{"id":"7b6b2273-d5be-4df9-8192-314d82108522","arxiv_id":"1908.11578","paper_version":3,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"NEGF with Böttiker probe scattering, calibrated on bulk silicon, reproduces molecular dynamics interface thermal resistance for Si/heavy-Si across mass ratios 1 to 10.","lead":"This paper tests a quantum-transport method, the non-equilibrium Green's function approach, for predicting heat flow across the interface between silicon and a heavier form of silicon. The method, once calibrated on plain silicon, matches molecular dynamics simulations for mass ratios from 1 to 10 and shows which phonon modes carry the heat.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Calibration transfer from Si to heavy-Si is the load-bearing assumption; the reported agreement does not independently test the mass dependence of the Böttiker scattering rate.","rationale":"The reader's weakest assumption is the transferability of B and C from Si to heavy-Si, and the manuscript itself contains the relevant admission around Eq. (3). This is the same load-bearing point I identify. The paper deserves credit for stating the limitation and for the consistency of using the same Tersoff potential in MD and NEGF; the agreement in Fig. 2(b) is nontrivial. However, the claim that the parameters 'proved to be transferable' is stronger than what the data show: the bulk fit is in Si only, and the interface benchmark is the one place where the ignored mass dependence and near-interface anharmonicity would enter. The direct missing check is a bulk heavy-Si thermal conductivity comparison; without it, the MR=1–10 agreement could be a calibrated result rather than an independent prediction. This argues for keeping the reader's CONDITIONAL verdict rather than accepting the paper's stronger reliability claim. I do not see a more load-bearing concern: the finite-size extrapolation, spectral analysis, and the absence of artificial resistance in homogeneous systems are either secondary or, in the latter case, essentially a property of the Böttiker-probe construction. If the proposed heavy-Si bulk test passes, the remaining caveat is limited to interface-specific anharmonicity, which the authors already flag as future work.","tokens_in":8244,"tokens_out":8040,"duration_ms":81633,"concrete_test":"Run the NEGF calculation for a homogeneous heavy-Si system at MR=4 and MR=10 using the same B=5e-20 s/K and C=430 K, extract the bulk thermal conductivity with the same procedure as Fig. 2(a), and compare with MD heavy-Si bulk thermal conductivity over 280–320 K. If the NEGF value deviates by more than the MD statistical error (the paper quotes 1.6–4.6% for interface resistances), the mass-independent Böttiker parameters in Eq. (3) are falsified and the interface agreement should be re-examined for cancellation; if it matches, the bulk part of the transferability assumption is supported, leaving only the near-interface anharmonicity caveat.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that NEGF with Böttiker probes quantitatively reproduces MD thermal boundary resistance for MR=1–10 stands or falls on whether the two fitted parameters B and C are transferable from homogeneous Si to heavy-Si and to the interface region. The paper is explicit about this: after Eq. (3), 'The same values of B and C are used in Si and heavy-Si. In this way, the dependence of the scattering rate on the atomic mass and deviations of the anharmonicity near the interface from the one of the volume materials are ignored.' The only calibration constraint is the MD thermal conductivity of bulk Si (Fig. 2a); no homogeneous heavy-Si bulk thermal conductivity is shown. At a mass-mismatched interface, a significant part of the resistance is set by high-frequency Si phonons that cannot propagate elastically into heavy-Si and must down-convert via the inelastic Böttiker self-energy. If the true heavy-Si scattering rate differs from the frequency scaling in Eq. (3), the thermal boundary resistance could in principle move outside the MD error bars. The agreement across MR=1–10 suggests the assumption may be adequate, but because the fit was performed in Si and applied in the interface system, the agreement is evidence of transferability only if the mass and interface dependence of the scattering rate is independently confirmed.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This manuscript presents a methodology for including anharmonic phonon scattering in non-equilibrium Green's function (NEGF) calculations of thermal transport, using Büttiker probe self-energies with a phenomenological scattering rate of the form B T ω² exp(−C/T). The two parameters B and C are fitted to reproduce the MD-computed bulk Si thermal conductivity. The method is then applied to Si/heavy-Si interfaces with mass ratios from 1 to 10. The predicted thermal boundary resistance (TBR) agrees quantitatively with MD simulations, and unlike the coherent Landauer approach, the NEGF with Büttiker probes yields zero TBR in homogeneous systems. Spectral analysis of energy-resolved currents and local phonon densities of states illustrates how inelastic scattering enables heat flow by high-frequency phonons above the heavy-Si cutoff and how modal overlap between regions controls the spectral current distribution.","tokens_in":8483,"tokens_out":9440,"duration_ms":87219,"significance":"The paper addresses a real bottleneck in phonon NEGF—the treatment of anharmonicity—and offers a computationally light, spectral alternative to MD. The benchmark against MD over a mass-ratio range of 1–10 is a strong and appropriate test, and the explicit demonstration that the Büttiker-probe NEGF removes the artificial interface resistance plaguing the Landauer approach is a useful contribution. The manuscript is honest about its key approximation (transferability of B and C), and the spectral results are informative. However, the central claim of predictive transferability rests on an assumption that is not directly validated, which limits the strength of the conclusions as currently stated.","major_comments":[{"comment":"The central claim that the Büttiker probe parameters fitted to bulk Si thermal conductivity transfer to Si/heavy-Si interfaces is not directly tested for the mass and interface dependence of the scattering rate. Eq. (3) uses a single functional form for the scattering lifetime, and the same constants B and C are applied in Si and heavy-Si, as stated in the paragraph following Eq. (3). The only calibration is the bulk Si conductivity (Fig. 2(a)), and no homogeneous heavy-Si conductivity is computed or compared. The thermal boundary resistance at high mass ratio is largely controlled by high-frequency Si phonons that must down-convert in heavy-Si through the inelastic Büttiker self-energy; therefore, the agreement in Fig. 2(b) could in principle be coincidental if the true mass dependence of the anharmonic scattering rate differs from the assumed ω² exp(−C/T) form. I request a direct validation: compute NEGF and MD thermal conductivities of homogeneous heavy-Si (for at least MR = 2, 4, 8) with the same B and C, and compare them. If these agree, the transferability claim is substantiated; if they do not, the agreement in Fig. 2(b) should be discussed in that light and the conclusion (which states the parameters 'proved to be transferable') should be softened.","section":"Methods, Eq. (3) and following paragraph; Results, Fig. 2(b)"},{"comment":"The claim of 'quantitative agreement' is not fully supported by the presented data. The MD standard deviations (1.6–4.6%) are reported in the text but are deliberately not plotted, and no numerical values for the relative difference (dotted line) are given. Without error bars or a table of the computed TBR values with uncertainties, the reader cannot assess whether the deviations between NEGF and MD are within statistical uncertainty. Please include the MD error bars in Fig. 2(b) (or provide a table of values with uncertainties for both methods and the relative differences) and state the maximum relative difference observed across MR = 1–10.","section":"Results, Fig. 2(b) and surrounding text"}],"minor_comments":[{"comment":"The text states that 'each discretized atom in the system has a Büttiker probe applied to it' but later says the RGF method 'requires the Büttiker probe self-energies to be equal throughout each slab.' Please clarify whether the probes are per atom or per slab, and if per atom, how the slab-wise constant approximation is justified.","section":"Methods, paragraph on Büttiker probes and RGF"},{"comment":"The claim that the present NEGF approach is 'numerically more efficient than MD' is not supported by any timing, scaling, or resource comparison in the manuscript. Please either include a quantitative efficiency comparison or qualify the claim.","section":"Conclusion and Abstract"},{"comment":"The notation '∑4i=1 Ri = 1' should be typeset properly as '∑_{i=1}^{4} R_i = 1' for clarity.","section":"Fig. 5 caption"},{"comment":"The paper does not specify how the local temperature profile is extracted in the NEGF calculations (as opposed to the MD simulations). Since Eq. (1) is phrased in terms of a local phonon number density, it would be helpful to state explicitly how the NEGF Green's functions are used to obtain the temperature profile for the TBR extraction.","section":"Methods, local temperature extraction"}],"recommendation":"major_revision","confidential_remarks":"The transferability concern is the crux of this manuscript. I believe the paper is potentially very good, but the authors should either add the heavy-Si bulk conductivity validation or substantially soften the 'proved to be transferable' claim. The error-bar presentation should also be improved. This is a well-written manuscript from a strong group; the requested additions are straightforward and within scope."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The one thing to know: this is a solid, transparent validation study, not a new method. The authors take the Büttiker-probe phonon NEGF approach from their earlier work and benchmark it systematically against MD for Si/heavy-Si interfaces over mass ratios 1–10. That systematic comparison is the real contribution, and the central quantitative claim holds up for the system studied.\n\nWhat is done well: the calibration is clean, with B and C fitted only to bulk Si thermal conductivity, then applied to the interface problem. The agreement with MD over a factor of ten in mass ratio is genuine evidence that the approach captures the relevant physics, at least for this mass-mismatch geometry. The demonstration that the artificial Landauer interface resistance disappears in the NEGF treatment is a nice, concrete point. The spectral analysis of which phonon modes carry current across the interface is also useful and follows naturally from the NEGF framework.\n\nThe soft spots are real but not fatal. The load-bearing assumption is that the fitted scattering parameters transfer from bulk Si to heavy-Si and to the interface region, explicitly ignoring mass dependence of the scattering rate and any change in anharmonicity near the interface. The authors state this clearly. The paper does not, however, show a check on the heavy-Si bulk thermal conductivity, which would be the natural way to test that transferability. Without that, some of the agreement across mass ratios could be partly coincidental. A sensitivity analysis on B and C, or NEGF error bars, would also strengthen the claim. The MD error bars are small, but the NEGF results are presented without any indication of their uncertainty. Finally, the conclusion that NEGF can “reliably predict phonon transport across interfaces” reaches further than one model system supports; the paper would be better served by a more cautious statement.\n\nThese are not disqualifying issues. The paper is honest, the method comparison is internally consistent, and the finite-size extrapolation is handled carefully. The one-interface limitation is a scope limitation, not a flaw in the results.\n\nWho this is for: people working on phonon NEGF, thermal boundary resistance simulation, or method benchmarking. It deserves a serious referee and would likely benefit from minor revision rather than rejection. I would recommend sending it to peer review, requesting the heavy-Si bulk conductivity check and a modestly toned-down general conclusion.","headline":"A clean, honest benchmark showing NEGF with Böttiker probes reproduces MD thermal boundary resistance across mass ratios 1–10, with the main caveat being the untested transferability of the fitted scattering parameters from bulk Si to heavy-Si.","tokens_in":9054,"tokens_out":1967,"would_cite":true,"duration_ms":19895,"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":"With Büttiker probes, non-equilibrium Green's function transport reproduces molecular-dynamics thermal boundary resistance for mass ratios 1 to 10.","keywords":["thermal boundary resistance","non-equilibrium Green's function","Büttiker probe","phonon transport","anharmonic scattering","molecular dynamics","Si/heavy-Si interface","spectral phonon current"],"falsifier":"Run the same NEGF setup with the same fitted scattering parameters for a mass ratio inside 1 to 10 that is not shown in the paper, such as MR=6, and compare with molecular dynamics; if the boundary resistance deviates substantially from the MD value, the claimed quantitative agreement across the mass-ratio range fails.","tokens_in":8006,"feed_emoji":"🔥","tokens_out":11392,"duration_ms":98559,"temperature":0.7,"pith_summary":"This paper argues that non-equilibrium Green's function (NEGF) phonon transport can include anharmonicity by attaching a Büttiker probe to each atom, and that the resulting method predicts thermal boundary resistance reliably. It tests this on a Si/heavy-Si interface, where the two materials differ only in atomic mass. After tuning two scattering parameters against a molecular dynamics (MD) bulk thermal conductivity calculation, the NEGF thermal boundary resistance agrees quantitatively with MD across mass ratios from 1 to 10. The method also removes the artificial interface resistance that a ballistic, scattering-free Landauer approach produces at a virtual interface in a homogeneous material. The practical payoff is a spectral, mode-resolved way to compute boundary resistance at finite temperature without the cost of full self-consistent Born scattering.","feed_headline":"Phonon Green's functions with Büttiker probes match molecular dynamics","feed_subtitle":"NEGF with fitted scattering predicts Si/heavy-Si boundary resistance for mass ratios 1 to 10","key_machinery":"The load-bearing object is the Büttiker probe self-energy, a complex, frequency-dependent insertion placed on every atom and vibrational direction: $\\Sigma^R_{BP(i,m)}(\\omega)=-i\\,2\\omega\\hbar^2/\\tau_{i,m}(\\omega)$, with the Umklapp lifetime $\\tau_i^{-1}(\\omega)=BT_i\\omega^2 e^{-C/T_i}$. The probe temperatures $T_i$ are unknowns solved iteratively so that each probe conserves energy, which turns the coherent, elastic NEGF calculation into a dissipative one within a recursive Green's function scheme. This object is what removes the spurious interface resistance in homogeneous systems and what lets the energy-resolved current be read directly from the steady-state Green's functions.","core_discovery":"The central claim is that NEGF with Büttiker probe self-energies, calibrated once on homogeneous bulk Si, gives correct thermal boundary resistance for the Si/heavy-Si interface across mass ratios 1 to 10, and yields zero interface resistance in a perfectly homogeneous system. The probes add a phenomenological scattering rate to each atom that enforces energy conservation through self-consistently solved probe temperatures; the calibration parameters are chosen so the NEGF bulk Si thermal conductivity matches MD. With this calibration, the boundary resistance rises exponentially with mass ratio and agrees with MD, while the artificial resistance of the ballistic Landauer approach disappears. The natural spectral output shows that inelastic scattering allows phonons above the heavy-Si cutoff to propagate, and that mode-resolved currents are controlled by how evenly a phonon mode is distributed across both sides of the interface.","pith_inferences":["This suggests the same fit-to-bulk calibration strategy could work for other homogeneous host materials, though the explicit neglect of mass-dependent scattering means recalibration is likely needed for very large mass contrasts.","Because NEGF is spectral and shares a common framework with electronic transport, coupling these phonon Büttiker probes to electron NEGF is a natural route to self-heating and thermoelectric simulations, a direction the paper motivates but does not demonstrate.","A testable extension is to vary the bath temperatures; the paper only benchmarks at 320 K and 280 K, so temperature transferability of the fitted parameters remains open.","The disappearance of artificial resistance also suggests Büttiker-probe NEGF could serve as a calibration reference for other approximate phonon transport methods that suffer from spurious contact resistances."],"forward_implications":["NEGF can now predict thermal boundary resistance at finite temperature with anharmonicity included at a fraction of the cost of self-consistent Born scattering.","Homogeneous systems produce exactly zero interface resistance, so spurious boundary terms that plague Landauer-type calculations are eliminated.","Energy-resolved, mode-resolved spectral currents come out directly, enabling interface resistance to be traced to specific phonon modes without transformation.","The approach is almost free of finite-size effects in the transport direction because it uses open boundary conditions with semi-infinite leads.","The fitted scattering parameters transfer from bulk calibration to the interface over mass ratios 1 to 10, supporting the use of Büttiker probes as a practical anharmonicity model."],"supporting_citations":[{"why":"Supplies the Büttiker probe concept that the scattering self-energies are built on.","marker":"27"},{"why":"Provides the phonon Büttiker probe self-energy and the Umklapp lifetime formula used in Eq. (3).","marker":"28"},{"why":"Supplies the recursive Green's function transport method and the local-temperature fitting used to extract boundary resistance.","marker":"21"},{"why":"Defines the interatomic potential whose harmonic force constants are shared by both the MD and NEGF Hamiltonians.","marker":"20"},{"why":"Establishes the 1/L extrapolation procedure used to extract the interface resistance.","marker":"22"},{"why":"The non-equilibrium Landauer approach whose artificial interface resistance the present NEGF method is designed to remove.","marker":"19"},{"why":"Documents the artificial interface resistance that arises in the Landauer approach in homogeneous systems.","marker":"6"},{"why":"Runs the molecular dynamics simulations that supply the benchmark thermal conductivity and boundary resistance.","marker":"23"}],"fun_headline_variants":["NEGF probes match MD for Si interface resistance","Büttiker probes fix Green's function thermal barriers","One calibration, many mass ratios: NEGF matches MD","Zero fake resistance: NEGF with probes hits MD","Inelastic probes predict interface resistance"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The whole calculation rests on the assumption that the scattering strength fitted in ordinary silicon also applies to the heavier silicon and to the atoms right at the interface, even though heavier atoms vibrate more slowly and the interface may not be as anharmonic as the bulk.","fun_headline_variants_meta":{"raw":{"variants":["NEGF probes match MD for Si interface resistance","Büttiker probes fix Green's function thermal barriers","One calibration, many mass ratios: NEGF matches MD","Zero fake resistance: NEGF with probes hits MD","Inelastic probes predict interface resistance"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000808,"raw_usage":{"total_tokens":3519,"prompt_tokens":893,"completion_tokens":2626,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":509,"completion_tokens_details":{"reasoning_tokens":2554}},"tokens_in":509,"tokens_out":2626,"duration_ms":21250,"temperature":1.0,"reasoning_tokens":2554,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T10:10:45.595029+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run the same NEGF setup with the same fitted scattering parameters for a mass ratio inside 1 to 10 that is not shown in the paper, such as MR=6, and compare with molecular dynamics; if the boundary resistance deviates substantially from the MD value, the claimed quantitative agreement across the mass-ratio range fails.","supporting_citations":[{"cited_title":"Sellan , author E","cited_arxiv_id":null,"evidence_quote":"Supplies the Büttiker probe concept that the scattering self-energies are built on."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the recursive Green's function transport method and the local-temperature fitting used to extract boundary resistance."},{"cited_title":"Luisier ,\\ title title Atomistic modeling of anharmonic phonon-phonon scattering in nanowires , \\ 10.1103/PhysRevB.86.245407 journal journal Phys","cited_arxiv_id":null,"evidence_quote":"Defines the interatomic potential whose harmonic force constants are shared by both the MD and NEGF Hamiltonians."},{"cited_title":"Stieger , author A","cited_arxiv_id":null,"evidence_quote":"The non-equilibrium Landauer approach whose artificial interface resistance the present NEGF method is designed to remove."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Runs the molecular dynamics simulations that supply the benchmark thermal conductivity and boundary resistance."}],"review_version":1}