A lower bound on the charge form of the equal-mass two-component unitary Fermi gas for all particle numbers, sharp for balanced systems
Theorem. We consider N fermions of one species and M of another, all with the same mass, in three dimensions, interacting only between species through a zero-range interaction with infinite scattering length. For the Ter-Martirosyan–Skornyakov charge form Tμ in the normalization of Finco and Teta we prove: (A) for all N, M ≥ 1, every μ > 0 and every charge ξ, TμN+M[ξ] ≥ (π²/√(N+M)) √μ ‖ξ‖²; (B) for N = M the constant is optimal: it is the value approached by N far-separated dimers, realized by trial charges with a logarithmic cutoff; (C) if Hypothesis (I1) below holds, the equal-mass Finco–Teta quadratic form at unitarity is bounded below, indeed nonnegative, for every N + M. Theorems A and B concern the explicitly defined charge form and use no hypothesis. Method. A local Pauli pair bound: for one orbital g, the pair operator Bg = c(g)a(g) satisfies Bg*Bg ≤ ⟨g, Dg⟩ (dΓ↑(A) + dΓ↓(A)), where A solves the Lyapunov equation DA + AD = |g⟩⟨g|. It follows from an exact vacancy identity in which each species contributes a factor between zero and ‖e−tA/2g‖², because an orbital can be occupied at most once. Two exact spatial resolutions with the same measure: one of the kinetic energy and one of a damped finite-range pair interaction. Integrating the local bound gives a damped all-sector inequality. A zero-range bridge, proved in the article: the charge form is the rescaled limit of the critically subtracted deficit of a finite-range Gram form, so a deficit bound of order ℓ³ passes to the charge form. The damping costs only order ℓ⁴. Scope. The result is a lower bound on the charge form only. Nothing is claimed about closability or closed extensions of the forms, closedness of the physical form, self-adjoint realizations, limits of regularized Hamiltonians, or the thermodynamic limit; the constant tends to zero as N + M grows, and for N = M this cannot be avoided. The proof uses a damped finite-range inequality and does not give the undamped one. Equal masses and Fermi statistics are both essential to the method; for bosons the key inequality fails. For N ≠ M the optimal constant is not known. Hypothesis (I1) and its support. (I1) states that, at equal masses and infinite scattering length, the charge form of the Finco–Teta quadratic form is NM · Tμ, and that every element of its domain splits into a regular part of nonnegative energy plus the potential of a charge. Only Corollary C and the physical reading of Theorem A depend on it. It is supported by the conversion of the formulas of Finco and Teta (arXiv:1012.0151v1, Sect. 2), by agreement with Moser and Seiringer for 2+2, by an independent derivation from the contact condition that was fixed by its hash before the comparison, and by numerical consistency checks (a recomputation with mass-dependent constants reproduces the known 2+1 critical mass ratio 13.607; this checks only the constant and sign of the single exchange). These derivations are formal, like the original construction; they are not a zero-range limit of microscopic Hamiltonians. Files. The article (PDF) and a supplementary archive with: the LaTeX source of the article; finite-dimensional tests of the local Pauli bound, variational computations with trial charges, and a numerical test of the damped inequality (side evidence, not part of any proof); the independent derivation of the charge form and its comparison with the literature; the review reports with the reviewers' scripts; a one-step check (check.py; the option --replay reruns four fast scripts and compares their results); a statement of trust and scope. AI tools. AI systems (Claude Opus 5.5, Claude Fable 5.2, GPT-6 Astra and GPT-6.2 Astra) did the computational and drafting work under the author's direction; the article's section on AI use describes it. Reviews crossed developers: the main argument, drafted with a GPT model, was reviewed blind twice by Claude Opus and audited by Claude Fable; the derivation and the paper were reviewed by models of both developers. No human expert has yet checked the proofs line by line.
Authors
- Neko2333OWO
Publication Details
- Journal
- Zenodo (CERN European Organization for Nuclear Research)
- Published
- 2026-11-18
- DOI
- https://doi.org/10.5281/zenodo.24347778
- Primary Topic
- Cold Atom Physics and Bose-Einstein Condensates
- Type
- preprint