{"id":"a8e43c24-50bb-4f4e-803f-6c4ae0233e41","arxiv_id":"2608.11615","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":29,"one_line_summary":"Graphene on a bulk SiO2 substrate acquires an anisotropic spin texture, and oxygen vacancies plus short-range disorder can drive spin relaxation to near-isotropy, matching measured anisotropies of 0.5 to 1.","lead":"Using first-principles simulations, this paper shows that a silicon dioxide (SiO2) support modifies electron spins in graphene in a more complex way than the standard Rashba model predicts. The simulations explain why experiments see spin lifetime anisotropies between 0.5 and 1, and point to substrate defects as a key cause.","discovery_kind":"first_principles","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The near-isotropic spin relaxation claim rests on one crystalline SiO2 interface model; a real amorphous substrate could average away the anisotropic spin texture and the local-SOC mechanism that produce it.","rationale":"I considered three candidate concerns: (i) crystalline versus amorphous interface, (ii) missing electron-phonon scattering for bulk SiO2, and (iii) the speculative local-SOC mechanism for oxygen vacancies. I agree with the reader that (i) is the most load-bearing. If the particular AA-stacked, Si-terminated, OH-passivated slab is not representative, then both routes to ζ≈1—the modified DP route through intervalley scattering of the nonuniform out-of-plane texture and the EY-like oxygen-vacancy route—are called into question, because both inherit the bond-resolved spin-orbit terms of Eq. (2) fitted to that one slab. Concern (ii) is a real gap, as the authors state in Sec. III C that bulk-slab phonons are unstable and e-ph is omitted; if included, it could pull ζ toward 1/2. But it affects the quantitative device comparison rather than the 'capable of yielding' claim, and it is already acknowledged. Concern (iii) is an interpretation label; the FPDMD impurity calculation would stand even if the local-SOC picture were refined. I nevertheless give credit: the TB parameters are fitted to DFT band structure and spin texture, not to spin lifetimes; the anisotropy emerges from spin dynamics; and two independent numerical approaches agree for the tested configurations. The conditional verdict is appropriate.","tokens_in":16068,"tokens_out":9633,"duration_ms":118145,"concrete_test":"Build an amorphous SiO2 slab by melt-quench ab initio or machine-learned-potential molecular dynamics, relax it in contact with graphene, and repeat the DFT/TB workflow of Secs. II–III: compute the K-point spin texture, fit Eqs. (1)–(2), and run the FPDMD oxygen-vacancy and TB puddle spin-transport calculations. If the ensemble-averaged in-plane anisotropy, the nonuniform out-of-plane spin component, and the EY-like ζ≈1 oxygen-vacancy result are not reproduced (e.g., ζ falls below 0.7), the crystalline-interface origin of the claim is confirmed as a model artifact. A cheaper first step is to run the same transport calculation with Eq. (2) parameters drawn from distributions over random bond environments rather than a single fitted set.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central claim—that SiO2 substrate SOC alone can yield spin lifetime anisotropies 0.5<ζ≲1—rests on the specific crystalline interface of Sec. II A and Fig. 1(b): a 2×2 graphene cell on a Si-terminated, OH-passivated crystalline SiO2 slab in AA stacking, chosen as the lowest-energy of a few tested terminations. The entire bulk-SiO2 result follows from this one configuration: the anisotropic in-plane and nonuniform out-of-plane spin textures in Fig. 2(f,h) are generated by the bond-resolved terms δt_kl, δλ_R,kl, and λ_in_R,kl in Eq. (2), whose fitted values (Table I) are at the few-meV/μeV scale, and the oxygen-vacancy EY-like spin relaxation (Fig. 4(a,b)) is computed for one vacancy in this periodic supercell. Real SiO2 substrates are amorphous, with random bond angles, silanol termination, trapped charges, and vacancies in varied local environments. It is not established that a random distribution of such local environments preserves either the preferred in-plane spin direction or the nonuniform out-of-plane component; the average spin-orbit field could revert to a Rashba-like helical form, giving ζ=1/2. Equally, the local-SOC mechanism invoked for oxygen vacancies may be specific to the ordered defect site chosen. The paper itself notes unstable slab phonons and does not sample interface configurations, so the quantitative comparison to measured ζ values is not robust to this modeling choice.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper investigates spin relaxation and its anisotropy in graphene supported by SiO2 substrates. The authors perform DFT for graphene on a 2D silica kagome layer and on a crystalline bulk SiO2 slab, extract a tight-binding model that includes bond-resolved anisotropic hopping and Rashba-like spin-orbit terms, and then compute spin lifetimes with first-principles density-matrix dynamics and with linear-scaling kernel-polynomial transport simulations in large systems containing electron-hole puddles. For the 2D SiO2 substrate they find a Rashba-type helical spin texture and D'yakonov-Perel' spin relaxation with anisotropy zeta = 1/2. For the bulk SiO2 substrate they find an anisotropic in-plane spin texture and a nonuniform out-of-plane spin component; the transport simulations yield 0.5 < zeta <= 1, with oxygen vacancies giving zeta ~ 1 through an Elliott-Yafet-like mechanism attributed to defect-induced local spin-orbit coupling, and short-range puddles with intervalley scattering giving 0.7 < zeta < 0.95. The central claim is that substrate spin-orbit coupling alone can produce near-isotropic spin relaxation in graphene on SiO2, offering a possible explanation for measured spin lifetime anisotropies close to 1.","tokens_in":16605,"tokens_out":7543,"duration_ms":85105,"significance":"If the claims hold, the manuscript provides a concrete alternative to magnetic-impurity or contact explanations for the near-isotropic spin lifetimes measured in SiO2-supported graphene spin valves, and it extends the standard Rashba picture of graphene/substrate interfaces by demonstrating how broken in-plane mirror symmetry can modify the spin texture. The strengths are the combination of two independent numerical approaches (first-principles density-matrix dynamics and linear-scaling tight-binding/KPM), the forward nature of the calculations (TB parameters are fitted to DFT band structures and spin textures, not to spin lifetimes or anisotropies), the consistency check against the analytic D'yakonov-Perel' model, and the falsifiable predictions for distinct disorder regimes. The bond-resolved TB Hamiltonian of Eq. (2) that reproduces the DFT spin texture is a useful methodological contribution. However, the quantitative comparison to experiments rests on a small number of crystalline interface configurations and on an inferred defect-induced local spin-orbit mechanism, so the generality of the central claim is not yet fully established.","major_comments":[{"comment":"The quantitative claim that a bulk SiO2 substrate generically yields 0.5 < zeta <= 1 is based on a single crystalline interface model: a 2x2 graphene cell on a Si-terminated, OH-passivated bulk SiO2 slab in AA stacking. Real SiO2 substrates are amorphous, and the anisotropic in-plane spin texture and the nonuniform out-of-plane component in Fig. 2(f,h), together with the fitted bond-resolved terms of Eq. (2), could average away under random bond angles, terminations, steps, and trapped charges. No ensemble of interface configurations or an amorphous model is studied. The authors should either restrict their conclusions to the specific crystalline interface or demonstrate that the qualitative features and the resulting anisotropy range are robust across representative interface configurations.","section":"Sec. II A, Fig. 1(b); Sec. IV"},{"comment":"The zeta ~ 1 result for oxygen vacancies rests on a single defect site and on an inferred mechanism. The paper proposes that oxygen vacancies induce a local spin-orbit coupling that randomizes electron spins during scattering, but no ab initio calculation of the local spin-orbit field, spin texture, or spin mixing around the vacancy is presented; the Elliott-Yafet-like scaling in Fig. 5(b) is obtained by scaling the full scattering matrix with A_scale, which cannot by itself identify the microscopic origin. A DFT-level characterization of the vacancy-induced local spin-orbit coupling, or tests with different vacancy sites and local environments, are needed to support the claim that SiO2 oxygen vacancies generically produce near-isotropic spin relaxation.","section":"Sec. III C and III D"},{"comment":"The oxygen-vacancy impurity density n_imp is not stated in the main text. Since the plotted spin lifetimes (tens of ns) and any comparison with experimental lifetimes depend on n_imp, this missing parameter prevents reproduction of the central zeta ~ 1 result. Please provide the value and its justification, or point to the specific section, equation, or table in the Supplemental Material where it is defined.","section":"Sec. III C, Fig. 4(a,b)"}],"minor_comments":[{"comment":"There is a typo: 'poseudopotentials' should be 'pseudopotentials', and the phrase 'fully-relativisitic' should be 'fully relativistic'.","section":"Sec. III A"},{"comment":"The phrase 'strongly constrasts' should read 'strongly contrasts'.","section":"Sec. III C"},{"comment":"The sentence 'The lifetime as a function of temperature-dependent (at mu = 25 meV above the CB minimum) is shown in the inset' is missing a noun; it should read 'The lifetime as a function of temperature (at mu = 25 meV above the CB minimum) is shown in the inset'.","section":"Fig. 3(a) caption"},{"comment":"The abstract states that the work quantifies electron-phonon scattering at the graphene/SiO2 interface, but for the bulk-SiO2 case electron-phonon scattering is not computed because of unstable slab phonons; this caveat should be acknowledged in the abstract or the wording should be adjusted.","section":"Abstract and Sec. III C"},{"comment":"The sign and normalization convention for the in-plane-field Rashba term lambda_in should be stated explicitly, since the text only notes that it couples to the z-component of the spin.","section":"Sec. II B, Eq. (2)"},{"comment":"Given the large number of fitted TB parameters, reporting fit uncertainties or at least specifying the fitted k-path and the weighting of band structure versus spin texture in the fits would improve reproducibility.","section":"Table I"}],"recommendation":"major_revision","confidential_remarks":"This is a strong numerical study that is within the scope of the journal, and I found no sign of circular fitting or unsupported citation claims. The main risk is not internal inconsistency but generalization: the anisotropic spin texture and the oxygen-vacancy local-SOC mechanism are demonstrated for one crystalline interface and one defect configuration. I would ask the authors to address the amorphous-interface question directly, even if only by a supplemental ensemble of high-symmetry configurations or by an explicit statement of the model's validity range. The missing oxygen-vacancy density is an easy fix but important for reproducibility."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Read it. This paper deserves a real referee. The new part is the anisotropic tight-binding model for graphene on bulk SiO2: bond-resolved corrections to hopping, Rashba, and an in-plane-field Rashba term that produces a nonuniform out-of-plane spin texture. That goes beyond the standard Rashba picture, and the modified D'yakonov-Perel' formula (Eq. 18) with the oscillating out-of-plane field is a clean analytical addition. The numerics look solid: two independent approaches (FPDMD and TB-KPM) agree, the DP model matches the simulated lifetimes, and the model parameters are fitted to DFT band structure and spin texture, not to the spin lifetimes or anisotropies they predict. So this is not a circular fit to the measurement. The oxygen-vacancy route to nearly isotropic relaxation is genuinely interesting and the scaling analysis in Fig. 5 separates it from intrinsic EY relaxation.\n\nThe main soft spot is the interface model. Every bulk-SiO2 result comes from a single crystalline slab geometry: AA-stacked, Si-terminated, OH-passivated. The anisotropic terms in Table I are at the few-meV level. Real SiO2 is amorphous. If random local bonding averages out the anisotropy, the zeta>1/2 behavior likely reverts to Rashba 1/2. The paper does not sample any amorphous or disordered interface, and the abstract's \"similar to measurements\" overstates what one configuration can establish. The authors are careful to say \"capable of yielding,\" and that scoping is defensible, but this limitation should be addressed head-on rather than left implicit.\n\nTwo smaller concerns. First, the EY-like relaxation from oxygen vacancies is attributed to local SOC induced by the defect, but that local SOC is never directly computed; the scaling evidence is suggestive but indirect. Second, the short-range puddle parameters are chosen to produce intervalley scattering, and the sensitivity of the anisotropy to those choices is not explored. Electron-phonon scattering is also absent for the bulk slab because of unstable phonons, so the phonon contribution in a real device is left open. Citation practice is fine; the self-citation to Ref. 13 is for an independently established mechanism.\n\nBottom line: this is a serious computational study with real novelty, and the central claim holds up on its own terms. I would send it to peer review and ask the authors to confront the amorphous-interface question—either by testing several plausible interface configurations or by softening the experimental comparison. If they do that, the paper becomes a solid contribution.","headline":"A serious computational study that shows substrate SOC alone can push spin lifetime anisotropy above the Rashba 1/2, but the quantitative link to real SiO2 rests on a single crystalline slab and needs an amorphous-interface test.","tokens_in":17303,"tokens_out":3787,"would_cite":true,"duration_ms":40749,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"The SiO2 substrate, through spin-orbit coupling alone, can produce the nearly isotropic spin lifetimes measured in graphene spin valves, removing the need to invoke magnetic impurities or contact effects.","keywords":["spin lifetime anisotropy","graphene","SiO2 substrate","spin-orbit coupling","spin texture","Rashba","Elliott-Yafet","electron-hole puddles"],"falsifier":"A decisive check would be a Hanle or oblique-precession measurement on a graphene spin valve on an ultraclean SiO2 surface, with magnetic impurities and contact effects independently ruled out, that returns $\\zeta = 1/2$ at all carrier densities; that would show the substrate alone cannot produce near-isotropic relaxation. The converse observation, $\\zeta$ clearly above 1/2 in a device with no magnetic impurities, would support the paper's mechanism. Computationally, an ab initio or tight-binding simulation of an amorphous SiO2 interface whose spin texture is purely helical would falsify the generality of the bulk-silica symmetry breaking.","tokens_in":15698,"feed_emoji":"🧲","tokens_out":6140,"duration_ms":62588,"temperature":0.7,"pith_summary":"This paper argues that a common SiO2 substrate, normally regarded as an inert support, can by itself control how long electron spins survive in graphene. The authors find two distinct regimes: a 2D SiO2 layer gives the standard Rashba helical spin texture and a spin-lifetime anisotropy of $\\zeta = 1/2$, while a bulk SiO2 surface breaks both out-of-plane and in-plane mirror symmetry, creating a spin texture with a preferred in-plane axis and a nonuniform out-of-plane component. With realistic disorder, this substrate-only mechanism yields anisotropy between 0.5 and 1, and oxygen vacancies in the SiO2 drive it to about 1. If the claim holds, the substrate alone explains the near-isotropic spin relaxation measured in graphene spin valves, and it will matter more as devices become cleaner.","feed_headline":"SiO2 can make graphene's spin lifetime nearly isotropic","feed_subtitle":"First-principles simulations lift the Rashba limit of ζ = 1/2 and match spin-valve experiments.","key_machinery":"The engine of the argument is the anisotropic tight-binding Hamiltonian $\\hat{H}_{\\mathrm{aniso}}$ of Eq. (2), whose sums run only over the graphene bonds that lie nearest to the SiO2 surface (the green bonds in Fig. 1(b)). Its three terms do three distinct jobs: $\\delta t$ shifts the band extrema away from the $K$ point in $k$-space, $\\delta\\lambda_R$ makes the helical in-plane spin texture anisotropic, and $\\lambda^{\\mathrm{in}}_R$, modeling a local in-plane electric field from broken mirror symmetry, produces the nonuniform out-of-plane spin component. Fitted to DFT band structures and spin textures, this Hamiltonian is used in two transport setups: first-principles Lindbladian density-matrix dynamics with electron-phonon and impurity scattering, and a linear-scaling real-space method that handles millions of atoms and models electron-hole puddles as Gaussian potentials.","core_discovery":"The central claim is that the SiO2 interface does more than induce the textbook Rashba spin-orbit coupling. On a bulk SiO2 surface, proximity to the substrate modifies graphene's nearest-neighbor bonds, so the authors extend the standard substrate Hamiltonian with three bond-selective terms: a hopping distortion $\\delta t$, a bond-dependent Rashba term $\\delta\\lambda_R$, and a local out-of-plane spin coupling $\\lambda^{\\mathrm{in}}_R$ that arises from broken in-plane mirror symmetry. These terms produce a spin texture whose in-plane part is no longer purely helical (it favors the $k_y$ axis) and whose out-of-plane component is finite and nonuniform around the $K$ point. Spin-transport simulations then show the anisotropy $\\zeta = \\tau_{s_z}/\\tau_{s_x}$ can leave the Rashba value of 1/2: electron-hole puddles that scatter between valleys give $0.7 < \\zeta < 0.95$, and oxygen vacancies give $\\zeta \\approx 1$ through a local SOC that randomizes spin during scattering in a way decoupled from momentum relaxation. The conclusion the paper draws is that spin-orbit-driven relaxation caused by the SiO2 substrate alone is capable of yielding spin lifetime anisotropies close to 1, matching experiments.","pith_inferences":["A direct test would be to simulate an amorphous SiO2/graphene interface: if the anisotropic spin texture and vacancy-induced local SOC survive disorder in the oxide, the mechanism is robust; if not, the quantitative agreement with experiments is specific to the crystalline model.","The same bond-selective anisotropic Hamiltonian could be applied to other dielectrics and van der Waals substrates (hBN, Al2O3, Si3N4), predicting which interfaces will show $\\zeta > 1/2$ without needing magnetic impurities.","The nonuniform out-of-plane spin texture is a small effect here but could be amplified in other interfaces; the added term $\\Omega_{z,\\mathrm{osc}}$ in the modified DP model is a general correction that should appear in any system with modulated out-of-plane spin polarization.","If vacancy-induced Elliott-Yafet-like relaxation is real, then controlling the stoichiometry of the oxide (e.g., via annealing) would provide a practical way to switch the dominant spin relaxation mechanism between DP and EY."],"forward_implications":["If the substrate alone can drive $\\zeta$ near 1, the interpretation of existing spin-valve experiments must be revisited: near-isotropic spin relaxation no longer implies magnetic impurities or contact effects.","The in-plane spin-lifetime anisotropy ($\\tau_{s_x} \\neq \\tau_{s_y}$) predicted for graphene on bulk SiO2 gives an experimental handle: measuring $\\zeta$ separately along $x$ and $y$ could confirm the broken in-plane symmetry.","Substrate engineering becomes a lever for spintronics: choosing or terminating the dielectric could tune $\\zeta$ between 1/2 and 1.","Oxygen vacancies are a concrete, controllable defect channel: varying vacancy density in the oxide should change the Elliott-Yafet-like contribution while leaving momentum relaxation largely unchanged.","As graphene devices get cleaner and contacts improve, substrate-induced SOC, not impurities, will set the upper bound on spin diffusion length."],"supporting_citations":[{"why":"Establishes that a combination of out-of-plane spin polarization and intervalley scattering boosts the anisotropy above 1/2, and supplies the modified DP model used here.","marker":"[13]"},{"why":"Reports the oblique spin precession measurements of nearly isotropic spin relaxation in graphene on SiO2 that the paper seeks to explain.","marker":"[22]"},{"why":"Provides the spin-precession-in-anisotropic-media formalism used to interpret the experimental anisotropy values.","marker":"[23]"},{"why":"Gives the competing magnetic-impurity explanation for isotropic spin relaxation that this paper argues is unnecessary.","marker":"[24]"},{"why":"Reports measured anisotropies of 0.76 to 0.96, the quantitative target reproduced by the substrate-only simulations.","marker":"[25]"},{"why":"Supplies the ab initio density-matrix dynamics (FPDMD) method used for first-principles electron-phonon and impurity spin transport.","marker":"[28]"},{"why":"Provides the linear-scaling real-space transport methodology used for the electron-hole puddle simulations.","marker":"[55]"},{"why":"Establishes that electron-hole puddles dictate spin dynamics in graphene and supplies the puddle parameters and DP interpretation.","marker":"[56]"}],"fun_headline_variants":["SiO2 lifts graphene spin anisotropy past Rashba 1/2","Graphene spin lifetime anisotropy tuned by SiO2 interface","Beyond Rashba: SiO2 revises graphene spin relaxation","SiO2 breaks graphene spin isotropy: anisotropy up to 1"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing assumption is that the two crystalline interface models used in the simulations, graphene on a 2D silica kagome layer and on an OH-passivated Si-terminated silica slab, capture the essential physics of a real amorphous SiO2 surface; if local bonding, termination, or trapped charges at an amorphous interface destroy the anisotropic spin texture or the vacancy-induced local SOC, the predicted $\\zeta$ near 1 would not transfer to experiments.","fun_headline_variants_meta":{"raw":{"variants":["SiO2 lifts graphene spin anisotropy past Rashba 1/2","Graphene spin lifetime anisotropy tuned by SiO2 interface","Beyond Rashba: SiO2 revises graphene spin relaxation","SiO2 breaks graphene spin isotropy: anisotropy up to 1"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000174,"raw_usage":{"total_tokens":1326,"prompt_tokens":1030,"completion_tokens":296,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":646,"completion_tokens_details":{"reasoning_tokens":225}},"tokens_in":646,"tokens_out":296,"duration_ms":3800,"temperature":1.0,"reasoning_tokens":225,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T00:34:42.165256+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A decisive check would be a Hanle or oblique-precession measurement on a graphene spin valve on an ultraclean SiO2 surface, with magnetic impurities and contact effects independently ruled out, that returns $\\zeta = 1/2$ at all carrier densities; that would show the substrate alone cannot produce near-isotropic relaxation. The converse observation, $\\zeta$ clearly above 1/2 in a device with no magnetic impurities, would support the paper's mechanism. Computationally, an ab initio or tight-binding simulation of an amorphous SiO2 interface whose spin texture is purely helical would falsify the generality of the bulk-silica symmetry breaking.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Reports the oblique spin precession measurements of nearly isotropic spin relaxation in graphene on SiO2 that the paper seeks to explain."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the spin-precession-in-anisotropic-media formalism used to interpret the experimental anisotropy values."},{"cited_title":"Kochan, M","cited_arxiv_id":null,"evidence_quote":"Gives the competing magnetic-impurity explanation for isotropic spin relaxation that this paper argues is unnecessary."},{"cited_title":"Ringer, S","cited_arxiv_id":null,"evidence_quote":"Reports measured anisotropies of 0.76 to 0.96, the quantitative target reproduced by the substrate-only simulations."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the ab initio density-matrix dynamics (FPDMD) method used for first-principles electron-phonon and impurity spin transport."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the linear-scaling real-space transport methodology used for the electron-hole puddle simulations."},{"cited_title":"Van Tuan, F","cited_arxiv_id":null,"evidence_quote":"Establishes that electron-hole puddles dictate spin dynamics in graphene and supplies the puddle parameters and DP interpretation."}],"review_version":1}