{"id":"c45000b7-4b2e-405f-8943-df587f4c7f53","arxiv_id":"1908.06017","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The BCS gap equations are claimed to have two equilibrium solutions at all temperatures below Tc, including a new small-gap solution in the symmetric textbook limit.","lead":"This paper studies the BCS equations for superconductors and claims they admit two different equilibrium solutions, even in the standard symmetric case. The second solution has a smaller energy gap and a nonzero population of excitations down to zero temperature.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The second solution at µ_R=µ is obtained only through a singular limit that does not solve the δ=0 BCS equations; without an independent derivation of Eq. (4c), the central claim is unsupported.","rationale":"The reader's weakest_assumption correctly identifies Eq. (4c) as the load-bearing premise. My stress-test sharpens this: the second branch at µ_R=µ is not a solution of the δ=0 equations because the F-dependent term is absent there, and the paper replaces the undefined ratio (5b) with a new constraint Fd=0 plus a finite M. This is a singular-limit construction, not a derivation from the symmetric BCS free energy. The T=0 ∆0/3 result is a genuine algebraic consistency check of the author's modified equations, which deserves credit, but it does not establish that the standard BCS problem has two solutions. The proposed test—an independent derivation of Eq. (4c) or a literal δ=0 numerical solve—would settle the issue. Since that test has not been performed, the reader's CONDITIONAL verdict is appropriate; if the test fails, the verdict should move to REJECT. I therefore leave the verdict unchanged.","tokens_in":9445,"tokens_out":23869,"duration_ms":253186,"concrete_test":"Re-derive Eq. (11) from Eq. (10) by differentiating the explicit E−µ_RN in the Bogoliubov basis, with ∆ eliminated through Eq. (5a). If the result at δ=0 is nξ=f(βϵξ) with no F term, then Eq. (4c) is not the BCS equilibrium condition; solving (5a) with nξ=f(βϵξ) then gives only the standard BCS root for all T. As a numerical cross-check, continue the second branch in δ from δ/∆0=10^-1 down to 10^-6 and test whether the limiting M satisfies the δ=0 stationarity condition of Eq. (10), rather than an imposed Fd=0 constraint; if the δ=0 system has no such root, the symmetric-limit claim fails.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim requires M=(µ_R−µ)F to remain nonzero as δ=µ_R−µ→0, with F diverging so that the denominator Fd in Eq. (5b) vanishes. But at δ=0, Eq. (5c) contains no correction at all: nξ=f(βϵξ), and the standard uniqueness results cited by the author [40,41] apply. The second branch instead uses occupations (8), nξ=f(β(ϵξ+M/ϵξ)), with M fixed by the extra constraint Fd=0 introduced after Eq. (9). That constraint is not part of the δ=0 form of Eqs. (5); Eq. (5b) is a 0/0 indeterminate at δ=0, and replacing it with Fd=0 defines a different model, not the symmetric BCS problem. Thus the claimed 'second solution at µ_R=µ' is the limit of a branch that exists for δ≠0, not a solution of the δ=0 self-consistency equations. The paper's own Eq. (12) shows ∂∆/∂n diverges on this branch, so the correction in Eq. (4c) is not a regular perturbation. Since Eq. (4c) is imported from Ref. [34] without re-derivation, the entire δ=0 second branch rests on an unverified singular limit. The T=0 ∆0/3 algebra shows only that the modified equations are internally consistent, not that the standard BCS problem has two equilibrium branches.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the BCS gap and quasiparticle-population equations as formulated in earlier work by the author, Eqs. (4)-(5), and reports two solution branches for the energy gap below the transition temperature. The main new claim is that even for an attraction band symmetric about the chemical potential (µR = µ), a second solution exists with a smaller gap, equal to ∆0/3 at T = 0, and nonzero quasiparticle populations down to zero temperature. The second branch is obtained as a limit µR → µ in which the correction term (µR − µ)F in the quasiparticle distribution tends to a finite nonzero value M while F itself diverges.","tokens_in":9703,"tokens_out":7168,"duration_ms":68244,"significance":"If correct, the result would imply non-uniqueness of the BCS equilibrium in the standard symmetric case and would identify a new class of superconducting states with residual quasiparticle populations. The paper does contain a genuine algebraic consistency check at T = 0 for the modified system: filling the interval |ξ| < ∆0/(3√3) with nξ = 1 and setting ∆ = ∆0/3 makes both the gap equation and the denominator Fd vanish. However, that check verifies only the modified equations, not the equilibrium conditions of the symmetric BCS problem, and the central claim therefore remains unsupported.","major_comments":[{"comment":"At µR = µ, the equilibrium condition (11) reduces to ln((1−nξ)/nξ) − βϵξ = 0, whose unique solution is nξ = f(βϵξ). The occupations (8) with M ≠ 0 give ∂(ln Z)/∂nξ = βM/ϵξ ≠ 0, so they are not stationary points of the partition function. The statement that a divergent F with finite M makes these occupations equilibrium distributions conflates the limit of maximizers for δ ≠ 0 with the maximizer of the δ = 0 functional; because F diverges, the convergence is not uniform and the limits do not commute. Consequently the second branch is not an equilibrium distribution of the symmetric BCS model.","section":"Discussion, Eq. (11) and Eq. (8)"},{"comment":"At δ = µR − µ = 0, the denominator Fd in Eq. (5b) is positive and F = 0 for the standard distribution nξ = f(βϵξ); the 0/0 limit used to define the second branch arises only along a sequence δ → 0 with Fd → 0. The additional condition Fd = 0 imposed after Eq. (9) is not part of the δ = 0 system (5). Replacing Eq. (5b) by the constraint Fd = 0 therefore defines a different model, and the paper does not show that the limiting branch solves the original δ = 0 self-consistency equations.","section":"Methods, Eq. (5b), and Discussion after Eq. (9)"},{"comment":"The quasiparticle chemical-potential correction in Eq. (4c) is imported from Ref. [34] without derivation, and all new results depend on it through the singular combination (µR − µ)F. Because the paper claims a result for the textbook BCS problem, it must either derive Eq. (4c) from the BCS Hamiltonian or otherwise justify the singular limit; the citation to Ref. [34] is insufficient, especially in view of the uniqueness theorems for the standard symmetric gap equation cited in Refs. [40,41].","section":"Methods, Eq. (4c)"}],"minor_comments":[{"comment":"The text contains several grammatical slips, including 'even at µR = µ' in the Conclusions (should be 'even for µR = µ') and 'Old theoretical predictions by made N. W. Ashcroft' in the Introduction.","section":"Conclusions"},{"comment":"The caption of Fig. 2 labels the plotted quantity as N1,2 − Nµ, while the axes and Eq. (6) indicate that the plot actually shows (N1,2 − Nµ)/(2σ0); please make the caption consistent.","section":"Fig. 2 and Eq. (6)"},{"comment":"The notation for the limiting M is inconsistent: Eq. (7b) defines M^(0)_1,2(T), while Fig. 3(b) labels the curve M2^(0); unify the notation.","section":"Fig. 3 and Eq. (7)"}],"recommendation":"reject","confidential_remarks":"The central claim rests entirely on the author's earlier Eq. (4c) and on a singular limit that is not justified from the δ = 0 variational problem. The algebraic T = 0 check is a useful consistency test of the modified system but does not establish that the second branch solves the symmetric BCS problem. Unless the author can re-derive Eq. (4c) from a microscopic theory and prove that the δ → 0 limit commutes with extremization, the result is not publishable as stated. There is also a concern that the manuscript does not squarely address the standard uniqueness results [40,41] beyond asserting a partial restoration of BCS theory."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear colleague,\n\nYou should know this paper makes a provocative claim: even in the symmetric attraction-band limit (µR=µ), the BCS equations have two equilibrium branches, and the second one has gap Δ0/3 at zero temperature. The finite-temperature persistence of that second branch is genuinely new relative to the author's own Refs. [34,35]; Ref. [34] stated standard BCS is restored at µR=µ, and Ref. [35] was T=0 only. The T=0 symmetric-limit algebra is internally consistent: filling |ξ|<Δ0/(3√3) satisfies the modified gap equation and makes the denominator Fd vanish. That is real, reproducible work.\n\nBut the central claim does not hold up. Everything sits on Eq. (4c), the quasiparticle chemical-potential correction imported from Ref. [34], and this paper neither re-derives it nor offers an independent check. Worse, at exactly µR=µ, Eq. (5c) has no correction term—nξ is just f(βϵξ). The second branch is obtained by a singular limit in which F diverges and M=(µR-µ)F stays finite, then imposing Fd=0. That is not a solution of the δ=0 self-consistency equations; it defines a different limiting model. The paper's own Eq. (12) shows ∂Δ/∂n diverges on this branch, so the correction is not a regular perturbation and the limit cannot be taken for granted. The author acknowledges the apparent contradiction but does not resolve it. The standard uniqueness theorems for the δ=0 gap equation [40,41] are cited but never reconciled.\n\nFinite-temperature existence is supported only by numerical plots; no convergence details or existence proof. So the reader's conditional verdict is fair, and the stress-test note lands: the second solution at µR=µ is a limit of branches that exist for δ≠0, not a solution of the textbook problem.\n\nWho should read this? People working on metastable states in BCS or on the author's modified formalism. It is not a paper about experimental superconductivity—no target material or preparation path is given. It deserves a serious referee because the singular-limit question is subtle and the derivation of Eq. (4c) should be scrutinized. I would send it out, but I would tell the referee to focus on exactly that equation and on whether the limit defines the same model. I would not accept the main claim as stated.","headline":"The finite-temperature second branch is a real extension of the author's earlier formalism, but the claim that standard symmetric BCS has a second equilibrium solution rests on a singular limit that is never justified; the Δ0/3 result is a consistency check of modified equations, not of the textbook problem.","tokens_in":10285,"tokens_out":4253,"would_cite":false,"duration_ms":40657,"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":"Even in the symmetric textbook BCS case, there are two equilibrium solutions for the energy gap, and at zero temperature the second gap is exactly one third of the standard one.","keywords":["BCS theory","superconductivity","energy gap","quasiparticle populations","chemical potential","attraction band","multiple equilibrium solutions","phase transition"],"falsifier":"Directly integrate the self-consistent gap and population equations for a sequence of small nonzero $\\mu_R-\\mu$ and extrapolate to $\\mu_R=\\mu$: the second branch is real only if $F_2$ diverges while $M_2=(\\mu_R-\\mu)F_2$ tends to a finite nonzero value, and a calculation that finds no such finite limit would falsify it.","tokens_in":9146,"feed_emoji":"⚛️","tokens_out":12706,"duration_ms":115293,"temperature":0.7,"pith_summary":"This paper argues that the standard BCS theory of superconductivity admits more equilibrium solutions than the textbook one. Solving the self-consistent gap and quasiparticle-population equations with a chemical-potential correction, the author finds two solution branches for every temperature below the phase transition, including in the symmetric case where the attraction band is centered on the chemical potential. One branch is the usual BCS solution; the other has a smaller energy gap and nonzero quasiparticle populations even at zero temperature, where its gap equals one third of the standard value. If this is right, the textbook BCS state is not the only equilibrium state, so metastable or additional superconducting states may appear in real materials and in other BCS-type paired systems.","feed_headline":"Even in textbook BCS, a second gap survives at one third the value","feed_subtitle":"The symmetric textbook problem has a second equilibrium branch with residual quasiparticles at zero temperature.","key_machinery":"The load-bearing object is the quasiparticle energy shift produced by the chemical-potential mismatch. In Eq. (5c) the occupation is $n_\\xi=[\\exp(\\beta(\\epsilon_\\xi-(\\mu_R-\\mu)(\\xi-F)/\\epsilon_\\xi))+1]^{-1}$, with $F$ the ratio of two integrals over the attraction band. In the symmetric limit $\\mu_R\\to\\mu$, the function $F_2$ diverges, but the product $M_2=(\\mu_R-\\mu)F_2$ tends to a finite nonzero value, so the combination $M_2/\\epsilon_\\xi$ survives in the exponent and creates a distinct population distribution. Equivalently, for the second branch the derivative $\\partial\\Delta/\\partial n_{\\xi i}$ diverges, so an infinitesimal chemical-potential difference can reorganize a finite number of quasiparticles; the paper uses this singular limit to obtain a second, unequal-gap equilibrium.","core_discovery":"The central claim is that the BCS equations, when quasiparticle energies carry the correction $\\tilde\\mu_k=(\\mu_R-\\mu)(\\xi-F)/\\epsilon_\\xi$, admit two families of solutions below a phase transition temperature $T_{ph}(\\mu_R-\\mu)$. In the symmetric limit $\\mu_R\\to\\mu$, the first family is the standard BCS solution with gap $\\Delta_0$ at $T=0$, while the second family has a smaller gap and satisfies $\\Delta_2(0)=\\Delta_0/3$. The second solution is an equilibrium state in the same grand-canonical sense: it maximizes the partition function, and it has fully occupied quasiparticle states in a band around the Fermi level at zero temperature, not an empty quasiparticle distribution. Both branches vanish at the same $T_{ph}$, so the new branch does not change the critical temperature but changes the gap magnitude and the quasiparticle content.","pith_inferences":["A natural next step not taken here is to test the dynamical stability of the second branch; if it is only metastable, it could be populated by fast cooling or by electromagnetic driving that tilts the effective band.","The exact $T=0$ ratio $\\Delta_0/3$ is a concrete experimental target: tunneling spectroscopy or specific-heat measurements on conventional superconductors could look for a second gap feature at one third of the main gap.","The same singular-limit mechanism may appear in other BCS-type settings, such as nuclear pairing, neutron-star superfluids, or ultracold Fermi gases, where symmetric bands and grand-canonical equilibrium are common assumptions.","Because the second branch is invisible when one sets $\\mu_R=\\mu$ before solving, numerical implementations should treat the product $M=(\\mu_R-\\mu)F$ as an independent variable rather than dropping the correction."],"forward_implications":["The symmetric BCS problem is not single-valued below $T_c$: a second equilibrium branch with smaller gap and persistent quasiparticles coexists with the standard state.","At $T=0$ the second branch has fully occupied quasiparticle states in $\\xi\\in[-\\Delta_0/(3\\sqrt{3}),\\Delta_0/(3\\sqrt{3})]$, so a sample in that branch would show residual normal excitations even at absolute zero.","Both branches share the same phase transition temperature, so the new branch is distinguished by gap size and quasiparticle population, not by where superconductivity appears.","For asymmetric attraction bands, the transition can be first order and is accompanied by a jump in the total particle number when $\\mu_R\\neq\\mu$.","Even the textbook symmetric limit retains the second branch, meaning that analyses which set $\\mu_R=\\mu$ before solving the equations will miss one equilibrium state."],"supporting_citations":[{"why":"Supplies the quasiparticle chemical-potential correction and the earlier finding of multiple gap and population solutions for asymmetric attraction bands that this paper extends.","marker":"[34]"},{"why":"Provides the zero-temperature analysis of the two solution families, including the value $\\Delta_2(0)=\\Delta_0/3$ and the corresponding occupation interval used to anchor the second branch.","marker":"[35]"},{"why":"Defines the standard BCS gap equation and textbook solution that the paper recovers as the first branch $\\Delta_1$.","marker":"[36]"},{"why":"The original BCS formulation, the baseline for the standard superconducting equilibrium state that the second branch is compared with.","marker":"[37]"},{"why":"Used to argue that the standard gap equation alone, with $F=0$ and $\\mu_R=\\mu$, has a unique solution, so the new branch must come from the chemical-potential correction.","marker":"[40]"},{"why":"Supplies the temperature dependence and uniqueness of the standard solution, the comparison curve against which the new branch is assessed.","marker":"[41]"}],"fun_headline_variants":["Even symmetric BCS has two solutions: a second gap at one third the classic value","Second BCS solution survives at zero temperature with gap one third of textbook value","Even textbook case has a second BCS gap, one third at zero temperature","Symmetric BCS yields two branches: the second gap is one third at absolute zero"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The second solution depends on a single load-bearing premise: the extra term that shifts quasiparticle energies when the chemical potential differs from the band center is the correct grand-canonical BCS thermodynamics, so that an infinitesimal difference $\\mu_R-\\mu$ can, through a divergent $F$, still leave a finite population shift.","fun_headline_variants_meta":{"raw":{"variants":["Even symmetric BCS has two solutions: a second gap at one third the classic value","Second BCS solution survives at zero temperature with gap one third of textbook value","Even textbook case has a second BCS gap, one third at zero temperature","Symmetric BCS yields two branches: the second gap is one third at absolute zero"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000752,"raw_usage":{"total_tokens":3321,"prompt_tokens":897,"completion_tokens":2424,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":513,"completion_tokens_details":{"reasoning_tokens":2337}},"tokens_in":513,"tokens_out":2424,"duration_ms":18430,"temperature":1.0,"reasoning_tokens":2337,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T13:16:39.501008+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Directly integrate the self-consistent gap and population equations for a sequence of small nonzero $\\mu_R-\\mu$ and extrapolate to $\\mu_R=\\mu$: the second branch is real only if $F_2$ diverges while $M_2=(\\mu_R-\\mu)F_2$ tends to a finite nonzero value, and a calculation that finds no such finite limit would falsify it.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the quasiparticle chemical-potential correction and the earlier finding of multiple gap and population solutions for asymmetric attraction bands that this paper extends."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the zero-temperature analysis of the two solution families, including the value $\\Delta_2(0)=\\Delta_0/3$ and the corresponding occupation interval used to anchor the second branch."},{"cited_title":"Introduction to Superconductivity","cited_arxiv_id":null,"evidence_quote":"Defines the standard BCS gap equation and textbook solution that the paper recovers as the first branch $\\Delta_1$."},{"cited_title":"Bardeen, L","cited_arxiv_id":null,"evidence_quote":"The original BCS formulation, the baseline for the standard superconducting equilibrium state that the second branch is compared with."},{"cited_title":"Watanabe","cited_arxiv_id":null,"evidence_quote":"Used to argue that the standard gap equation alone, with $F=0$ and $\\mu_R=\\mu$, has a unique solution, so the new branch must come from the chemical-potential correction."},{"cited_title":"Watanabe","cited_arxiv_id":null,"evidence_quote":"Supplies the temperature dependence and uniqueness of the standard solution, the comparison curve against which the new branch is assessed."}],"review_version":1}