This paper shows that, for a standard simplified model of superconducting pairing, the Bardeen–Cooper–Schrieffer (BCS) variational energy gives the correct ground-state energy per volume once the system is made large. In other words, the simple BCS formula that physicists use to estimate the lowest energy of paired fermions matches the exact many-body ground-state energy density for this reduced model in the thermodynamic limit.

The authors study a reduced BCS Hamiltonian acting on the full fermionic Fock space. The model has spin-1/2 fermions in a periodic box and an attractive pairing interaction that only acts on momenta inside a thin energy shell around the Fermi surface (the set of momenta whose kinetic energy is close to the chemical potential). For fixed chemical potential they prove a finite-volume estimate: the exact ground-state energy differs from the minimum of the corresponding finite-volume BCS functional by an error that does not grow with the box volume. From this they deduce that, as the box size goes to infinity, the ground-state energy per volume equals the sum of the free Fermi-sea energy density and the BCS energy density.

At a high level, the argument revisits the old “approximating Hamiltonian” idea. One replaces the collective pair operator by a complex parameter, which produces a quadratic (solvable) Bogoliubov Hamiltonian plus a fluctuation term. The paper works in a finite-volume operator framework and makes explicit estimates on those fluctuations. The main technical result is a uniform, volume-independent bound on the difference between the exact many-body energy and the BCS variational energy.

The result matters because it puts the familiar BCS variational principle on firmer mathematical ground for this class of mean-field pairing models. It quantifies how close the simple variational approximation is to the true many-body energy for any large but finite system, and it shows that the usual BCS correction to the free energy density survives in a suitable high-density scaling. The authors also give asymptotic expansions in the high-density regime showing that the BCS correction remains nontrivial and that the gap parameter (the BCS “order parameter”) stays of order one under their scaling.