Abstract
Free quantum gases have two classical instabilities, Bose–Einstein condensation and Cooper pairing. For a density of states N(ξ) ∝ ξa both are governed by Mellin transforms: Γ(a+1)ζ(a+1) for the Bose density and 2Γ(a)η(a) for the fermionic pair kernel. At a = 0 these factors sit at the two poles of the completed zeta function Λ(s) = π−s/2Γ(s/2)ζ(s), at s = 1 and s = 0, each with residue 1. Their Laurent constants are 0 and ln(π/2) − γ, and the second one is the BCS constant. For 0 < a < 2 mean-field pairing has a threshold at the coupling λ_c = a. Our main theorem is the closed limit (Δ₀/Tc)a → 2a+1 η(a) Γ(1 + a/2)/Γ(1 − a/2) as λ ↓ λ_c. It interpolates between the BCS value πe−γ (a → 0) and 2 ln 2 for Dirac cones (a = 1). It gives the exact constant π²η(1/2)²/Γ(3/4)⁴ = 1.6015… for semi-Dirac dispersion (a = 1/2), where η(1/2) lies in the critical strip, and new constants for triple-Weyl (a = 2/3) and three-dimensional semi-Dirac (a = 3/2) nodes. The ratio tends to zero at the next gamma pole, a = 2. A short proposition shows that a = Σᵢ 1/nᵢ − 1 for any dispersion that scales with degrees nᵢ, which places these cases on one line. As a frame for the series of papers to which this one belongs, we prove that left of the critical strip the n-th derivative of an L-function at a pole of its gamma factor is a combination of derivatives of order at most n − 1 on the right, while away from such a pole order n is needed. This explains which derivatives in physics reduce to values. A primer and a plain-language summary make the paper readable without a background in number theory or superconductivity. Labels separate proved, cited, numerical and interpretive statements. The paper ends with a list of possible implications.