interpretation of the TMR-vs-energy comparison revised accordingly.","section":"Section III.B (text after Fig. 6)"},{"comment":"The TMR ratio is reported at a single k-mesh (151×151) for NTiO2 = 4 and 101×101 otherwise, and the ratio is highly sensitive to energy and barrier thickness. No convergence checks with respect to k-mesh or number of barrier layers are presented, and no error estimates are given. Because the central quantitative claim is the ~300% peak at x = 0.45, a convergence study (e.g., increasing the k-mesh to 201×201 for x = 0.45, and varying NTiO2 over a wider range) is needed to show that the peak is not a numerical artifact.","section":"Section III.B (Figs. 4 and 7)"}],"minor_comments":[{"comment":"Typographical errors: 'magnetizaton' and 'antiferromagentic' should be 'magnetization' and 'antiferromagnetic'.","section":"Abstract"},{"comment":"The interfacial distance between Ru1−xCrxO2(110) and TiO2(110) is set as the average of the bulk layer distances. This is a free parameter; at least a brief discussion of its sensitivity or a check with a relaxed interface would strengthen the results.","section":"Section II"},{"comment":"The spin index σ in Tσ(k∥, E) should be defined more precisely with respect to the electrode magnetization direction, especially for the antiparallel configuration where spin channels in the two electrodes are not globally aligned.","section":"Eq. (1)"},{"comment":"The schematic in Figs. 7(c) and 7(d) is helpful, but the description of in-plane coordinates (y-offset by aRuO2/2) is somewhat confusing in the caption. A clearer statement of the registry would improve readability.","section":"Fig. 7"}],"recommendation":"major_revision","confidential_remarks":"The paper builds on the authors' prior work [26] and the general altermagnet TMR framework. The main concern is the unverified magnetic ground state: the altermagnetic order is imposed, not predicted, while the literature the authors cite reports nonmagnetic RuO2. I would ask for a magnetic-stability check and a robustness analysis before publication. The transport calculations themselves are technically sound and the physical picture is clear; the result would be a useful contribution if the magnetic premise is justified."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: this is a solid, parameter-free DFT transport study with a genuinely new result — the (110) Cr-doped RuO2 stack gives a TMR peak near 300% at x = 0.45, exceeding both the undoped (110) and Cr-doped (001) stacks. The paper does a good job tracing the TMR to momentum-dependent spin polarization and interfacial sublattice alignment, and the barrier DOS check is a nice sanity test.\n\nThe soft spot is the one the authors themselves flag: RuO2 magnetism is under debate, and they simply assume the collinear altermagnetic order for all x. There is no total-energy comparison with a nonmagnetic or other magnetic state. That is a load-bearing assumption, because if Cr doping doesn't stabilize the order, the whole TMR prediction collapses. I'm not saying the assumption is wrong — prior work suggests Cr doping pushes the system toward altermagnetism — but a paper whose headline is a number should at least show the assumed phase is metastable in its own calculations.\n\nThe other approximations are in a less serious category: VCA for Cr substitution is a coarse way to treat a dopant whose main effect is supposed to be correlation enhancement; lattice constants are fixed at RuO2 values; there are no convergence checks or SOI estimates. None of these individually kill the result, but together they cap the quantitative confidence.\n\nOverall: the transport machinery is standard and carefully applied, the analysis is honest, and the novelty is real. It deserves a serious referee, and the referee should ask for stability checks and probably a supercell calculation at x = 0.45 to validate the VCA. I'd send it to review, with the expectation of a revision.","headline":"A useful computational prediction of ~300% TMR in Cr-doped RuO2(110) tunnel junctions, but the result hinges on an assumed altermagnetic order that the paper never tests.","tokens_in":14284,"tokens_out":2300,"would_cite":false,"duration_ms":24346,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["73.40.Gk","75.70.Cn","71.15.Mb"],"model":"deepseek-v4-flash","headline":"This paper predicts that a magnetic tunnel junction built from Cr-doped RuO2 with (110) crystal orientation acts as a spin-polarized switch despite having almost no net magnetization.","keywords":["tunnel magnetoresistance","altermagnet","RuO2","Cr doping","antiferromagnetic spintronics","magnetic tunnel junction","spin polarization","first-principles transport"],"falsifier":"Measure the TMR of an epitaxial Ru1−xCrxO2(110)/TiO2(110)/Ru1−xCrxO2(110) junction at x≈0.45 with, say, four TiO2 layers. If the ratio is far below the predicted ~300%, or if magnetization measurements (such as muon spin rotation or neutron diffraction) show no collinear altermagnetic order in Cr-doped RuO2 at this composition, the prediction fails. A related check: the TMR should oscillate between odd and even TiO2 layer counts; absence of such oscillation at the Fermi energy would contradict the interfacial-sublattice mechanism.","tokens_in":13448,"feed_emoji":"🧲","tokens_out":10144,"duration_ms":87360,"temperature":0.7,"pith_summary":"This paper predicts that a magnetic tunnel junction built from Cr-doped RuO2 with (110) crystal orientation acts as a spin-polarized switch despite having almost no net magnetization. Using first-principles density-functional and scattering calculations, the authors show that the collinear altermagnetic order of Ru1−xCrxO2 produces a momentum-dependent spin polarization, and in the (110) direction this gives a spin-polarized current that supports a tunnel magnetoresistance (TMR) effect. The TMR grows with Cr content and peaks at x=0.45 at around 300%, larger than the 100–200% found for the (001)-oriented junctions. The paper argues that the full effect cannot be read off from bulk spin polarization alone; momentum-resolved tunneling and the magnetic alignment at the electrode/barrier interface are equally decisive, and the latter causes an oscillation in the TMR with barrier thickness.","feed_headline":"Altermagnet junction tops 300% TMR at x=0.45 Cr doping","feed_subtitle":"The predicted 300% signal comes from electrodes with zero net magnetization—a path to stray-field-free spintronics.","key_machinery":"The key object is the rutile altermagnet Ru1−xCrxO2 stacked along the (110) direction. In the assumed collinear antiferromagnetic order, the two Ru/Cr sublattices are inequivalent, so the band structure has spin-split states that swap under a combined time-reversal and lattice-rotation operation—the defining property of an altermagnet. Along (110), the momentum-resolved spin polarization of the conduction channels does not cancel, so a spin-polarized current flows even though the net magnetization is zero. The transport analysis uses a scattering-theory formalism to compute momentum-resolved transmissions and polarizations, and the final TMR emerges from the interplay of these bulk quantitie","core_discovery":"The central claim is that a finite, sizable tunnel magnetoresistance emerges in Ru1−xCrxO2(110)/TiO2(110)/Ru1−xCrxO2(110) magnetic tunnel junctions with altermagnetic electrodes, reaching roughly 300% at the Fermi energy for Cr fraction x=0.45. This exceeds the 100–200% range reported for the (001)-oriented junctions, and it arises because the (110) orientation converts the momentum-dependent spin splitting of the altermagnet into a net spin-polarized bulk current. The parallel- and antiparallel-configuration transmissions are computed with a scattering-theory transport method, and the TMR ratio is traced to spin-polarized tunneling that depends on momentum, to the interfacial magnetic struc","pith_inferences":["If the altermagnetic order persists in thin-film form, these results suggest a practical all-antiferromagnetic tunnel junction design; the paper does not address magnetic switching or thermal stability, but the large ratio implies a measurable signal.","The odd/even barrier-thickness oscillation could be used experimentally to verify the interfacial mechanism: fabricate junctions with 3, 4, and 5 TiO2 layers and look for alternating TMR at the Fermi energy.","The same analysis should generalize to other d-wave-like altermagnets in tetragonal or rutile lattices, potentially yielding even higher ratios if the interface termination is engineered.","Because the calculation models Cr substitution with the virtual crystal approximation, a full supercell treatment that includes explicit disorder might shift the optimal concentration."],"forward_implications":["A large TMR (≈300%) is possible with compensated antiferromagnetic electrodes, offering a route to memory and sensor devices without stray magnetic fields.","The (110) orientation of the rutile structure outperforms (001), roughly doubling the TMR, so crystal orientation is a key design parameter for altermagnetic tunnel junctions.","Because the TMR peaks at an intermediate Cr concentration while bulk spin polarization falls monotonically, optimal device performance requires tuning both the electrode composition and the junction interface.","The predicted oscillation of the TMR with barrier thickness (odd vs even TiO2 layers) provides a clear, testable signature of the interfacial magnetic structure's role."],"fun_headline_variants":["Altermagnet TMR hits 300% without net magnetization","Zero-magnetization altermagnet junction delivers 300% TMR","Cr-doped RuO2 altermagnet TMR: 300% at x=0.45","Altermagnet junction: 300% TMR with zero net spin","300% TMR from a Cr-doped RuO2 altermagnet junction"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The predicted effect relies on the collinear altermagnetic order being the true magnetic ground state of Cr-doped RuO2(110), which the paper assumes without a stability calculation and which remains debated for pure RuO2.","fun_headline_variants_meta":{"raw":{"variants":["Altermagnet TMR hits 300% without net magnetization","Zero-magnetization altermagnet junction delivers 300% TMR","Cr-doped RuO2 altermagnet TMR: 300% at x=0.45","Altermagnet junction: 300% TMR with zero net spin","300% TMR from a Cr-doped RuO2 altermagnet junction"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000953,"raw_usage":{"total_tokens":3957,"prompt_tokens":856,"completion_tokens":3101,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":600,"completion_tokens_details":{"reasoning_tokens":3000}},"tokens_in":600,"tokens_out":3101,"duration_ms":20928,"temperature":1.0,"reasoning_tokens":3000,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-01T08:28:14.684505+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the TMR of an epitaxial Ru1−xCrxO2(110)/TiO2(110)/Ru1−xCrxO2(110) junction at x≈0.45 with, say, four TiO2 layers. If the ratio is far below the predicted ~300%, or if magnetization measurements (such as muon spin rotation or neutron diffraction) show no collinear altermagnetic order in Cr-doped RuO2 at this composition, the prediction fails. A related check: the TMR should oscillate between odd and even TiO2 layer counts; absence of such oscillation at the Fermi energy would contradict the interfacial-sublattice mechanism.","supporting_citations":[],"review_version":1}