{"id":"0af44a35-573b-411b-8e07-25db41ab5ee4","arxiv_id":"2411.12810","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"A non-perturbative model using atomic d-electron momentum distributions reproduces low-energy proton stopping in six transition metals and extends to 100 MeV when combined with standard shell contributions.","lead":"A new model describes how loosely bound d-electrons slow down protons in transition metals such as copper, silver, and gold. It combines atomic momentum distributions with a standard scattering formula to cover stopping power from very low to high energies.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The central claim rests on an untested assumption that free-atom d-orbital momentum distributions represent the solid d-electron response; a DFT-based d-projected momentum density calculation would settle it.","rationale":"I reviewed the paper in good faith. The model is clearly presented, and the comparison strategy with the IAEA database and with TDDFT results is appropriate; the equations are explicit and the numerical approach is reproducible in principle, though code and complete parameter tables are not provided. The reader's weakest assumption is also the one I identify: the use of free-atom d-orbital momentum distributions for the solid d-electron response. This is the load-bearing link between the atomic-structure input and the low-velocity stopping. My concern is not that atomic orbitals are outside current consensus; for localized d-bands in late transition metals they are a reasonable starting point. The issue is more specific: the low-momentum behavior of the distribution controls low-velocity stopping, and that is exactly where band effects and s-d hybridization are expected to modify the isolated-atom distribution. The paper provides no independent check of f_nd(p) against a solid-state calculation. I also note that the NFEG/Nd inference from plasmon frequencies is an integer-rounding procedure whose sensitivity is not quantified. These are testable concerns rather than fatal flaws. If the DFT-based test shows that the d-projected momentum density closely matches the atomic f_nd(p) in the sampled momentum range, the central claim would be supported; if not, the agreement with experiment would likely be fortuitous. I therefore keep the reader's conditional verdict rather than moving to accept or reject.","tokens_in":12801,"tokens_out":8131,"duration_ms":87188,"concrete_test":"For a representative metal (e.g., Cu and Au), compute the occupied d-projected momentum density from a converged plane-wave DFT calculation: n_d(p) = (1/N_k) Σ_{nk} |∫ d^3r ψ_{nk}(r) e^{-ip·r} P_d(r)|^2, with P_d a projection onto atomic d orbitals (e.g., via Wannier-projected d bands), spherically averaged and normalized to the same Nd as in Table I. Replace f_nd(p) in Eqs. (5)-(6) with (2π)^{3/2} n_d(p) and recompute the low-energy S(v). If the resulting curves deviate by more than about 10% from the model below v=0.5, or if the model's agreement with the experimental/TDDFT datasets in Figs. 3 and 7 is lost, the atomic-momentum assumption is invalid. Also run the sensitivity test NFEG→NFEG±1 (with Nd adjusted) for these two elements; if the total S(v) shifts by more than about 10% at v<0.5, the integer count is not robustly determined by the plasmon frequency.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim rests on Eq. (7): the d-electron momentum distribution in the solid is taken from free-atom Hartree-Fock or relativistic wave functions, with the number Nd of bound d-electrons fixed by the integer NFEG inferred from the experimental plasmon frequency (Table I). This assumption is load-bearing because Eq. (5) computes the low-velocity stopping S(v) as an integral over f_nd(p) weighted by the transport cross section; at v<1 the integrand samples the distribution at momenta p≈v, precisely the low-momentum tail of the atomic distribution. In a transition metal, the occupied d-like states are Bloch states, and s-d hybridization and d-band dispersion add low-momentum weight that is absent in an isolated-atom distribution; the d-band occupation is not a rescaled atomic subshell. If this low-momentum weight is significant, the d-electron contribution at low velocities, and therefore the claimed agreement with experiment and TDDFT, would change. The authors' own caveat about a possible band-gap overestimation at the lowest velocities (Sec. III A) is a symptom of the same atomic-versus-solid gap, but the shape of f_nd(p) itself is the more fundamental issue. The electron-count assignment (NFEG, Nd) is a coupled uncertainty: integer rounding of the plasmon frequency changes Nd (e.g., Cu 3 vs 1, Au 7 vs 11) and directly rescales the d contribution. Neither the atomic-shape assumption nor the count assignment is tested against a solid-state calculation in the paper.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes a non-perturbative model for the d-electron contribution to the electronic stopping cross section of the late transition metals Ni, Pd, Pt, Cu, Ag, and Au. The stopping integral of Eq. (1) is evaluated with an atomic \"inhomogeneous\" momentum distribution f_nl(p) built from Hartree-Fock or relativistic wave functions (Eq. (7)), and this d contribution is combined with a free-electron-gas description of the valence electrons and a shellwise local approximation for inner shells. The authors compare their low-velocity results with experimental data from the IAEA database and with available TDDFT calculations, and present extended-energy curves from 0.1 keV to 100 MeV. The central claim is that the d-electron stopping response can be captured by atomic momentum distributions rather than by a free-electron gas or full TDDFT, and that this yields good agreement with experiment and with TDDFT where available.","tokens_in":13091,"tokens_out":16713,"duration_ms":162886,"significance":"If validated, the model is attractive because it is simple and analytic, it requires no fitting to stopping data, and the electron-count inputs (N_FEG, N_d) are inferred from experimental plasmon frequencies. The paper's strengths include a transparent derivation, extensive comparison with a large experimental database, and direct comparison with TDDFT results for several targets. The main obstacle is that the central assumption — that free-atom d-orbital momentum distributions represent the solid-state d-electron response — is not tested against any solid-state calculation, and the sensitivity of the results to the N_FEG/N_d partition is not quantified. These issues are addressable, but they are load-bearing for the claimed agreement at low velocities.","major_comments":[{"comment":"The central assumption is that the d-electron response in the solid is described by the free-atom momentum distribution f_nl(p). In the low-velocity limit, Eqs. (5)-(6) weight the distribution at p ~ v, so the low-energy stopping power is controlled by the low-momentum tail of f_nd(p). In a transition metal this tail is affected by band structure, s-d hybridization, and d-band occupation, none of which is captured by an isolated-atom orbital. The authors' caveat at the end of Sec. III A that a band-gap effect may cause overestimation at the lowest velocities is a symptom of this atomic-versus-solid gap, but it does not test the shape of f_nd(p) itself. I request a quantitative comparison with a DFT-based d-projected momentum density (or equivalent solid-state calculation) for at least one Group-10 and one Group-11 target, with a discussion of how hybridization changes the low-p weight and how that would affect the computed stopping curves.","section":"Sec. II, Eq. (7); Sec. III A"},{"comment":"The partition of outer electrons into N_FEG and N_d is based on integer rounding of the experimental plasmon frequency through the free-electron relation, together with the free-atom configuration; it is not validated by a solid-state occupation calculation. Because the d contribution is proportional to N_d through the normalization in Eq. (7), the low-velocity results depend directly on this assignment. The authors state that N_FEG and N_d 'may be the subject of discussion' and compare with an N_FEG=10 variant for Pt in Figs. 6 and 12, but this is not a systematic sensitivity analysis for the present model. I would like to see either a quantitative sensitivity study over the admissible integer choices of N_FEG/N_d for all six targets, or an independent estimate of the solid d-occupation from DFT projected densities of states.","section":"Table I; Sec. III"},{"comment":"As printed, the integration limits in I'(v_r), g5, and g7 are reversed. After exchanging the order of integration in Eqs. (3)-(4), the p-integral should run from |v_r - v| to v_r + v. The printed limits |v_r + v| to |v_r - v| decay over the integration interval and would give the negative of the intended value. This is likely typographical, but it appears in the central derivation and in the analytical expressions of Appendix A, so it must be corrected and the numerical implementation must be confirmed to use the correct limits.","section":"Eqs. (6), (A11)-(A13)"}],"minor_comments":[{"comment":"The Heaviside function is written as Θ(p - p_F), but the text says the step function lies 'within the Fermi sphere'; this should be Θ(p_F - p). In addition, the numerical constant (2π)^{3/2} on the right-hand side is inconsistent with the Fermi-sphere volume factor: with p_F = (3π^2 n_e)^{1/3}, one has ∫ d^3p = 4π^3 n_e, not (2π)^{3/2} n_e. Please clarify the normalization convention and reconcile Eq. (9) with Eq. (8).","section":"Eq. (9)"},{"comment":"The phrase 'the dump value γ_p^exp' appears to be a typo for 'damping value'; the text around Table I should be corrected.","section":"Table I caption"},{"comment":"The comparison with TDDFT for Au is complicated by the fact that the cited TDDFT results are channelling calculations for Au<100>; the authors note this, but the figure caption could make the distinction clearer to avoid the impression of a direct off-channelling comparison.","section":"Sec. III A, Figs. 2-7"}],"recommendation":"major_revision","confidential_remarks":null},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThis paper does something genuinely new: it computes the d-electron contribution to proton stopping in six transition metals using the inhomogeneous momentum distribution of atomic d-orbitals, rather than the usual free-electron-gas sphere. The kinetic-theory framework is standard, but the input distribution is not, and the analytical machinery in the appendix for obtaining f_nl(p) and folding it into the stopping integral is a real step forward. For Ni, Pd, and Pt the model's agreement with recent experiments and with TDDFT is good enough to take seriously. The combined FEG + d + inner-shell description spanning 0.1 keV to 100 MeV is practically useful for ion-beam analysis.\n\nThe soft spots, in order of importance. First, the load-bearing assumption is that free-atom Hartree-Fock/relativistic wave functions represent the solid's bound d-electrons. The stress-test worry about the low-momentum tail is legitimate: in the solid, d-like Bloch states hybridize with s/p states and acquire low-momentum weight that an isolated atom lacks, and Eq. (5) samples exactly p≈v at low velocity. The authors' own caveat about a possible overestimation at lowest velocities from the band gap (Sec. III A) is related, but a DFT calculation of the d-projected momentum density would be the direct test. Second, the electron counts NFEG and Nd are inferred from experimental plasmon frequencies and rounded to integers; the paper does not quantify how sensitive the stopping curves are to that rounding, and a sensitivity study would be cheap and useful. Third, reproduction is harder than it should be: the Slater expansion coefficients for the six atoms are not tabulated and no code is provided. Fourth, the group 11 comparisons are genuinely ambiguous because the low-energy data themselves split into two clusters; the model lands between them, which is honest but not clean validation.\n\nNone of these are fatal. The model is a plausible first-principles-based approximation, the comparisons are with the best available data and TDDFT, and the authors are transparent about the open questions. A serious referee should engage with it, and the required fixes are well-defined. I would recommend sending this to peer review, with the expectation that the referee asks for the DFT momentum-density test and a sensitivity analysis of the electron-count rounding.","headline":"A genuinely new d-electron stopping model for transition metals; the atomic-to-solid momentum distribution assumption is the key thing a referee should test.","tokens_in":13640,"tokens_out":6551,"would_cite":true,"duration_ms":64493,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["34.50.Bw"],"model":"deepseek-v4-flash","headline":"The low-energy stopping of protons in six transition metals is reproduced by the atomic d-subshell momentum distribution, not by a free-electron gas.","keywords":["stopping power","transition metals","d-electrons","inhomogeneous momentum distribution","non-perturbative model","low-energy ion stopping","plasmon frequency","electronic stopping cross section"],"falsifier":"Measure proton stopping in a single crystal of Ni or Au along a well-characterized off-channelling direction at velocities below about 0.3 atomic units with a few-percent uncertainty and compare with the model curve: a deviation larger than the stated few-percent agreement, or the appearance of a sharp slope break that the model excludes, would falsify the free-atom d-electron description. A less costly test is to recompute $f_{3d}(p)$ using a band-structure d-projected wavefunction and check whether the low-velocity total stopping shifts by more than the model's current scatter.","tokens_in":12599,"feed_emoji":"⚛️","tokens_out":21323,"duration_ms":166937,"temperature":0.7,"pith_summary":"This paper proposes a way to compute the d-electron contribution to the stopping power of the later transition metals (Ni, Pd, Pt, Cu, Ag, and Au) without perturbation theory: treat the partially filled d-subshell as a gas of electrons with the momentum distribution of the free atom, $f_{nl}(p)=(2\\pi)^{3/2}|\\Phi_{nl}(p)|^2$, rather than as part of the homogeneous free-electron gas. Only the number of electrons actually promoted to the conduction band is taken from the measured plasmon frequency; the rest stay bound and respond to the ion through an effective collision cross section built from a velocity-dependent screened potential. The model is claimed to reproduce the low-energy stopping cross sections with good agreement with experiment and with time-dependent density functional theory (TDDFT), and, when combined with free-electron-gas and inner-shell contributions, to give coherent total stopping cross sections from 0.1 keV to 100 MeV. The d-electron contribution appears as a smooth nonlinearity in the velocity dependence, with a maximum near the mean d-electron velocity, rather than as the sharp break of linearity discussed in older experiments. The authors do not rule out a small overestimation at the lowest velocities because the model contains no explicit band gap.","feed_headline":"d-electron momentum distributions match ion stopping in six metals","feed_subtitle":"Uses only atomic wavefunctions plus the measured plasmon frequency to match stopping from 0.1 keV to 100 MeV.","key_machinery":"The load-bearing object is the inhomogeneous momentum distribution $f_{nl}(p)=(2\\pi)^{3/2}|\\Phi_{nl}(p)|^2$: the squared Fourier transform of the atomic d-subshell wavefunction, normalized to the number $N_d$ of bound d electrons. It replaces the Fermi-sphere step function of the free-electron-gas model, so d-electrons of all momenta contribute, and it is what carries the low-energy d-electron response. The stopping cross section is then computed by folding this distribution with the transport cross section $\\sigma_{\\mathrm{tr}}(v_r)$ (the momentum-transfer cross section) generated by a velocity-dependent screened potential that satisfies the cusp condition; the Fourier transforms are evaluated analytically after expanding the wavefunctions in exponential-type basis functions. The model becomes a full stopping theory by adding free-electron-gas and inner-shell contributions, covering 0.1 keV to 100 MeV.","core_discovery":"The paper's central discovery is that the stopping of protons in the later transition metals is not governed by how the d-electrons sit in the Fermi sphere but by their full atomic momentum profile. Concretely, the d-subshell stopping cross section is obtained by replacing the Fermi step function with $f_{nl}(p) = (2\\pi)^{3/2}|\\Phi_{nl}(p)|^2$ in the transport-cross-section integral, with $\\Phi_{nl}(p)$ the Fourier transform of the atomic d wavefunction normalized to the number $N_d$ of d electrons that remain bound. The integer $N_{\\mathrm{FEG}}$ of electrons in the free-electron gas is taken as the integer closest to the value implied by the measured plasmon frequency, so the only solid-state input is that frequency. Adding the resulting d-curve to a free-electron-gas curve and to inner-shell curves yields total stopping cross sections that match recent low-energy experiments closely for Ni, Pd, and Pt, lie between the conflicting data sets for Cu, Ag, and Au, and agree with available TDDFT results. The authors interpret the low-energy 'break of linearity' seen in group 11 measurements as a soft nonlinearity generated by the inhomogeneous momentum distribution, with the d-curve peaking near the mean d-electron velocity $v_d$.","pith_inferences":["Beyond the paper: the same inhomogeneous-momentum construction could be tested on other d-block metals (for example, group 9 or group 12), where low-energy stopping data are sparser; the only inputs needed are a d-subshell wavefunction and a measured or computed plasmon frequency.","Beyond the paper: replacing the free-atom d wavefunctions in Eq. (7) with solid-state d-projected wavefunctions would isolate the size of band-structure and band-gap effects the authors identify as a possible source of low-velocity overestimation.","Beyond the paper: a systematic comparison of d-only stopping curves (not total stopping) between this model and TDDFT across all six metals would test whether the atomic momentum profile captures the same physics as explicit time-dependent electron dynamics."],"forward_implications":["For Ni, Pd, and Pt the calculated total stopping cross sections fall within a few percent of recent low-energy measurements, so the model provides a quantitative account of the group 10 low-energy response without a slope break.","For Cu, Ag, and Au the calculated curves sit between the two conflicting families of low-energy data, clarifying that the historical 'break of linearity' is not reproduced as a sharp kink but as a smooth d-electron-driven rise.","The d-subshell contribution to the stopping cross section is largest when the projectile speed is close to the mean d-electron speed $v_d$; this places the d-curve maximum at a distinct impact energy for each metal (about 350 keV/amu for Ni, about 100 keV for Pt, with the other metals in between).","Combining the d-model with free-electron-gas and inner-shell models yields a fully theoretical stopping-power curve over 0.1 keV to 100 MeV, with the only electron-count parameter fixed by the measured plasmon frequency."],"supporting_citations":[{"why":"Supplies the velocity-dependent screened potential (with cusp condition) used to build the transport cross-section in Eq. (2).","marker":"[4]"},{"why":"Provide the transport-cross-section formalism and the momentum-integral expression for the stopping cross-section in Eq. (1).","marker":"[33, 34]"},{"why":"Atomic wavefunctions for Ni and Cu from which the 3d momentum distributions are built.","marker":"[35]"},{"why":"Relativistic atomic-structure calculations for Pd, Ag, Pt, and Au from which the 4d and 5d momentum distributions are built.","marker":"[36, 37]"},{"why":"Analytic integral identities used to evaluate the Fourier transforms of the exponential-expanded wavefunctions.","marker":"[39]"},{"why":"Experimental plasmon frequencies that determine the integer FEG electron number and hence the number of bound d electrons.","marker":"[40]"},{"why":"Experimental stopping cross-section data against which every presented curve is compared.","marker":"[28, 29]"},{"why":"Dielectric free-electron-gas stopping model used for the valence contribution above the plasmon threshold.","marker":"[30]"},{"why":"Inner-shell stopping model used to include the deepest subshells in the extended-energy curves.","marker":"[31, 32]"},{"why":"TDDFT off-channelling stopping results used as the ab initio comparison at low and intermediate energies.","marker":"[22–25]"}],"fun_headline_variants":["Atomic d-wavefunctions explain low-energy ion stopping in metals","Full atomic momentum profile predicts stopping power in six metals","d-electron momentum shape drives stopping from keV to MeV","Inhomogeneous d-momentum explains metal stopping power break","Atomic d-state momentum captures low-energy stopping in metals"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the d electrons that stay bound in the solid respond like free-atom d electrons, with the number of such electrons fixed by rounding the conduction-electron count from the measured plasmon frequency to an integer; if solid-state band effects or that electron-count assignment are wrong, the low-energy d contribution changes.","fun_headline_variants_meta":{"raw":{"variants":["Atomic d-wavefunctions explain low-energy ion stopping in metals","Full atomic momentum profile predicts stopping power in six metals","d-electron momentum shape drives stopping from keV to MeV","Inhomogeneous d-momentum explains metal stopping power break","Atomic d-state momentum captures low-energy stopping in metals"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00038,"raw_usage":{"total_tokens":2045,"prompt_tokens":1001,"completion_tokens":1044,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":617,"completion_tokens_details":{"reasoning_tokens":965}},"tokens_in":617,"tokens_out":1044,"duration_ms":7988,"temperature":1.0,"reasoning_tokens":965,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T17:10:21.276467+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure proton stopping in a single crystal of Ni or Au along a well-characterized off-channelling direction at velocities below about 0.3 atomic units with a few-percent uncertainty and compare with the model curve: a deviation larger than the stated few-percent agreement, or the appearance of a sharp slope break that the model excludes, would falsify the free-atom d-electron description. A less costly test is to recompute $f_{3d}(p)$ using a band-structure d-projected wavefunction and check whether the low-velocity total stopping shifts by more than the model's current scatter.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the velocity-dependent screened potential (with cusp condition) used to build the transport cross-section in Eq. (2)."},{"cited_title":"Nagy and A","cited_arxiv_id":null,"evidence_quote":"Atomic wavefunctions for Ni and Cu from which the 3d momentum distributions are built."},{"cited_title":"Bar-Shalom, M","cited_arxiv_id":null,"evidence_quote":"Analytic integral identities used to evaluate the Fourier transforms of the exponential-expanded wavefunctions."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Experimental plasmon frequencies that determine the integer FEG electron number and hence the number of bound d electrons."},{"cited_title":"Electronic Stop- ping Power of Matter for Ions Graphs, Data, Comments and Programs,","cited_arxiv_id":null,"evidence_quote":"Dielectric free-electron-gas stopping model used for the valence contribution above the plasmon threshold."}],"review_version":1}