{"id":"fc196db2-b800-4894-8bf7-4d99a9f2a3e9","arxiv_id":"2505.10225","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"A sodium-intercalated √7×√7 Moiré stack of hexagonal boron nitride is predicted to support plasmons that avoid first-order electron-phonon decay in a narrow 1.08 to 1.16 eV window, with electron-electron scattering setting the actual lifetime at about 10^14 s^-1.","lead":"Sodium atoms placed between two layers of hexagonal boron nitride create a nearly flat electronic band, and the authors predict that plasmons in this band do not lose energy to crystal vibrations near 1 eV. The paper then shows these plasmons decay mainly by splitting into lower-frequency plasmons at a rate near 10^14 per second, which sets a practical limit on how low the losses really are.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The dominant ~10^14 s^-1 electron-electron decay rate rests entirely on the untested plasmon-pole approximation (Eq. 5); if that approximation is quantitatively wrong, the reported lifetime and the practical low-loss claim change.","rationale":"The reader's weakest-assumption analysis identifies exactly the load-bearing point: the dominant electron-electron loss rate is computed within the plasmon-pole approximation, a simplification the authors themselves flag as untested. This is the correct central concern because the first-order electron-phonon immunity is a kinematic consequence of the band structure and phonon spectrum, whereas the claimed 10^14 s^-1 rate controls the actual lifetime and quality factor. A concrete RPA-level recomputation would settle whether the number is credible. This stress-test does not change the reader's verdict; it reinforces the condition that the paper should be read as a predictive proposal pending validation of the electron-electron decay channel.","tokens_in":8154,"tokens_out":4320,"duration_ms":44496,"concrete_test":"Using the same Wannier Hamiltonian, recompute Im Sigma(k, omega') for the isolated band with the full dynamically screened RPA response W(q, omega), or a Mermin approximation with the same band structure, and then reevaluate tau^-1(omega) from Eq. (2) for omega in the 1.08-1.16 eV window. Compare the resulting decay rate with the plasmon-pole value, and test sensitivity to the q-grid density and to the hBN dielectric screening. If the rate changes by more than roughly a factor of 2 away from 10^14 s^-1, the central lifetime claim is not supported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central quantitative prediction most exposed is the ~10^14 s^-1 electron-electron decay rate reported in the section 'Plasmon to Plasmon Scattering.' It is computed by inserting the plasmon-pole self-energy of Eq. (5) into the general lifetime formula, Eq. (2), and the authors explicitly state in the Outlook that this prediction rests on the validity of the plasmon-pole approximation. This is the dominant intrinsic loss in the 1.08-1.16 eV window; if the approximation misestimates the phase space or coupling strength, the lifetime and the 'low-loss' characterization could change by an order of magnitude or more. The paper offers no independent validation of this rate, such as a full RPA or sum-rule comparison. By contrast, the first-order electron-phonon 'lossless' window is a kinematic statement based on the computed bandwidth and maximum phonon energy, and is more robust; the practical value of the material, however, is governed by the unvalidated electron-electron channel.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This manuscript proposes a class of Moiré heterostructures of hexagonal boron nitride intercalated with alkali atoms as platforms for low-loss plasmons. The authors screen five alkali species and two Moiré angles, and identify sodium intercalated in a √7×√7 hBN Moiré structure as the only candidate with an isolated, moderately flat band at the Fermi level. Using DFT for band structures and phonons and a three-layer model for the plasmon dispersion, they find that for hole dopings n=0.1 and n=0.2, plasmons in the energy window 1.08–1.16 eV are immune to first-order electron-phonon decay. They then compute higher-order electron-phonon decay rates via Eq. (2) and find rates around 10^7 s^-1 in that window. Finally, using the plasmon-pole approximation for the electron-electron self-energy, Eq. (5), they report that plasmon-to-plasmon scattering dominates, with a rate around 10^14 s^-1. The paper concludes that the electron-phonon channel is effectively suppressed but the electron-electron channel sets the practical lifetime.","tokens_in":8204,"tokens_out":3812,"duration_ms":39295,"significance":"If the quantitative predictions hold, the paper provides a concrete, structurally stable, nonmagnetic material in which a specific plasmon frequency window is immune to first-order electron-phonon decay, addressing a limitation of previous defect-lattice proposals in hBN. The kinematic argument for the lossless window is transparent and rests on computed band and phonon properties rather than on parameters fitted to the target decay rates. The comparison of Eqs. (2), (3), and (4) offers a useful internal consistency check for the electron-phonon channel, and the prediction of a 1.08–1.16 eV lossless window is falsifiable. The main weakness is that the dominant loss channel, the electron-electron plasmon-plasmon scattering rate of about 10^14 s^-1, is obtained solely from the plasmon-pole approximation and is not validated by an independent calculation or a quantitative error estimate. Because this channel governs the practical lifetime, the central low-loss claim is not yet fully established.","major_comments":[{"comment":"The dominant decay rate of about 10^14 s^-1 is computed by inserting the plasmon-pole electron-electron self-energy, Eq. (5), into Eq. (2). This is the only calculation of the electron-electron channel in the paper, and the Outlook explicitly states that the prediction rests on the validity of the plasmon-pole approximation. Since this channel dominates the lifetime in the 1.08–1.16 eV window, the central practical claim of controlled losses is not yet quantitatively established. I recommend adding an independent validation, such as a full RPA calculation of the loss function, a comparison with the f-sum rule, or a quantitative assessment of the plasmon-pole approximation error for the relevant doping and frequency range. A qualitative caveat in the Outlook is not sufficient for a reported rate of 10^14 s^-1.","section":"Plasmon to Plasmon Scattering, Eq. (5), Fig. 4"},{"comment":"The manuscript uses only the imaginary part of the electron self-energy and states in footnote [23] that the real part contributes to a change in the effective bandwidth. The first-order lossless window, however, is defined by the computed bandwidth (0.9 eV), the maximum phonon energy, and the band-edge position relative to the Fermi level. A real-part correction of a few tens of meV could shift the 1.08 eV lower edge or the 1.16 eV upper edge for n=0.1 and materially change the predicted window. The authors should quantify the real-part contribution or argue with specific numbers that it is negligible.","section":"Plasmonic Losses, Eq. (2), Footnote [23]"},{"comment":"For n=0.1, the first-order lossless window spans only 1.08 to 1.16 eV, a margin of about 80 meV. The paper reports no error bars or convergence tests for the DFT bandwidth, the maximum phonon frequency, or the band-edge energies that set both edges of this window. Because the window is the manuscript's core prediction, its robustness to computational parameters (k-point sampling, Wannier interpolation, exchange-correlation functional, and van der Waals correction) should be demonstrated. A short convergence or sensitivity study would substantially strengthen the claim.","section":"Introduction, Fig. 2, lossless window"}],"minor_comments":[{"comment":"There are typographical errors, including 'subsitutional' instead of 'substitutional' and inconsistent spelling of 'Moiré' as 'Moire'.","section":"Introduction"},{"comment":"The typeset form of Eq. (1) in the manuscript is difficult to parse; please ensure the three-layer plasmon dispersion equation is rendered clearly and that all symbols are defined at first use.","section":"Eq. (1)"},{"comment":"The caption for Fig. 2(c) should state explicitly what the pink shaded region represents and what the solid arrows denote, since the current description is ambiguous.","section":"Fig. 2(c)"},{"comment":"Reference [33] is incomplete; it should include the full author list and a proper citation for the JJDFTX.jl repository.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The stress-test concern about the plasmon-pole approximation lands: the electron-electron decay rate is load-bearing and currently rests entirely on Eq. (5). The reliance on the authors' earlier work [9] is legitimate and not circular, since no parameter is fitted to the target decay rates. The manuscript would be suitable for the journal after the electron-electron channel is validated or its uncertainty quantified, and after the sensitivity of the lossless window to computational inputs is addressed."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nShort version: the paper predicts that sodium intercalated into a √7×√7 Moiré hBN stack creates an isolated, moderately flat band that supports plasmons lossless at first order in the electron-phonon interaction in a 1.08–1.16 eV window. Higher-order e-ph decay is negligible (~10^7 s^-1) in that range, but electron-electron scattering into lower-frequency plasmons dominates at ~10^14 s^-1. For me the key takeaway: the first-order lossless claim is solid; the practical lifetime is not.\n\nWhat's actually new: the authors carry the flat-band lossless-plasmon idea from their previous defect-lattice work [9] into a more experimentally feasible material class, and they generalize the treatment of e-e plasmon decay beyond the equal-frequency case used for graphene. That is a genuine step. The workflow (DFT, phonons, Wannier-based plasmons, three-layer model) is coherent, and the lossless window follows from computed bandwidth and phonon energies rather than from fitting the target rates. The cross-checks in Eqs. (2)–(4) and the graphene comparison give reasonable confidence in the machinery.\n\nSoft spots, in proportion: the dominant ~10^14 s^-1 e-e decay rate rests entirely on the plasmon-pole self-energy, Eq. (5), which the authors explicitly flag in the Outlook. If that approximation misestimates phase space or coupling, the lifetime and any practical low-loss claim shift by an order of magnitude. That is the load-bearing uncertainty. There are also no error bars or convergence data anywhere, which matters for a paper making quantitative predictions. And the 'only sodium' conclusion is based on five alkali atoms and two Moiré angles—fine for a candidate, not for a general rule. The title's 'low-loss' framing is also doing work: at ~1 eV and 10^14 Hz the Q is around 20, so this is a design principle and a concrete candidate, not a demonstrated low-loss material.\n\nIf you work on 2D plasmonics or flat-band physics, this is worth reading and citing as a predicted system. It deserves a serious referee: the central first-order lossless window is robust, and the e-e rate, while approximate, is honestly labeled as such. I'd accept it with a request for uncertainty estimates and, ideally, a full RPA check of the e-e decay.\n\nRegards.","headline":"Predicts a concrete candidate for first-order lossless plasmons in sodium-intercalated Moiré hBN, but the e-e decay that sets the real lifetime rests on an unvalidated approximation.","tokens_in":8904,"tokens_out":3586,"would_cite":true,"duration_ms":32880,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["73.20.Mf","71.15.Mb","63.22.-m"],"model":"deepseek-v4-flash","headline":"Sodium intercalation makes a Moiré hBN heterostructure host plasmons that evade first-order electron-phonon decay, with lifetimes set by electron-electron scattering near 10^14 s^-1.","keywords":["Moiré heterostructures","hexagonal boron nitride","alkali intercalation","plasmons","low-loss plasmonics","electron-phonon interaction","flat bands","electron-electron scattering"],"falsifier":"Compute the electron-electron self-energy without the plasmon-pole approximation (for example, with a full RPA dielectric function or a quantum Monte Carlo approach) and compare the resulting plasmon decay rate in the 1.08–1.16 eV window to the paper's ~$10^{14}$ $s^{-1}$ prediction; a rate differing by an order of magnitude would invalidate the central lifetime claim. Alternatively, if a fabricated sodium-intercalated √7×√7 hBN sample shows first-order phonon-assisted plasmon decay below 1.16 eV in optical or near-field spectroscopic measurements, the band-structure picture would be wrong.","tokens_in":7847,"feed_emoji":"⚡","tokens_out":3892,"duration_ms":38010,"temperature":0.7,"pith_summary":"This paper proposes a concrete material platform—sodium atoms intercalated into a √7×√7 Moiré heterostructure of hexagonal boron nitride—that should support plasmons with deep subwavelength confinement that are immune to the usual decay channel of one-phonon emission in a frequency window near 1 eV. If the calculations hold, this would be a practical route to low-loss plasmonics in a structurally stable, nonmagnetic, fabricable system, moving past earlier proposals that were ferromagnetic, unstable, or difficult to make. The key design principle is an isolated, moderately flat band at the Fermi level: with bandwidth 0.9 eV and a maximum phonon energy of 0.18 eV, plasmons above 1.08 eV cannot lose energy by emitting one phonon within that band. The paper shows that higher-order electron-phonon decay is negligible in that window, but that electron-electron interactions cause plasmons to decay to lower-frequency plasmons at a rate around $10^{14}$ $s^{-1}$, which sets the practical lifetime.","feed_headline":"Sodium opens a lossless plasmon window at 1.1 eV","feed_subtitle":"In a √7×√7 hBN Moiré heterostructure, plasmons in a 1.08–1.16 eV band evade first-order phonon decay.","key_machinery":"The central object is the isolated, moderately flat electronic band at the Fermi level created by sodium intercalation into the √7×√7 Moiré pattern of hBN. The argument rests on two threshold conditions: a plasmon with energy ℏω is protected from intraband one-phonon decay when ℏω > W + ℏω_ph (with bandwidth W = 0.9 eV and maximum phonon energy ℏω_ph ≈ 0.18 eV), and from interband one-phonon decay when ℏω < |E_{v,c} − E_Fermi| − ℏω_ph. The quantitative machinery includes a three-layer dielectric model for the plasmon dispersion (Eq. 1), the general self-energy-based decay-rate formula of Allen (Eq. 2) that accounts for finite bandwidth and higher-order electron-phonon processes, and the plasmon-pole approximation for the electron-electron self-energy (Eq. 5) that yields the dominant plasmon-plasmon scattering rate. The same framework is validated against the one-phonon formula (Eq. 4) and against graphene, where all three methods agree.","core_discovery":"The central claim is that in the √7×√7 sodium-intercalated hBN Moiré heterostructure, the Moiré potential traps the sodium atoms and produces an isolated band of width 0.9 eV at the Fermi level, with the nearest other bands far enough away that plasmons in the 1.08–1.16 eV window cannot decay through any one-phonon-assisted intraband or interband transition. For hole dopings of n = 0.1 and n = 0.2, the paper finds a first-order lossless window from 1.08 eV up to 1.16 eV, above which interband one-phonon processes turn on. Using self-energy-based decay formulas, the authors compute higher-order electron-phonon decay rates near $10^{7}$ $s^{-1}$ in that window, and then, using a plasmon-pole approximation for the electron-electron self-energy, compute a dominant plasmon-to-plasmon scattering rate of about $10^{14}$ $s^{-1}$. The conclusion is that, in a clean sample, the plasmon lifetime in this material is controlled not by phonons but by intrinsic electron-electron scattering, and that the material is dynamically stable with no imaginary phonon frequencies.","pith_inferences":["The paper's own caveat is that the dominant decay rate rests on the plasmon-pole approximation; if a full random-phase-approximation or diagrammatic Monte Carlo treatment shifts that rate significantly, the practical lifetime and the claim of 'low-loss' would change, though the first-order phonon immunity would remain.","The Moiré-potential trapping mechanism suggests a broader design space: varying the intercalant size, Moiré angle, and substrate layers could tune the bandwidth and the gap to neighbouring bands, potentially opening lossless windows at other frequencies.","The 1.08–1.16 eV window sits near the telecom and near-infrared region, so a successful experimental realization could enable low-loss nanophotonic devices, but the fabrication of large-angle Moiré hBN with controlled sodium doping remains an untested experimental challenge.","The authors mention impurity and two-plasmon scattering as additional channels; the predicted 10^14 s^-1 rate therefore represents an upper bound on lifetime in ideal clean samples, and real devices would likely show additional losses."],"forward_implications":["Plasmons in the 1.08–1.16 eV window of the sodium-intercalated heterostructure would propagate with decay rates set by electron-electron scattering (~10^14 s^-1), yielding quality factors orders of magnitude higher than conventional two-dimensional metals at comparable confinement.","The design rule 'isolated flat band at the Fermi level plus small maximum phonon energy' becomes a systematic search principle: among the five alkalis tested, only sodium produces a sufficiently isolated band, and the √7×√7 angle is the one that avoids ferromagnetism.","The decay-rate evaluation protocol (combining finite-bandwidth self-energy formulas with one-phonon checks) provides a template for predicting plasmon lifetimes in other bandwidth-limited two-dimensional metals.","The lossless window is bounded: plasmons above 1.16 eV acquire interband one-phonon decay channels, so applications must deliberately operate inside the 1.08–1.16 eV range.","If realized, the material would offer near-infrared subwavelength plasmonics in a stable, nonmagnetic van der Waals heterostructure, a regime previously inaccessible in hBN-based systems."],"supporting_citations":[{"why":"Prior work by the same group establishing the flat-band design principle and the one-phonon loss formula for defect lattices in hBN, which the current paper extends to Moiré heterostructures.","marker":"[9]"},{"why":"Allen's general self-energy expression for infrared conductivity, which is the basis of Eq. (2) for the plasmon decay rate including finite bandwidth and higher-order electron-phonon processes.","marker":"[21]"},{"why":"Jablan and Chang's treatment of multiplasmon absorption in graphene, which the authors generalize to derive the plasmon-plasmon scattering contribution to the decay rate.","marker":"[25]"},{"why":"Prior work on Li-intercalated hBN films showing a strong two-dimensional plasmon with low damping, providing the experimental and theoretical context the authors differentiate from their large-angle Moiré approach.","marker":"[17]"},{"why":"The DFT code used for structural relaxation and band-structure calculations, on which the identification of the sodium-induced isolated band and the phonon stability test depend.","marker":"[26]"}],"fun_headline_variants":["Sodium-stuffed hBN Moire silences phonon decay for 1 eV plasmons","Moire-trapped sodium gives lossless plasmons at 1 eV","Sodium intercalation kills phonon losses in Moire plasmons","hBN Moire with sodium: phonon-free plasmon window at 1.1 eV"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The prediction that electron-electron scattering sets the plasma decay rate at about $10^{14}$ $s^{-1}$ rests on the plasmon-pole approximation for the electron-electron self-energy; if that approximation is quantitatively wrong, the lifetime and the practical value of the material would change.","fun_headline_variants_meta":{"raw":{"variants":["Sodium-stuffed hBN Moire silences phonon decay for 1 eV plasmons","Moire-trapped sodium gives lossless plasmons at 1 eV","Sodium intercalation kills phonon losses in Moire plasmons","hBN Moire with sodium: phonon-free plasmon window at 1.1 eV"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001097,"raw_usage":{"total_tokens":4641,"prompt_tokens":1068,"completion_tokens":3573,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":684,"completion_tokens_details":{"reasoning_tokens":3485}},"tokens_in":684,"tokens_out":3573,"duration_ms":28128,"temperature":1.0,"reasoning_tokens":3485,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T21:14:28.143018+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the electron-electron self-energy without the plasmon-pole approximation (for example, with a full RPA dielectric function or a quantum Monte Carlo approach) and compare the resulting plasmon decay rate in the 1.08–1.16 eV window to the paper's ~$10^{14}$ $s^{-1}$ prediction; a rate differing by an order of magnitude would invalidate the central lifetime claim. Alternatively, if a fabricated sodium-intercalated √7×√7 hBN sample shows first-order phonon-assisted plasmon decay below 1.16 eV in optical or near-field spectroscopic measurements, the band-structure picture would be wrong.","supporting_citations":[{"cited_title":"Ghorashi, N","cited_arxiv_id":null,"evidence_quote":"Prior work by the same group establishing the flat-band design principle and the one-phonon loss formula for defect lattices in hBN, which the current paper extends to Moiré heterostructures."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Allen's general self-energy expression for infrared conductivity, which is the basis of Eq. (2) for the plasmon decay rate including finite bandwidth and higher-order electron-phonon processes."},{"cited_title":"Jablan and D","cited_arxiv_id":null,"evidence_quote":"Jablan and Chang's treatment of multiplasmon absorption in graphene, which the authors generalize to derive the plasmon-plasmon scattering contribution to the decay rate."},{"cited_title":"Lon ˇcari´c, Z","cited_arxiv_id":null,"evidence_quote":"Prior work on Li-intercalated hBN films showing a strong two-dimensional plasmon with low damping, providing the experimental and theoretical context the authors differentiate from their large-angle Moiré approach."},{"cited_title":"Sundararaman, K","cited_arxiv_id":null,"evidence_quote":"The DFT code used for structural relaxation and band-structure calculations, on which the identification of the sodium-induced isolated band and the phonon stability test depend."}],"review_version":1}