{"id":"eda17633-1736-414b-a491-8291f4606183","arxiv_id":"2412.14839","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"A theoretical study predicts oscillatory time-domain photoemission from bound-state excitations, a revised adiabaticity criterion for Fermi gases, and a semi-quantitative sticking coefficient for H on Cu.","lead":"This thesis analyzes three out-of-equilibrium electron gas problems. It predicts that x-ray photoemission from metals should show decaying oscillations in time, derives a new criterion for when a metal responds adiabatically to a slowly growing potential, and computes the sticking probability of hydrogen atoms on copper surfaces, finding a peak near 0.3 eV.","discovery_kind":"first_principles","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The gapless-limit adiabaticity break and critical Kc rest on a one-pair truncation and a fitted 4.4 prefactor that are not tested at the threshold as Δε→0; an eNRG run at a still smaller gap would settle it.","rationale":"The reader's weakest_assumption identifies the single-particle-hole truncation and the fitted prefactor 4.4 as the basis of the adiabaticity criterion. My stress test agrees and sharpens it: the missing check is whether the truncation and the 4.4 prefactor survive as the gap is reduced toward the gapless limit, which is exactly where the claimed violation of the adiabatic theorem lives. The paper contains genuine supporting evidence — direct diagonalization and eNRG cross-checks in Chapter 3, and eNRG agreement with Eq. (4.9) for |c0|^2 down to about 0.6 in Fig. 22 — so the concern is not that the numerics are fabricated or that the truncated equations are useless. It is that the most novel and most sweeping claim, the existence of a critical Kc in the continuum limit, rests on an extrapolation beyond the validated range. The algebraic discrepancy between Eq. (4.13) at T=Tm and Eq. (4.15) is a concrete internal warning that the critical-value derivation is not fully transparent. A single eNRG run at N=13, at and around the claimed critical boundary, would directly test whether the truncation and the prefactor hold at smaller Δε. Since this is an addressable numerical check rather than a demonstrated contradiction, the reader's CONDITIONAL verdict remains appropriate; no verdict change is required.","tokens_in":57859,"tokens_out":12866,"duration_ms":115618,"concrete_test":"Use the eNRG method described in §3.4 with λ=2 and N=13 (Δε≈1/3370) to compute |c0(T)|^2 for a ramp of duration T=Tm=ℏ/Δε and amplitude set to the claimed critical value ρ|Kc|=η/√4.4 with η=0.1, plus one point at 20% above and one at 20% below, and compare with Eq. (4.10) and with the N=11 eNRG results from Fig. 23. If the eNRG trace distance at the claimed boundary exceeds η by a margin larger than the N=11 finite-size scatter, or if the extracted effective prefactor shifts by more than 10% between N=11 and N=13, the one-pair truncation plus fitted 4.4 do not survive the continuum extrapolation and the critical-Kc conclusion is not established. If instead the N=13 points fall on the same scaling curve, the extrapolation is supported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim of §4 — that a gapless Fermi gas has a critical ramp amplitude ρ|Kc|=η/√4.4 beyond which evolution is never adiabatic — depends on Eq. (4.10), which in turn rests on the single-particle-hole truncation of Eq. (4.9). The eNRG comparison in Fig. 22 validates the truncated equations only for |c0|^2≳0.6, whereas the adiabatic threshold η=0.1 corresponds to |c0|^2≈0.99, a different corner of parameter space; the manuscript does not show that two-pair excitations remain negligible there as Δε→0. The prefactor 4.4 is fitted to the truncated numerical solution, not derived, and Fig. 23 tests only two finite gaps (Δε=1/210 and 1/843), so the step 'we expect this expression to hold as Δε→0' is an unsupported extrapolation. Additionally, setting T=Tm=ℏ/Δε in Eq. (4.13) gives ρ|K|≤η/√(ln 4.4), not η/√4.4, so the stated critical value appears algebraically inconsistent unless an approximation is missing. The claim of violating the adiabatic theorem therefore rests on precisely the least-tested ingredients of the model.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript, based on the author's doctoral thesis, studies three out-of-equilibrium electron-gas problems. Chapter 3 treats x-ray photoemission from a metal after a sudden core-hole potential, deriving analytical expressions for the fidelity and spectral function that include excitations out of the bound state; it predicts oscillations in the time-domain fidelity with frequency |ε_B|/ℏ and a satellite peak obeying the Nozières–De Dominicis power law, supported by direct diagonalization and eNRG. Chapter 4 studies a Fermi gas with a linearly ramped local potential and claims that adiabaticity is controlled by the number of participating energy scales rather than by the ramp rate, leading to Eq. (4.10), the threshold Eq. (4.13), and a critical potential Eq. (4.15). Chapter 7 applies NRG and Crank-Nicolson to H-Cu collisions and computes a sticking coefficient with a maximum near 0.3 eV, in semi-quantitative agreement with experiment.","tokens_in":58161,"tokens_out":6603,"duration_ms":45716,"significance":"The photoemission part is the strongest: the bound-state/unplugged-state decomposition yields explicit, falsifiable predictions—oscillations at |ε_B|/ℏ whose amplitude decays faster than the Doniach-Sunjic envelope—and these are checked against direct diagonalization and eNRG. The eNRG/block-diagonal methodology is a practical computational contribution. The adiabaticity claim, if correct, is conceptually important because it challenges the Quantum Adiabatic Criterion in gapless systems. However, the central quantitative conclusions in §4 rest on a fitted prefactor and an extrapolation, and the critical-value formula contains an algebraic inconsistency. The sticking-coefficient model is less analytically developed but includes a transparent numerical treatment and a semi-quantitative comparison with experiment.","major_comments":[{"comment":"Setting T=Tm=ℏ/Δε in Eq. (4.13) yields ρ|K| ≤ η (ln 4.4)^{-1/2} ≈ 0.675 η, not η/√4.4 ≈ 0.477 η. Eq. (4.15) therefore does not follow algebraically from Eq. (4.13). Since the critical amplitude is a central claim of the chapter, the manuscript must either correct Eq. (4.15) and the associated text (including the statement 'ρ| ¯K|≤ η√4.4' in §4.2) or identify the missing approximation that changes ln 4.4 into 4.4. This is not a peripheral typo: the existence and value of Kc are used to argue for violation of the adiabatic theorem in the gapless limit.","section":"§4.2, Eqs. (4.13) and (4.15)"},{"comment":"The constant 4.4 in Eq. (4.10) is introduced by rescaling the horizontal axis in Fig. 21 so that the numerical solution of the truncated system Eq. (4.9) collapses onto a single curve; it is therefore a fit to the very equations used to generate |c0(T)|^2. The exponents are derived in Appendix F, but the prefactor is not. Because Eq. (4.15) inherits this prefactor, the quantitative critical potential ρ|Kc|=η/√4.4 is not a parameter-free prediction. The manuscript should present the determination of 4.4 as an empirical constant, estimate its uncertainty or sensitivity, or derive it analytically.","section":"§4, Eq. (4.10) and Fig. 21"},{"comment":"The eNRG comparison validates Eq. (4.9) down to |c0|^2≈0.6 for two finite gaps (Δε=1/210 and 1/843), while the adiabatic threshold η=0.1 corresponds to |c0|^2≈0.99, a different corner of parameter space. The sentence 'We expect therefore this expression to hold as Δε→0' is an extrapolation beyond the tested regime: as Δε→0 the number of low-energy single-particle levels diverges, and the validity of the single-pair truncation at the threshold is not established. A direct eNRG run at a smaller gap (e.g., Δε∼1/3000 with appropriate λ and N) targeting the threshold region (T close to Tm, |c0|^2≈0.99) would substantially strengthen the claim. Without it, the critical-Kc conclusion rests on the least-tested ingredient of the model.","section":"§4.2, Figs. 22–23 and text after Eq. (4.13)"}],"minor_comments":[{"comment":"The manuscript has numerous typos and grammatical errors, including 'INTRODCUTION' in the Contents, 'beown plots' in §4.2, and 'sticking coefficient for atoms impinging on a metallic surfaces' in the title; a careful language edit is needed.","section":"Global"},{"comment":"The numerical constants 600 and 1.25 in Eq. (3.13) are introduced without derivation in the main text; Appendix B should state explicitly whether they are fitted or derived from an asymptotic expansion.","section":"Eqs. (3.13)–(3.14)"},{"comment":"The caption says both the solid and dashed lines are computed by Eq. (2.8); one of them should refer to the constant-δ approximation or another distinct expression.","section":"Fig. 39 caption"},{"comment":"The notation 'atan' should be replaced by 'arctan' or 'tan^{-1}', and the sign convention for the phase shift δ should be stated once and used consistently.","section":"Eq. (2.8)"},{"comment":"Several parameter values in §7 are attributed to DFT calculations without complete citations; please add full references for the DFT-derived values of V0, zim, D, and the Cu surface parameters.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is clearly a doctoral thesis and is much longer than a typical research article, with textbook-style chapters (Ch. 1, Ch. 2, Ch. 5) that could be condensed. The photoemission part is publishable and well supported. Before external review, the authors should correct the algebraic slip connecting Eqs. (4.13) and (4.15), reframe the prefactor 4.4 as a fitted constant, and provide the suggested eNRG test at a smaller gap. The fit between the manuscript and the journal scope is reasonable for cond-mat.mtrl-sci, but the presentation needs substantial streamlining."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"It's a PhD thesis, not a compact paper, and the three parts are of uneven quality. The photoemission work is the strongest. The time-domain fidelity oscillation from interference between plugged and unplugged states (Eq. 3.38) is new, and the paper supports it with both direct diagonalization and eNRG. The analytic bound-state expressions (Eqs. 3.8, 3.9) are checked numerically, and the extension beyond Combescot-Nozières and Ohtaka-Tanabe is real. The silver XPS satellite comparison is qualitative but honest about its limits.\n\nThe adiabaticity section is the flagship, and it's where I have the most trouble. The criterion in Eq. (4.10) rests on the single-particle-hole truncation of Eq. (4.9) and on a fitted prefactor 4.4. The eNRG comparison is good for |c0|^2 above about 0.6, but the adiabatic threshold lives near |c0|^2 ≈ 0.99, which is a different corner of parameter space. The extrapolation to Δε → 0 is asserted, not demonstrated. There is also an algebraic inconsistency: setting T = Tm in Eq. (4.13) gives ρ|K| ≤ η/√(ln 4.4), not η/√4.4. The stated critical Kc therefore does not follow from the displayed equations unless an approximation is missing. That needs to be fixed or explained before the claim about violating the adiabatic theorem in the gapless limit is publishable. I don't think the claim is crazy, but it's resting on the least-tested ingredients.\n\nThe sticking coefficient calculation is suggestive, not definitive. The spinless model, the heavily truncated band, and the absence of code, data, and error bars make the semi-quantitative agreement with experiment a proof of principle rather than a quantitative result. Still, the eNRG machinery and the block-diagonal approximation are carefully tested against exact diagonalization in the appendices, and that is real, reproducible method work.\n\nWho is this for? People working on time-domain XPS, nonequilibrium Fermi gases, and atom-surface collisions. The photoemission part deserves a serious referee. The adiabaticity claim should be revised or softened, and the Kc algebra must be corrected. I would send this to peer review, with a request for a smaller-gap eNRG run or an explicit statement of the truncation's limitation, plus release of the code and data. If those are provided, the adiabaticity claim could become testable rather than merely interesting.","headline":"A thesis with one solid photoemission result, one interesting but under-supported adiabaticity claim, and one suggestive sticking calculation; referee-worthy but needs revision before the gapless-breakdown claim is published.","tokens_in":58707,"tokens_out":1899,"would_cite":false,"duration_ms":15674,"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":"This paper claims that the adiabatic theorem fails for a gapless Fermi gas beyond a critical localized-potential strength, and that the same electron-gas physics produces oscillatory XPS and a 0.3 eV sticking peak for hydrogen on copper.","keywords":["sticking coefficient","x-ray photoemission","Anderson orthogonality catastrophe","adiabatic theorem","numerical renormalization group","bound state","hydrogen-copper collision","Doniach-Sunjic power law"],"falsifier":"Rerun the eNRG calculation with two-particle-hole excitations included for parameter sets on the adiabatic boundary of Fig. 23; if the boundary shifts substantially, the single-pair truncation and the critical potential are falsified, and a time-resolved XPS scan looking for the $|\\epsilon_B|/\\hbar$ oscillation would test the photoemission interference claim.","tokens_in":57603,"feed_emoji":"⚛️","tokens_out":10612,"duration_ms":76010,"temperature":0.7,"pith_summary":"The paper sets out to unify three out-of-equilibrium electron-gas problems: x-ray photoemission from metals, the response of a Fermi gas to a slowly growing localized potential, and the collision of a neutral hydrogen atom with a copper surface. Its central claim is that the conventional adiabatic criterion, which says slow ramps are adiabatic, is wrong for a gapless electron gas: what matters is the number of energy scales involved in screening, with a universal formula giving the ground-state survival probability and a critical potential beyond which adiabatic evolution is impossible. In the same framework, the paper explains the time-dependent photoemission rate as interference between two decay channels, and computes a sticking coefficient that peaks near 0.3 eV, in semi-quantitative agreement with experiment. A sympathetic reader would care because the results give testable predictions for XPS satellite peaks and for when non-adiabatic effects in atom-surface collisions cannot be ignored.","feed_headline":"A critical potential makes a gapless Fermi gas never adiabatic","feed_subtitle":"Beyond a threshold scattering strength, no ramp time keeps the system in its ground state, reshaping the adiabatic criterion.","key_machinery":"The central object is the instantaneous many-body eigenbasis of the time-dependent Hamiltonian, together with the derivative coupling $\\langle\\varphi_n|\\partial_t|\\varphi_m\\rangle$ that drives transitions between instantaneous eigenstates. For a continuous ramp this coupling connects only states differing by one particle-hole pair, which reduces the many-body Schrödinger equation to the closed set of equations (4.9); numerical solution of that set collapses onto the universal law (4.10), from which the adiabatic threshold and the critical potential follow. In the photoemission problem the same basis is split into 'plugged' states, with the bound level occupied and decaying by the Doniach-Sunjic power law, and 'unplugged' states, with the bound level empty and decaying by the Nozières-De Dominicis power law; their phase difference is what produces the predicted oscillations.","core_discovery":"For an electron gas subject to a localized potential ramped linearly from zero to $\\bar K$ over time $T$, the probability that the system remains in the instantaneous ground state is $|c_0(T)|^2 = (4.4\\hbar/(\\Delta\\varepsilon T))^{-(\\delta/\\pi)^2(1+(\\delta/\\pi)^2)}$, where $\\delta$ is the scattering phase shift and $\\Delta\\varepsilon$ the single-particle level spacing. Adiabaticity is therefore controlled by the combination $\\Delta\\varepsilon T/\\hbar$ and by the potential strength, not by the ramp rate $\\bar K/T$ as the Quantum Adiabatic Criterion would have it. The adiabatic region is bounded by $\\rho|\\bar K| \\leq \\eta (\\log(4.4\\hbar/(\\Delta\\varepsilon T)))^{-1/2}$, and in the continuum limit there is a critical potential $\\rho|\\bar K_c| = \\eta/\\sqrt{4.4}$ beyond which no ramp time $T \\leq \\hbar/\\Delta\\varepsilon$ yields adiabatic evolution, a violation of the naive adiabatic theorem in a gapless system. The same bound-state physics, in which a strong core-hole potential creates a level below the band, makes the photoemission fidelity oscillate at frequency $|\\epsilon_B|/\\hbar$ as two power-law decay channels interfere, and the collision calculation yields a sticking coefficient peaked near 0.3 eV.","pith_inferences":["If the bound-state interference picture is correct, time-resolved XPS on any simple metal with a strong core-hole potential should see the predicted $|\\epsilon_B|/\\hbar$ oscillations, offering a direct test outside the collision context.","The critical-potential result suggests that gapless many-body systems beyond the single-particle picture, such as interacting metals near quantum critical points, may also fail to follow slow ramps once a local perturbation exceeds a threshold.","The universal prefactor 4.4 is fitted to a single-particle-hole calculation; a calculation that includes two-particle-hole excitations could shift the threshold and the critical potential, so the quantitative value should be treated with caution.","The spinless H-Cu model could be extended to include spin degrees of freedom and phonon channels; the semi-quantitative agreement with experiment suggests those extensions would refine, not overturn, the peak near 0.3 eV."],"forward_implications":["X-ray photoemission spectra of simple metals should show a satellite peak displaced from the main threshold by the bound-state energy $|\\epsilon_B|$, decaying with the Nozières-De Dominicis exponent.","Time-domain photoemission measurements should display oscillations at frequency $|\\epsilon_B|/\\hbar$ whose amplitude decays faster than the average current.","The Quantum Adiabatic Criterion is unreliable for gapless fermionic systems; adiabaticity estimates should be replaced by the threshold $\\rho|\\bar K| \\leq \\eta (\\log(4.4\\hbar/(\\Delta\\varepsilon T)))^{-1/2}$.","For a strictly gapless band, no slow ramp can adiabatically turn on a localized potential stronger than $\\rho|\\bar K_c| = \\eta/\\sqrt{4.4}$, whereas a finite gap restores adiabaticity for sufficiently long ramp times.","The H-Cu sticking coefficient is controlled by a tradeoff between growing non-adiabatic energy loss and traversal time through the interaction region, peaking near 0.3 eV."],"supporting_citations":[{"why":"supplies the Anderson orthogonality catastrophe that makes many excited states essential after a sudden potential change.","marker":"(44)"},{"why":"provides the Nozières-De Dominicis power law used for the unplugged bound-state contribution.","marker":"(46)"},{"why":"provides the Doniach-Sunjic power law used as the baseline for the plugged contribution to photoemission.","marker":"(47)"},{"why":"established the bound-state influence on x-ray spectra, which the paper extends to photoemission and the time domain.","marker":"(48)"},{"why":"showed the extra peak in x-ray photoemission spectra that the paper makes quantitative.","marker":"(51)"},{"why":"states the adiabatic theorem that the paper argues fails for a gapless Fermi gas.","marker":"(64)"},{"why":"supplies the numerical renormalization group iterative diagonalization used for the electronic Hamiltonians.","marker":"(2)"},{"why":"supplies the real-space eNRG variant whose logarithmic discretization makes long-time dynamics tractable.","marker":"(77)"},{"why":"supplies the smoothing procedure that removes discretization artifacts in eNRG results.","marker":"(78)"},{"why":"supplies the Crank-Nicolson time-stepping used to evolve the nuclear wave packet in the collision.","marker":"(43)"}],"fun_headline_variants":["Gapless Fermi gas never adiabatic beyond critical potential","Critical potential forbids adiabatic evolution in electron gas","No ramp time saves adiabaticity past critical scattering strength","Photoemission decays with oscillations from two interfering power laws","Sticking coefficient on copper peaks near 0.3 eV"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The adiabaticity criterion assumes that during a slow ramp the electron gas never needs more than one particle-hole pair at a time; if multiple pairs contribute significantly near the threshold, the universal formula and the critical potential do not follow.","fun_headline_variants_meta":{"raw":{"variants":["Gapless Fermi gas never adiabatic beyond critical potential","Critical potential forbids adiabatic evolution in electron gas","No ramp time saves adiabaticity past critical scattering strength","Photoemission decays with oscillations from two interfering power laws","Sticking coefficient on copper peaks near 0.3 eV"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001138,"raw_usage":{"total_tokens":4816,"prompt_tokens":1128,"completion_tokens":3688,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":744,"completion_tokens_details":{"reasoning_tokens":3608}},"tokens_in":744,"tokens_out":3688,"duration_ms":22736,"temperature":1.0,"reasoning_tokens":3608,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T11:51:44.886717+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Rerun the eNRG calculation with two-particle-hole excitations included for parameter sets on the adiabatic boundary of Fig. 23; if the boundary shifts substantially, the single-pair truncation and the critical potential are falsified, and a time-resolved XPS scan looking for the $|\\epsilon_B|/\\hbar$ oscillation would test the photoemission interference claim.","supporting_citations":[],"review_version":1}