{"id":"10213a66-e894-4a53-b89f-a592cff25663","arxiv_id":"2411.11021","paper_version":1,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"A kink in the trial nucleon density at the center of cylindrical or plane-parallel Wigner-Seitz cells makes the fourth-order extended Thomas-Fermi energy divergent; only spherical-cell kinks are harmless.","lead":"This paper shows that numerical calculations of dense neutron star matter with the extended Thomas-Fermi method can blow up if the trial density profiles are not smooth enough. It derives the exact smoothness condition and a bound on the accuracy of such calculations.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"No significant objection identified.","rationale":"The paper's central claim is a mathematical statement about the fourth-order ETF functional: if one uses this functional, kinked trial densities produce divergent energy integrals. The proof is built on the standard expression for τ^(4)_{q,v} (Eq. 11), and the divergence is identified in the (∇²n)²/n^{5/3} term. The analysis correctly separates the volume and surface contributions, showing that the surface term on an artificial boundary is finite and cannot cancel the volume divergence. The spherical case is treated carefully and found to be integrable, which is consistent with the local behavior (∇²n ~ 1/r, volume element r² dr). The Sec. 5 lower bound provides an independent, regularization-based confirmation for the plane-parallel case, and the numerical example is appropriately limited to the D=1 (lasagna) geometry. The reader's weakest assumption—that the derivation relies on the fourth-order ETF functional being an adequate description—is a fair external-validity caveat, but it does not threaten the internal correctness of the claim. The paper explicitly acknowledges the dependence on the employed nuclear functional and the approximation M* = M, so it does not overclaim. I therefore find no load-bearing flaw and agree with the reader's ACCEPT verdict.","tokens_in":14699,"tokens_out":30270,"duration_ms":361136,"concrete_test":"Numerically integrate the fourth-order ETF kinetic energy density τ^(4)_{q,v} for a slab trial profile n(z) = n0 + A|z| on a uniform mesh with spacing h, representing the cusp by one-sided derivatives at the center. Verify that the integral grows as 1/h as h→0. Repeat for a cylindrical profile n(r) = n0 + A r and verify logarithmic growth ln(R/h). If these scalings are reproduced, the divergence claim is confirmed numerically.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim—that a kink in a trial density at the center of a cylindrical or plane-parallel WS cell, or on any smooth surface, makes the fourth-order ETF gradient correction divergent—is internally sound. The divergence follows directly from the (∇²n)²/n^{5/3} term in the volume part of the fourth-order kinetic energy density (Eq. 11). For cylindrical symmetry, the Laplacian of n0 + n′0 r⊥ is n′0/r⊥, so the integrand behaves as 1/r⊥ and the integral diverges logarithmically (Eq. 27). For plane-parallel symmetry, the Laplacian contains a delta function whose square is non-integrable (Eq. 29). The argument that surface terms cannot cancel the divergence is valid because one can isolate the kink by a smooth artificial surface; the surface integral on that surface is finite while the volume integral diverges. The Sec. 5 lower bound independently confirms a 1/ε divergence for the slab case. The only meaningful caveat is external: the fourth-order gradient expansion may not adequately describe the pasta regime, so the divergence is a property of the approximate functional rather than the physical system. The paper explicitly limits its numerical example to a specific functional and does not overclaim physical relevance. No internal inconsistency or missing step in the proof was found.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper examines smoothness conditions for parametrized trial nucleon density distributions used in restricted variational calculations with the fourth-order extended Thomas-Fermi (ETF) functional, with emphasis on Wigner-Seitz cells of spherical, cylindrical, and slab symmetry in neutron-star pasta phases. The central result is that a kink (discontinuous first derivative) of the trial density at the center of a cylindrical or plane-parallel cell, or across any smooth interior surface, makes the fourth-order gradient correction divergent: the term (∇²n)²/n^(5/3) in τ^(4)_q,v is logarithmically divergent in the cylindrical case and non-integrable (delta-squared) in the slab case. A kink at the center of a spherical cell is shown to give only an integrable contribution. The paper also derives a lower bound for the contribution of a locally smoothed kink on a numerical mesh, showing that the bound grows at least as 1/ε in the slab case, and it applies this estimate to the BSk24 damped-Fermi parametrization to argue that previously reported pasta-phase results are not invalidated.","tokens_in":14962,"tokens_out":13396,"duration_ms":158084,"significance":"If accepted, the result is practically important for ETF calculations of neutron-star crust and pasta: it establishes a rigorous, easy-to-check smoothness constraint on trial functions and explains why certain widely used parametrizations (e.g., simple Fermi or modified Fermi forms) can give divergent fourth-order corrections, while damped-Fermi forms with vanishing gradient at the cell center are safe. The derivation is transparent and self-contained: the spherical, cylindrical, and plane-parallel cases are treated explicitly, and the Sec. 5 lower bound provides an independent regularized confirmation of the divergence. The numerical estimates are carefully hedged: the authors state that the example is specific to the BSk24 functional and that the M*=M approximation makes the numbers order-of-magnitude only. The paper does not overclaim physical relevance beyond the approximate ETF functional it analyzes.","major_comments":[],"minor_comments":[{"comment":"The sentence stating that the minimal contribution of the smoothed-out central kink 'does not exceed 0.1 keV' is logically inconsistent with Eq. (37), which is a lower bound: a lower bound cannot establish an upper bound on the actual contribution. Please rephrase to say that the lower-bound estimate remains below 0.1 keV for ε ≳ 10^-3 fm, and, if desired, add an explicit estimate of the actual smoothed contribution for the mesh used in Ref. [74].","section":"Sec. 6"},{"comment":"There are several typographical errors, including 'Saint Peter sburg' in the affiliation and 'diverge s' and 't o add' in the Introduction; a careful proofreading pass is needed.","section":"Title page and Introduction"},{"comment":"The surface integrand τ^(4)_q,s is not written explicitly. Since the later argument that surface terms are finite and cannot cancel the volume divergence depends on this quantity, an explicit expression or a precise reference for it would make the proof easier to verify.","section":"Sec. 2, Eqs. (9)-(10)"},{"comment":"The symbol rendered as '/greaterorsimilar' should be typeset as '≳' and its meaning ('greater than or approximately equal to, up to a factor 1+O(ε)') should be stated in the text.","section":"Sec. 5, Eqs. (36)-(37)"},{"comment":"The use of the squared Dirac delta function δ(z)^2 is formal; a parenthetical note that the argument is made rigorous by the regularized smoothing construction in Sec. 5 would improve mathematical clarity.","section":"Sec. 4.2, Eq. (29)"},{"comment":"When discussing the simple and modified Fermi parametrizations, the paper notes that they do not satisfy condition (12); it would also be useful to state explicitly that these same parametrizations have a kink at the cell center, which is the relevant issue for cylindrical and plane-parallel cells.","section":"Sec. 3"}],"recommendation":"minor_revision","confidential_remarks":"This is a well-scoped and technically sound paper. The central divergence result is correct and should be useful to practitioners of ETF calculations. The only substantive issue I found is the lower-bound/upper-bound wording in Sec. 6, which is local and easily fixed; the remaining points are presentation. I recommend minor revision."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The useful new thing here is the explicit geometry-dependent divergence analysis. A kink in the trial density at the center of a cylindrical or plane-parallel WS cell makes the fourth-order ETF gradient correction diverge (logarithmically for cylinders, via a squared delta function for slabs), while the same kink in a spherical cell is integrable. The authors also give a lower-bound estimate, Eq. (37), showing that smoothing a kink over a mesh scale ε injects at least ~1/ε of kinetic energy per nucleon. That is a concrete, reproducible criterion for practitioners: if your parametrization has a central cusp and you use a fine mesh, you are not converging to a finite ETF energy. That is the main point, and it is proved cleanly. The stress-test note is right that the derivation is internally sound; I checked the cylindrical log divergence and the slab delta-squared argument against Eqs. (24)–(29) and found no gap. The Sec. 5 bound is a legitimate addition, not just a restatement of the divergence — it tells you how badly a smoothed kink hurts as ε shrinks.\n\nThe paper is also honest about its own limits. The numerical example uses the BSk24 functional and M* = M, so the 10^-3 fm estimate is order-of-magnitude only. More importantly, the authors state up front that previous phase-diagram results are not invalidated because the divergence sits below their actual accuracy. That is the right scope: this is a methodological caution for future calculations, not a revision of existing pasta phase boundaries.\n\nSoft spots: the divergence is a property of the fourth-order gradient expansion, not of the physical system. If the semiclassical expansion itself is suspect in the pasta regime, the practical force of the warning weakens. The authors do not engage with that possibility beyond noting they use a concrete functional. That is a mild limitation, not a flaw — they never claim physical relevance beyond the ETF framework. The lasagna/anti-lasagna symmetry discussion (Eqs. (20)–(22)) is useful but somewhat orthogonal to the divergence proof; it reads as a secondary point. Also, the paper leans on the boundary condition (12) to drop the surface term; they argue correctly that the volume divergence cannot be cancelled by a finite surface term, so that is not a real weakness.\n\nThis is a solid, narrow methods paper. It earns its acceptance. The target audience is anyone doing restricted variational ETF calculations of neutron star crust or pasta — they should read this before choosing a parametrization. I would cite it if I worked in that area. Send it to a referee with nuclear structure or ETF expertise; it will survive scrutiny and may prompt a useful discussion of whether the fourth-order functional itself is adequate in pasta. Serious referee: yes. Reading group: worth a slot if anyone in the group does ETF or pasta work.","headline":"A clean, self-contained proof that fourth-order ETF calculations need trial densities with continuous first derivatives, with the pasta-phase impact honestly bounded; worth refereeing.","tokens_in":15457,"tokens_out":698,"would_cite":true,"duration_ms":10841,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["26.60.-c","26.60.Gj","31.15.bt","64.10.+h"],"model":"deepseek-v4-flash","headline":"Kinks—discontinuities in the first derivative of a parametrized nucleon density—make the fourth-order extended Thomas-Fermi gradient correction diverge in cylindrical and slab geometries, so restricted variational trial functions must…","keywords":["extended Thomas-Fermi method","gradient corrections","nuclear pasta","Wigner-Seitz cell","trial density parametrization","smoothness condition","neutron star crust","density functional theory"],"falsifier":"Take a slab-symmetric trial density $n(z)=n_0+n_1|z|$ in a small interval $|z|<\\varepsilon$, matched continuously to an outer profile, and evaluate the fourth-order ETF kinetic energy (Eqs. 5-11) as $\\varepsilon\\to 0$ with the outer profile fixed. If the energy does not grow at least as $1/\\varepsilon$—or if, on a uniform mesh, the computed energy does not increase without bound as the mesh spacing goes to zero—the claimed divergence is absent.","tokens_in":14562,"feed_emoji":"⚛️","tokens_out":10118,"duration_ms":97954,"temperature":0.7,"pith_summary":"Many calculations of nuclear pasta in neutron-star crusts use the extended Thomas-Fermi method, an approximation that adds gradient corrections to the statistical Thomas-Fermi model of nucleons, with parametrized density profiles inside Wigner-Seitz cells. This paper shows that if such a trial profile has a kink—a jump in its first derivative—at the center of a cylindrical or slab-shaped Wigner-Seitz cell, or across any smooth surface such as a cell boundary, the fourth-order gradient correction to the kinetic energy diverges. A kink at the center of a spherical cell is harmless because the divergence is integrable. The practical conclusion is that restricted variational trial functions must have continuous first derivatives everywhere, and that discrete-mesh calculations that smooth kinks implicitly face a fundamental accuracy limit. If correct, this identifies and resolves a subtle failure mode in a widely used approximation for dense-matter equations of state.","feed_headline":"Kinked trial densities make fourth-order Thomas-Fermi energy diverge","feed_subtitle":"Nuclear-pasta trial densities need continuous first derivatives, or the fourth-order gradient correction blows up.","key_machinery":"The load-bearing object is the volume part of the fourth-order kinetic-energy correction, $\\tau^{(4)}_{q,v}$ from Eq. (11), and specifically its first term $\\tau^{(4a)}_q \\propto n_q^{1/3}(\\nabla^2 n_q/n_q)^2$. This term is singular precisely where the density slope changes discontinuously: the Laplacian of a kinked profile contains a delta-like contribution in the normal direction, and the square of that contribution is not integrable unless the slope jump is zero. The argument also uses the Ostrogradsky-Gauss reduction that replaces the original fourth-order expression (Eq. 7) by Eq. (11) plus a surface term, and the elementary inequality $\\langle x^2\\rangle \\ge \\langle x\\rangle^2$ to convert the divergence into a lower bound on the energy of any locally smoothed kink (Eq. 36).","core_discovery":"The paper's central claim is that the fourth-order extended Thomas-Fermi kinetic-energy functional is finite only for trial nucleon densities whose gradient is continuous everywhere. In cylindrical (spaghetti) geometry the Laplacian of a profile with nonzero central slope behaves like $n'_{q0}/r$ near the axis, and the term proportional to $(\\nabla^2 n/n)^2$ makes the integral $\\int dr/r$ diverge. In plane-parallel (lasagna) geometry the kink produces a Dirac delta in $\\nabla^2 n$, and the integral of the squared delta function is not defined; for a kink on any smooth surface, the tangent-plane argument gives the same nonintegrable squared delta. The only escape is to require the normal derivative of $n_q$ to be continuous at every smooth surface and, because of symmetry, zero at the center of cylindrical and slab cells. A kink at the center of a spherical cell does not diverge, because the volume element $r^2 dr$ tames the singularity.","pith_inferences":["Editorial: The proof only uses the algebraic structure of the fourth-order gradient term, so the same $C^1$ requirement should apply to any density functional containing $n^{1/3}(\\nabla^2 n/n)^2$; a generic coefficient could be substituted in the $\\varepsilon\\to 0$ estimate to test this.","Editorial: The geometry dependence suggests a simple dimensional diagnostic: a kink at an isolated point is integrable in three dimensions but not in one or two, because the volume element near the singularity is $r^{D-1}dr$; this could guide the design of trial profiles for other cell shapes.","Editorial: One practical extension is to add an explicit $C^1$ core (for instance a quadratic dependence on $\\xi$ near the center) to existing parametrizations and compare the minimized energies; agreement would confirm that the smoothness condition is not restrictive in practice, while disagreement would call for revisiting earlier restricted-variational results.","Editorial: The lower bound of Eq. (37) can be repurposed as a mesh-selection criterion: before trusting an energy difference between pasta phases, check that the smoothed-kink bound is below the target precision of the calculation."],"forward_implications":["Woods-Saxon and modified-Fermi parametrizations, which have a nonzero slope at the cell center, are not admissible for fourth-order ETF calculations in cylindrical and slab pasta phases unless the kink is treated as a distribution; in spherical geometry they remain integrable.","For cylindrical and slab cells the trial density must satisfy $d n_q/d\\xi \\to 0$ at the cell center, and at any smooth interior surface or true polyhedral cell boundary the normal derivative must be continuous.","Enforcing the lasagna/anti-lasagna inversion symmetry of Eq. (21) automatically gives a zero gradient at the slab center and removes a source of spurious numerical minima.","Discrete-mesh codes that implicitly smooth kinks over a length $\\varepsilon$ cannot refine the mesh arbitrarily: the smoothed-kink contribution grows at least as $\\varepsilon^{-1}$ for slabs, so mesh refinement near a kink drives the energy upward without bound.","For the BSk24 calculations examined in the paper, this accuracy limit lies below 0.1 keV per nucleon for $\\varepsilon \\gtrsim 10^{-3}$ fm, so the previously reported pasta-phase energies are not invalidated."],"supporting_citations":[{"why":"Derived the fourth-order gradient correction and showed the surface term can be removed by partial integration; this supplies the functional form used in the divergence analysis.","marker":"[10]"},{"why":"Canonical derivation of the extended Thomas-Fermi expansion and the whole-space integration that justifies dropping the surface term; the paper's second-order and fourth-order expressions rest on it.","marker":"[12]"},{"why":"Provided the explicit volume form $\\tau^{(4)}_{q,v}$ and stressed that the fourth-order expressions require vanishing density gradients at the boundary; this is the functional being tested.","marker":"[49]"},{"why":"The discrete-mesh ETF calculation with the damped-Fermi parametrization that provides the numerical example and the estimated accuracy limit for lasagna phases.","marker":"[74]"},{"why":"Compared parametrizations and introduced the two-sided damping and soft-damping forms; used here to discuss symmetry and smoothness of trial profiles.","marker":"[75]"},{"why":"Source of the Oyamatsu parametrization (15), whose sharp cut-off profile illustrates the kind of non-smooth trial function the paper analyzes.","marker":"[59]"},{"why":"Introduced the damped-Fermi parametrization (19) used in practical pasta calculations and in the paper's numerical estimate.","marker":"[22]"},{"why":"Showed that fourth-order ETF corrections reproduce nuclear binding energies without empirical adjustment, establishing why the fourth-order term matters.","marker":"[28]"}],"fun_headline_variants":["Kinked densities blow up Thomas-Fermi energy","Fourth-order Thomas-Fermi diverges on kinked trial densities","Smoothness condition for Thomas-Fermi nuclear pasta densities","Trial density kinks ruin extended Thomas-Fermi accuracy"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The claim rests on the fourth-order gradient-expanded kinetic-energy density, with the surface term dropped and coefficients set by Eqs. (5)-(11), being the correct energy functional for nuclear pasta; if that semiclassical expansion is not accurate in the pasta regime, the kink divergence is an artifact of the approximation rather than a property of the physical energy.","fun_headline_variants_meta":{"raw":{"variants":["Kinked densities blow up Thomas-Fermi energy","Fourth-order Thomas-Fermi diverges on kinked trial densities","Smoothness condition for Thomas-Fermi nuclear pasta densities","Trial density kinks ruin extended Thomas-Fermi accuracy"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000948,"raw_usage":{"total_tokens":3979,"prompt_tokens":809,"completion_tokens":3170,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":425,"completion_tokens_details":{"reasoning_tokens":3100}},"tokens_in":425,"tokens_out":3170,"duration_ms":27236,"temperature":1.0,"reasoning_tokens":3100,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T19:01:16.642420+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take a slab-symmetric trial density $n(z)=n_0+n_1|z|$ in a small interval $|z|<\\varepsilon$, matched continuously to an outer profile, and evaluate the fourth-order ETF kinetic energy (Eqs. 5-11) as $\\varepsilon\\to 0$ with the outer profile fixed. If the energy does not grow at least as $1/\\varepsilon$—or if, on a uniform mesh, the computed energy does not increase without bound as the mesh spacing goes to zero—the claimed divergence is absent.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Derived the fourth-order gradient correction and showed the surface term can be removed by partial integration; this supplies the functional form used in the divergence analysis."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Canonical derivation of the extended Thomas-Fermi expansion and the whole-space integration that justifies dropping the surface term; the paper's second-order and fourth-order expressions rest on it."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provided the explicit volume form $\\tau^{(4)}_{q,v}$ and stressed that the fourth-order expressions require vanishing density gradients at the boundary; this is the functional being tested."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Source of the Oyamatsu parametrization (15), whose sharp cut-off profile illustrates the kind of non-smooth trial function the paper analyzes."},{"cited_title":"Semi-classical equation of state and specific heats for neutron-star inner crust with proton shell corrections","cited_arxiv_id":"0806.0296","evidence_quote":"Introduced the damped-Fermi parametrization (19) used in practical pasta calculations and in the paper's numerical estimate."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Showed that fourth-order ETF corrections reproduce nuclear binding energies without empirical adjustment, establishing why the fourth-order term matters."}],"review_version":1}