BCS pairing and superconductivity

Assume an effective attractive interaction within an energy window $\Omega$ around the Fermi surface. In conventional superconductors this attraction can emerge from electron-phonon dynamics after other degrees of freedom are treated appropriately; it is not the assertion that bare Coulomb repulsion becomes attractive at every distance. The idealized reduced model takes a constant pairing matrix element in that window and uses the normal-state density of states as approximately constant. Weak coupling means its dimensionless product with the attraction is small.

The Cooper problem shows that the available states near an otherwise filled Fermi sea make arbitrarily weak attraction unstable to pairing in the idealized channel. BCS goes beyond one added pair by varying coherent amplitudes for all paired momenta. A factor $u_k+v_k c^\dagger_{k\uparrow}c^\dagger_{-k\downarrow}$ mixes an empty pair sector with an occupied one. Normalize each factor so $|u_k|^2+|v_k|^2=1$. This state does not have a fixed particle number, although number projection is possible and bulk number-conserving observables can be treated consistently.

Define $\xi_k$ as normal-state single-particle energy measured from the chemical potential. Decoupling the interaction in the pair channel produces a quadratic Hamiltonian for particle and hole amplitudes. Its positive excitation energy is $E_k=\sqrt{\xi_k^2+|\Delta|^2}$. The transformation preserves fermionic anticommutation relations. Although both Bose and Fermi calculations use the name Bogoliubov, their algebra and normalization differ; the identical name is not permission to reuse a bosonic formula.

The gap is determined self-consistently rather than inserted as a free extra interaction. At zero temperature in the weak-coupling constant-density-of-states model, its scale is exponentially small compared with the cutoff window. This exponential dependence explains why modest interaction changes can greatly change a predicted transition scale. A fully gapped isotropic s-wave state has minimum excitation energy $|\Delta|$; creating two quasiparticles typically requires at least $2|\Delta|$. Other pairing symmetries can have nodes, so a universal hard gap should not be inferred for every superconductor.

A gap or a two-body bound state alone does not prove a material carries a robust supercurrent. Phase stiffness describes the energy cost of spatially twisting the collective phase. Coupling that phase consistently to electromagnetism produces screening and the Meissner response. The pair charge has magnitude twice the electron charge. Calling this simply a broken local gauge symmetry is misleading: gauge transformations are redundancies, and observable statements concern correlations and electromagnetic response.

The usual finite-temperature BCS transition is a mean-field result, most reliable away from the narrow fluctuation region in suitable three-dimensional weak-coupling systems. In strictly two-dimensional neutral systems, or suitable thin-film limits, phase fluctuations and vortex physics can control the transition instead of the mean-field pair-formation temperature. Strong coupling, disorder, multiband structure, and unconventional pairing require extensions. Compare the assumptions with an experiment before interpreting a fitted gap as evidence for one unique pairing mechanism.