BCS Theory

Definition

BCS theory is the microscopic, quantum-mechanical theory of conventional superconductivity. It states that below a critical temperature T_c, electrons near the Fermi surface bind into Cooper pairs through an attractive interaction carried by lattice vibrations (phonons). These pairs, behaving as bosons, condense into a single coherent quantum state. The condensation opens an energy gap Delta in the excitation spectrum, and that gap is what produces zero electrical resistance and the expulsion of magnetic fields (the Meissner effect).

Named for John Bardeen, Leon Cooper, and Robert Schrieffer, who published it in 1957 — 46 years after superconductivity was discovered and after Einstein, Bohr, Feynman, and Landau had all failed to crack it.

How it works — the three ideas

BCS rests on three pillars, each contributed by one initial of the acronym, stacked in sequence.

1. The phonon attraction (the "glue"). Electrons repel via Coulomb's law, so how can they bind? The trick is the lattice. As an electron streaks through the crystal it drags the heavy positive ions slightly toward it. The ions are roughly 100,000 times more massive than the electron, so they respond sluggishly and the dimple of excess positive charge lingers after the first electron has flown on. A second electron, arriving milliseconds later in lattice terms, feels that leftover positive region and is drawn in. The two electrons never sit in the same place at the same time, so they sidestep their mutual repulsion. In the language of quantum field theory, this delayed, retarded interaction is the exchange of a virtual phonon.

electron 1 --emits--> phonon --absorbed by--> electron 2
net effect: weak attraction for energies within hbar*omega_D of E_F

2. Cooper pairing (the instability). In 1956 Cooper proved a startling result: take a filled Fermi sea and add just two extra electrons with any attractive interaction, no matter how weak. They always form a bound state — a Cooper pair — with energy below the Fermi level. The normal metallic state is therefore unstable: the system will always lower its energy by pairing. The pair partners have equal and opposite momentum (k, −k) and opposite spin, giving zero net momentum and zero net spin, so the pair is a boson.

3. Condensation and the gap (the macroscopic state). Schrieffer wrote down a single many-body wavefunction in which all the electrons near the Fermi surface are paired coherently, with one common quantum phase. Because pairs are bosons, there is no Pauli exclusion stopping them from all occupying the same state — they form a macroscopic condensate, a single quantum object the size of the whole sample. Breaking a pair to make an ordinary electron excitation now costs a minimum energy 2Delta. That forbidden zone — no single-particle states within Delta of the Fermi level — is the energy gap.

The energy gap and its universal ratio

Solving the BCS gap equation self-consistently at zero temperature gives one of the most celebrated results in condensed matter physics:

Delta(0) = 1.76 * k_B * T_c

equivalently:  2*Delta(0) / (k_B * T_c) = 3.52   (universal)

The factor 1.76 (precisely π/e^γ ≈ 1.764, with γ the Euler-Mascheroni constant) is material-independent in the weak-coupling limit. Whether you measure lead, tin, or aluminium, the ratio of the gap to the transition temperature lands near 3.52. The gap is temperature-dependent: it equals 1.76 k_B T_c at absolute zero, shrinks as you warm the sample, and closes smoothly to zero exactly at T_c, where superconductivity disappears.

Worked example — the gap of niobium

Niobium is the workhorse superconductor of MRI magnets and particle accelerators, with T_c = 9.3 K. Let us compute its zero-temperature gap and the energy scale it sets.

Given:   T_c = 9.3 K
         k_B = 8.617e-5 eV/K

Delta(0) = 1.76 * k_B * T_c
         = 1.76 * 8.617e-5 eV/K * 9.3 K
         = 1.76 * 8.014e-4 eV
         = 1.41e-3 eV
         = 1.41 meV

Pair-breaking energy 2*Delta(0) = 2.82 meV

So it takes only about 1.4 milli-electronvolts to break a Cooper pair in niobium — roughly a thousand times smaller than a typical chemical bond. The measured tunnelling gap of niobium is about 1.5 meV, slightly above the BCS value because niobium is a moderately strong-coupling superconductor (its real ratio 2Delta/k_B T_c is about 3.8 rather than 3.52). This is why superconductors only work when cold: thermal energy k_B T must stay well below 2Delta, or the gas of thermal phonons simply shatters the pairs as fast as they form.

Now the isotope effect. BCS predicts T_c ∝ M^-1/2. Mercury offers a clean test: comparing mercury-198 and mercury-202, the predicted fractional shift in T_c is

T_c(202)/T_c(198) = (198/202)^(1/2) = 0.990
shift = 0.990 * 4.16 K - 4.16 K = -0.04 K

The measured shift matched this to within experimental error — direct proof that phonons, whose frequencies scale as M^-1/2, set the transition temperature.

Variants and regimes

Regime / class	Pairing glue	Gap symmetry	2Δ/k_B T_c	Fits BCS?
Weak-coupling (Al, Sn, In)	Phonons	s-wave (isotropic)	≈ 3.5	Yes, textbook
Strong-coupling (Pb, Nb, Hg)	Phonons	s-wave	3.8 – 4.5	Yes, via Eliashberg
MgB₂ (T_c = 39 K)	Phonons	Two s-wave gaps	≈ 1.7 and 4.5	Yes, multiband BCS
Heavy-fermion (CeCoIn₅)	Spin fluctuations	d-wave	~ 5 – 6	BCS-like, non-phonon
Cuprates (YBCO, T_c ≈ 92 K)	Likely magnetic	d-wave (nodal)	≈ 5 – 8	No — beyond BCS
Iron pnictides	Spin fluctuations	s± (sign-changing)	≈ 4 – 7	No — beyond BCS

The Eliashberg extension keeps the BCS skeleton but treats the phonon retardation exactly, capturing strong-coupling metals like lead. Bose-Einstein-condensate physics sits at the opposite extreme of the so-called BCS-BEC crossover, where pairs become tightly bound molecules rather than overlapping Cooper pairs.

Common pitfalls and misconceptions

• "Cooper pairs are little molecules." No. The two electrons in a pair are separated by the coherence length — hundreds of nanometres in conventional superconductors — so a single pair envelops millions of other pairs. They overlap massively and act collectively. A pair is more like a correlation than a bond.
• "The electrons attract directly." Never directly. The bare Coulomb interaction is always repulsive. The attraction is entirely a second-order, lattice-mediated effect, and it only wins for electrons within hbar·omega_D (a few tens of meV) of the Fermi surface.
• "Superconductivity is just very low resistance." It is exactly zero, and qualitatively different. The energy gap forbids the small-energy scattering events that cause resistance — there is literally nowhere for a paired electron to scatter to without paying 2Delta.
• "BCS explains all superconductors." It explains the conventional ones. The cuprate and iron-based high-T_c materials have d-wave or sign-changing gaps, near-zero isotope effect, and ratios far above 3.52 — strong signs the glue is not phonons.
• "The gap is a band gap like in semiconductors." Different beast. A semiconductor's band gap is fixed by chemistry and exists at all temperatures. The superconducting gap is a many-body, temperature-dependent gap that opens at T_c and closes again above it.
• "Pairs carry the current like charged balls." The supercurrent is carried by the coherent phase of the condensate wavefunction. It is the rigidity of that single macroscopic phase — not ballistic pair transport — that makes the current dissipationless.

Applications

• MRI and NMR magnets. Niobium-titanium and niobium-tin BCS superconductors carry the huge persistent currents that generate the 1.5–7 tesla fields inside every hospital MRI scanner, with no ohmic heating.
• Particle accelerators. The LHC's bending magnets and superconducting RF cavities run on Nb and Nb₃Sn, kept below their gap temperatures with superfluid helium.
• SQUID magnetometers. Josephson junctions between BCS superconductors detect magnetic fields a hundred billion times weaker than Earth's — used in magnetoencephalography to read brain activity.
• Superconducting qubits. Many quantum computers (the transmon) use Cooper-pair tunnelling across Josephson junctions; the energy gap protects the qubit from quasiparticle noise.
• Maglev and power. Magnetic levitation trains and lossless power-transmission demonstrators exploit persistent supercurrents and the Meissner effect.
• Single-photon detectors. Superconducting nanowire detectors sense one photon by the local breaking of pairs — the energy gap sets the detection threshold.

Derivation and performance analysis

The heart of BCS is a variational ground state in which each pair state (k, −k) is either occupied by a pair (amplitude v_k) or empty (amplitude u_k):

|BCS> = product over k of ( u_k + v_k * c†(k,up) * c†(-k,down) ) |0>
with |u_k|^2 + |v_k|^2 = 1

Minimising the reduced BCS Hamiltonian energy with respect to the amplitudes yields the self-consistent gap equation:

Delta = V * sum_k  Delta / (2 * E_k) * tanh( E_k / (2 k_B T) )
where E_k = sqrt( epsilon_k^2 + Delta^2 )  is the quasiparticle energy

Here epsilon_k is the normal electron energy measured from E_F, V is the attraction strength, and E_k is the energy to create a single excitation — note it can never be smaller than Delta, which is the gap made manifest. Two limits drop out cleanly:

• At T = 0: the tanh saturates to 1 and the integral gives Delta(0) = 2 hbar·omega_D · exp(−1/(N(0)V)), where N(0) is the density of states at the Fermi level. The exponential, non-perturbative form explains why no power-series (perturbative) attempt for 46 years could find the answer.
• At T = T_c: setting Delta → 0 gives k_B T_c = 1.13 hbar·omega_D · exp(−1/(N(0)V)). Dividing the two expressions cancels the messy material constants and leaves the clean universal ratio 2Delta(0)/k_B T_c = 3.52.

The same machinery predicts a discontinuous jump in the electronic specific heat at T_c of (C_s − C_n)/C_n = 1.43 — another parameter-free number confirmed across many metals. The fact that one variational wavefunction reproduces the gap, the Meissner effect, the specific-heat jump, the isotope effect, and the exponentially activated low-temperature behaviour, all without adjustable fudge factors, is why BCS is regarded as one of the most successful theories in physics — and why it earned the 1972 Nobel Prize.