{"id":"c5651a43-11ee-45dd-bfe3-f92dbde80c38","arxiv_id":"2507.07500","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"Resonance scattering off disordered alkali-metal ions on black phosphorus produces a backbending band and a pseudogap, with sodium giving p-wave and potassium, rubidium, and cesium giving d-wave behavior.","lead":"A team measured the electronic structure of black phosphorus with alkali metal atoms sprinkled on its surface and found that the electrons' band bends back near zero momentum with a pseudogap, a signature long predicted for disordered liquid-like metals. The result gives a concrete experimental example of resonance-scattering band renormalization and offers a possible explanation for waterfall-like dispersion seen in cuprates and other doped insulators.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The central claim rests on an untested structural premise: the ARPES gap/backbending is attributed to disorder-induced resonance, but no in-situ structural data at the measured densities establish the required liquid-like layer, and the paper concedes d-wave gaps are indistinguishable from…","rationale":"The reader's weakest_assumption captures the same issue, and I agree. The paper's external phase-shift calculation (ED Fig. 8) is real independent evidence that the partial-wave character can differ between Na and K/Rb/Cs, but it cannot distinguish a resonance-scattering pseudogap from an ordinary hybridization or symmetry-breaking gap, because the latter can also produce a species-dependent gap at EF. The strongest direct evidence for the liquid-metal mechanism is the backbending in Na; for K/Rb/Cs, the data are only a gap, and the authors explicitly concede the ambiguity. The structural premise is even more directly load-bearing: the entire argument that Er sits at EF relies on kr tracking kF, which is asserted from hard-sphere simulations and one low-density STM measurement. The extrapolation to the densities of the ARPES data is uncontrolled; alkali adlayers are known to order at higher coverages on many surfaces. The proposed LEED/STM check would settle the matter. Because the paper is already CONDITIONAL in the reader's verdict and the concern is exactly the stated condition, no change in verdict is needed, though the condition should be stated as a required experimental check rather than a request for fitting details alone.","tokens_in":13811,"tokens_out":13624,"duration_ms":161613,"concrete_test":"Perform in-situ LEED and/or STM on the same alkali-doped black phosphorus surfaces at the exact coverages, temperatures, and dosing rates used for the ARPES data in Fig. 2 and ED Fig. 9. Test two quantitative signatures: (1) absence of any sharp superstructure spots (no long-range order) and (2) presence of a diffuse structure-factor ring whose half-radius kr equals kF of the C1 band extracted from the simultaneously measured ARPES dispersion, within the resonance linewidth (roughly +/- 0.08 A-1). If an ordered phase appears at these coverages, or if kr deviates from kF, the resonance-scattering interpretation is falsified in favor of a conventional gap.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The identification of the C1 pseudogap/backbending as liquid-metal resonance scattering hinges on two conditions: (i) the alkali layer is actually disordered with a short-range-order structure-factor peak 2kr matching 2kF, and (ii) no conventional gap mechanism (band bending, surface charging, adsorbate hybridization, ordered reconstruction, or matrix-element suppression) produces the same spectra. The paper's structural evidence is extrapolated: Extended Data Fig. 1 and the cited STM study (Ref. 25) give a radial/anisotropic structure factor at nd = 1.8e13 cm-2, whereas the key ARPES data in Fig. 2 are at nd ~ 1e14 cm-2 (Methods; ED Fig. 9). No LEED/STM data at these higher coverages are shown. For K, Rb, and Cs, the only observed feature is a gaplike suppression; the paper itself states that such d-wave pseudogap spectra are difficult to distinguish from genuine energy gaps unless band folding or Bogoliubov bands are seen (Discussion), and no such signatures are presented. The systematic p-wave/d-wave variation provides indirect support via the external screened-potential calculation (ED Fig. 8), but it does not establish that the dopant layer is disordered at the experimental coverages, nor that the gap is not a conventional hybridization gap. Thus the most load-bearing premise — the disorder origin of the pseudogap — is plausible but not uniquely supported.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This manuscript reports angle-resolved photoemission spectroscopy (ARPES) measurements of bulk black phosphorus whose surface is doped by Na, K, Rb, or Cs. The authors observe that the C1 conduction band, instead of simply crossing the Fermi level as expected from rigid band filling, exhibits a gaplike suppression near E_F, with a backward-bending dispersion for Na and a more abrupt gap-like suppression for K, Rb, and Cs. They interpret this as the long-predicted band-structure renormalization and pseudogap of liquid-metal theory, caused by resonance scattering of the doped electrons from disordered alkali ions. A single-site step-potential model is used to compute partial-wave phase shifts, classify the Na case as p-wave and the K/Rb/Cs cases as d-wave resonance, and simulate ARPES spectra. The doping dependence of the pseudogap is compiled into a phase diagram, and the model is also applied to reproduce the waterfall dispersion in a cuprate.","tokens_in":14077,"tokens_out":8116,"duration_ms":89628,"significance":"If correct, this would be the first direct observation of the k-space renormalization and pseudogap predicted by Edwards, Ziman, Anderson-McMillan, and others for electrons in liquid-like disordered media, and it would provide a unifying framework for puzzling spectral features in doped insulators. The experimental core is solid in several respects: the doping series is systematic, the photon-energy and sample dependence are checked (Extended Data Fig. 4), and the raw ARPES images (Fig. 2d-o, Extended Data Fig. 9) show a reproducible gap-like feature that is not an artifact of a single cleave or photon energy. The external screened-potential phase-shift calculation (Extended Data Fig. 8), which yields p-wave for Na and d-wave for K/Rb/Cs without using the ARPES spectra as input, is a genuine independent prediction and a significant strength. The main weaknesses are that the structural disorder premise is not directly established at the experimental coverages, and that the quantitative simulations are fit-based, so the classification is less secure than the presentation suggests.","major_comments":[{"comment":"The attribution of the pseudogap to resonance scattering from disordered alkali ions requires that the dopant layer is actually liquid-like (short-range order only) at the coverages of the ARPES data. The only structural evidence cited is the STM study (Ref. 25) at nd = 1.8 x 10^13 cm^-2, while the key ARPES data in Fig. 2 and Extended Data Fig. 9 are at nd ~ 1 x 10^14 cm^-2 and above. Extended Data Fig. 1 is explicitly a hard-sphere simulation, not a measurement, and it shows that a radial structure factor is always obtained under that assumption. No LEED, STM, or diffraction data at the experimental coverages are presented. Because band bending, surface charging, ordered reconstruction, or adsorbate hybridization could produce a similar spectral suppression, the disorder origin is not uniquely supported. Please provide structural characterization at the relevant nd range, or state a concrete falsifiable prediction that distinguishes the liquid-like layer from these alternatives.","section":"Doping dependence and phase diagram; Methods: Structural simulations"},{"comment":"The Discussion concedes that d-wave pseudogap spectra are difficult to distinguish from genuine energy gaps unless band folding or Bogoliubov bands are observed. For K, Rb, and Cs, the only observed signature is the gap-like suppression, and no folded bands or Bogoliubov-like features are shown. The classification of these species as d-wave therefore rests on the fitted single-site model and on the external phase-shift calculation in Extended Data Fig. 8, rather than on a uniquely identifying experimental signature. The systematic trend across alkali species is suggestive but does not exclude a conventional hybridization or charge-order gap whose magnitude scales with ionic radius. An additional observable, such as a control with an ordered alkali layer (e.g., low-temperature ordered phase or different deposition protocol) or a search for the predicted spectral weight inside the gap at higher sensitivity, would be needed to secure the claim.","section":"Discussion; Fig. 2g-o and Fig. 4h-j"},{"comment":"The simulated spectra in Fig. 4e-j and Extended Data Fig. 7 are produced with V0 and eta optimized to reproduce the measured pseudogap magnitudes (V0 = 7.15 eV for Na and 16.26 eV for K, eta = 0.03-0.14 Angstrom^-1), and the same fitted parameters are then presented as agreement with experiment. This is parameter fitting rather than an a priori prediction, and it makes the p-wave/d-wave classification of the simulations partly circular. The genuinely independent input is the screened-potential calculation in Extended Data Fig. 8; it should be made the primary quantitative evidence. Please separate fitted from predicted quantities and report the sensitivity of the classification and of the pseudogap magnitude to the assumed rs (fixed at 2.1 Angstrom for all species) and to the functional form of eta.","section":"Methods: Spectral simulations; Fig. 4d"}],"minor_comments":[{"comment":"The word 'APRES' in the caption of Fig. 4 should be corrected to 'ARPES'.","section":"Fig. 4 caption"},{"comment":"The phrase 'unexceptionally observed' should be replaced by a standard expression such as 'without exception' or 'invariably observed'.","section":"Main text, pages 3 and 5"},{"comment":"The ML unit is defined using the close-packed density of K atoms but is applied to Na, Rb, and Cs as well; please state the justification or use species-specific densities.","section":"Methods: Sample preparation and surface doping"},{"comment":"The definition of the pseudogap magnitude as the energy at which the integrated spectral weight drops by half relative to E_F should be justified against background subtraction and the k-integration range; a small sensitivity analysis would help.","section":"Fig. 4d and Methods: ARPES experiments and analysis"},{"comment":"The grey region in Fig. 5b is not explicitly defined in the main text; please state the criterion (e.g., pseudogap greater than zero) and the uncertainty of the phase boundary.","section":"Fig. 5b"}],"recommendation":"major_revision","confidential_remarks":"The raw experimental observation is reproducible and of considerable interest. The main risk is that the disorder-resonance interpretation is not uniquely supported without structural data at the relevant coverages. The fit-based simulations should be repositioned as an illustration rather than as confirmation. If the authors can supply in-situ structural evidence or a decisive control, the paper would be a strong candidate for publication."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"First thing to know: this paper has a genuinely new experimental result. The ARPES data from alkali-doped black phosphorus show a reproducible backbending and gap-like suppression in the C1 band, with magnitudes that vary systematically from ~235 meV for Na down to ~30 meV for Cs. The data are shown across doping series, photon energies, and multiple samples, and the raw signatures look robust. If the authors are right, this is the first observation of the resonance-scattering band renormalization and pseudogap predicted for liquid metals fifty years ago.\n\nWhat is good beyond the data: the partial-wave classification is supported by an external screened-potential calculation (ED Fig. 8), which is the strongest non-circular evidence. The step-potential model is crude, but it captures the difference between Na (p-wave, with visible backbending) and K/Rb/Cs (d-wave, with a cleaner dip). The authors also explicitly acknowledge the main ambiguity—d-wave pseudogap spectra are hard to tell apart from ordinary energy gaps without band folding or Bogoliubov bands—and they lean on the systematics across alkali species.\n\nNow the soft spots, in proportion. The most load-bearing premise is that the alkali layer is actually disordered with a structure-factor peak at 2kr matching 2kF. The structural evidence is extrapolated: the cited STM study and the simulations are at nd ~ 4e13 or lower, while the key ARPES data in Fig. 2 are at ~1e14. No in-situ LEED or STM at those higher coverages is shown, so we don't know whether the layer remains liquid-like, forms an ordered phase, or clusters. That matters because an ordered reconstruction or a hybridization gap could produce similar spectra. The stress-test note has this right.\n\nSecond, V0 and eta are fitted to the measured pseudogap magnitudes, and the simulations use those same fitted parameters. That is not fatal, but it makes the quantitative agreement less impressive than it looks. The screened-potential calculation helps, but it doesn't fix the structural uncertainty.\n\nThird, the cuprate waterfall comparison in ED Fig. 10 is a post-hoc analogy with a few adjustable inputs (V0, rs, eta). It should not be read as evidence for the main claim, and the paper would be stronger if it were framed as a speculative remark.\n\nWho this is for: anyone working on ARPES of doped insulators, on the electronic structure of disordered adsorbates, or on the old liquid-metal theory. It deserves a serious referee: the observation is new, the data are shown honestly, and the interpretation is a leading candidate even if not uniquely established. I would encourage engagement, with the expectation that the structural premise will need direct testing.","headline":"A credible first observation of the long-predicted liquid-metal pseudogap and backbending in a disordered dopant layer; the interpretation is plausible but the structural premise is extrapolated and the d-wave cases are hard to separate from ordinary gaps.","tokens_in":14659,"tokens_out":3057,"would_cite":true,"duration_ms":34501,"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":"Disordered alkali atoms on black phosphorus bend the conduction band backward and open a 30–240 meV pseudogap at the Fermi level — the liquid-metal band renormalization seen for the first time.","keywords":["pseudogap","resonance scattering","backbending dispersion","black phosphorus","alkali doping","liquid-metal band theory","ARPES","disordered dopants"],"falsifier":"Check the real-space arrangement of the alkali atoms at the doping densities used here (e.g., by scanning tunnelling microscopy or low-energy electron diffraction) and compare the structure factor with the assumed liquid-like radial form with $k_r$ matching $k_F$. If the overlayer is ordered at these densities, or if a band-bending calculation without any resonance scattering reproduces the same backbending and gap, the central claim would be refuted. A complementary test: if the same electron density supplied by an ordered intercalant or an external gate also produces the pseudogap, the disorder-resonance explanation is undercut.","tokens_in":13578,"feed_emoji":"⚛️","tokens_out":14737,"duration_ms":131102,"temperature":0.7,"pith_summary":"Alkali atoms deposited on the surface of black phosphorus are disordered, and this paper argues that the disorder itself is the engine of the observed spectrum. Instead of the rigid downward shift expected from simple surface doping, the C1 conduction band bends back toward zero momentum and loses spectral weight near the Fermi level, leaving a pseudogap of 30–240 meV depending on the alkali species. The paper attributes this to resonance scattering of electrons by randomly arranged alkali ions, a band-structure renormalization predicted for liquid metals fifty years ago but never directly observed. The gap and backbending are reproduced by a single-site multiple-scattering model, with Na giving p-wave resonance and K, Rb, Cs giving d-wave resonance, and the same simulation reproduces the waterfall dispersion measured in a cuprate.","feed_headline":"Alkali atoms on black phosphorus bend electron bands backward","feed_subtitle":"50-year-old prediction of electron-band renormalization by resonance scattering is seen at a doped insulator surface.","key_machinery":"The load-bearing object is the complex wavenumber shift $\\Delta k$ acquired by an electron wave multiply scattered by disordered ions. In the thin-slab approximation $\\Delta k \\sim 2\\pi n_i f_l / k$, with partial-wave amplitude $f_l = \\sin\\delta_l\\, e^{i\\delta_l}/k$, so $\\mathrm{Re}(\\Delta k) \\propto \\sin(2\\delta_l)/k^2$ and $\\mathrm{Im}(\\Delta k) \\propto \\sin^2(\\delta_l)/k^2$. Each alkali ion is modelled as a spherical step potential of depth $V_0$ and radius $r_s$; when the phase shift $\\delta_l$ passes through $\\pi/2$ in a nonzero partial wave, the real part distorts the free-electron parabola into a sinusoidal backbending form and the imaginary part spreads the momentum distribution, corresponding to quasi-bound states around the ions. The resulting local minimum in the density of states is the pseudogap. Tuning $V_0$ across alkali species selects which partial wave resonates, giving the p-wave (Na) and d-wave (K, Rb, Cs) cases that match the measured spectra.","core_discovery":"On the paper's account, the C1 conduction band of alkali-doped black phosphorus exhibits a Fermi-level pseudogap and a backbending dispersion, with the band renormalized by a complex wavenumber shift $\\Delta k$ produced by resonance scattering from disordered alkali ions. The gap magnitude is about 235 meV for Na and 65 meV for K and Rb, 33 meV for Cs, and the band bends back toward the $\\Gamma$ point near $-$0.2 eV rather than crossing the Fermi level with a sharp cutoff. The same data, simulated with the single-site liquid-metal multiple-scattering model, show that Na favours p-wave resonance ($l=1$), giving an incomplete wider pseudogap with visible backbending, while K, Rb and Cs favour d-wave resonance ($l=2$), giving a sharper, smaller pseudogap with little spectral weight on the backbending branch. The pseudogap persists across photon energies, samples, and a wide doping range, and its magnitude tracks the balance between $k_F$ and the half structure-factor radius $k_r$ of the dopant layer. The paper further argues that the same mechanism reproduces the waterfall dispersion measured in a cuprate.","pith_inferences":["If the mechanism is general, other doped layered insulators with monovalent disordered dopants should show the same backbending and pseudogap, with a testable prediction that the gap size scales with the alkali ion scattering strength and the dopant layer's $k_r$ rather than with the bare band structure.","A direct local probe could check the picture beyond photoemission: scanning tunnelling spectroscopy should find the pseudogap suppression spatially correlated with the disordered ions, and tunnelling spectra should reveal p-wave versus d-wave character for Na versus K, Rb, Cs.","The authors hint at a possible pairing instability in the renormalized band; if real, strongly resonant disordered dopants on an insulator could become a controllable platform for superconductivity, but that is an untested speculation.","The cuprate waterfall resemblance implies that some 'high-energy anomalies' attributed to electronic correlations might receive a disorder-resonance contribution; varying the dopant disorder or species in those materials would separate the two."],"forward_implications":["This is the first direct observation of the $k$-renormalization and pseudogap predicted by liquid-metal band theory, realized at a crystalline-insulator/disordered-dopant interface.","The pseudogap magnitude and shape are controlled by the alkali species: Na gives p-wave resonance with a wider pseudogap of about 235 meV, while K, Rb and Cs give d-wave resonance with sharper pseudogaps of about 65, 65 and 33 meV, matching screened-potential partial-wave calculations.","The pseudogap is isotropic in magnitude even though the Fermi surface is anisotropic, because the dopant structure factor is itself anisotropic with its half-radius $k_r$ matched to $k_F$.","The same resonance-scattering simulation reproduces the waterfall dispersion measured in a cuprate, suggesting a common origin for puzzling high-energy spectral anomalies in doped insulators.","As doping increases and the C2 band crosses the Fermi level, the balance between $k_r$ and $k_F$ breaks and the pseudogap shrinks toward zero, defining a pseudogap phase in the dopant-density phase diagram."],"supporting_citations":[{"why":"Foundational theory of electronic structure in disordered systems that introduces the complex wavenumber shift and renormalized band used as the interpretive frame.","marker":"[1]"},{"why":"Self-consistent multiple-scattering model whose backbending can reach k = 0, cited for the extreme Na case.","marker":"[5]"},{"why":"Single-site structure-independent model and Equation (3.13) used for the density-of-states and spectral simulations.","marker":"[12]"},{"why":"Defines the pseudogap and the localized states inside it, used to interpret the spectral suppression.","marker":"[14]"},{"why":"Provides the experimental band structure and C1/C2 Fermi-surface assignment for the bare bands of black phosphorus.","marker":"[23]"},{"why":"STM evidence for the radial and anisotropic structure factor of alkali dopants on black phosphorus, used to justify kr and its anisotropy.","marker":"[25]"},{"why":"Pseudo-atom phase shifts of alkali ions that show Na favours p-wave and K/Rb/Cs favour d-wave resonance.","marker":"[30]"},{"why":"Documents the waterfall dispersion in cuprates that the paper's resonance-scattering simulation reproduces.","marker":"[16]"}],"fun_headline_variants":["50-year-old electron band bend finally seen in doped insulator","Alkali atoms bend electron bands on black phosphorus","Disordered metals reveal hidden pseudogap in black phosphorus","Backbending electron bands confirm 50-year-old prediction","Resonance scattering bends bands in doped black phosphorus"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the missing spectral weight and the backbending come from resonance scattering by randomly arranged alkali ions, not from an ordinary energy gap due to band bending, surface charging, an ordered reconstruction, or photoemission matrix-element effects, and that the disordered layer keeps a short-range-order length $k_r$ that tracks the Fermi wavevector $k_F$ so the resonance sits at the Fermi level.","fun_headline_variants_meta":{"raw":{"variants":["50-year-old electron band bend finally seen in doped insulator","Alkali atoms bend electron bands on black phosphorus","Disordered metals reveal hidden pseudogap in black phosphorus","Backbending electron bands confirm 50-year-old prediction","Resonance scattering bends bands in doped black phosphorus"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000958,"raw_usage":{"total_tokens":4119,"prompt_tokens":1020,"completion_tokens":3099,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":636,"completion_tokens_details":{"reasoning_tokens":3020}},"tokens_in":636,"tokens_out":3099,"duration_ms":21971,"temperature":1.0,"reasoning_tokens":3020,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T18:39:17.583257+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Check the real-space arrangement of the alkali atoms at the doping densities used here (e.g., by scanning tunnelling microscopy or low-energy electron diffraction) and compare the structure factor with the assumed liquid-like radial form with $k_r$ matching $k_F$. If the overlayer is ordered at these densities, or if a band-bending calculation without any resonance scattering reproduces the same backbending and gap, the central claim would be refuted. A complementary test: if the same electron density supplied by an ordered intercalant or an external gate also produces the pseudogap, the disorder-resonance explanation is undercut.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Foundational theory of electronic structure in disordered systems that introduces the complex wavenumber shift and renormalized band used as the interpretive frame."},{"cited_title":"& Lloyd, P","cited_arxiv_id":null,"evidence_quote":"Single-site structure-independent model and Equation (3.13) used for the density-of-states and spectral simulations."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Pseudo-atom phase shifts of alkali ions that show Na favours p-wave and K/Rb/Cs favour d-wave resonance."},{"cited_title":"M., Yi, M., Chen, Y., Moore R","cited_arxiv_id":null,"evidence_quote":"Documents the waterfall dispersion in cuprates that the paper's resonance-scattering simulation reproduces."}],"review_version":1}