Cryogenic Systems

Cooper Pair

Q: Why do two electrons attract each other to form a Cooper pair?

Although two electrons repel through the Coulomb interaction, in a metal below the critical temperature they experience a net attraction mediated by lattice vibrations (phonons). A moving electron distorts the positive ion lattice, leaving a transient region of higher positive charge density. A second electron with opposite momentum and spin is drawn toward that distortion. Leon Cooper showed in 1956 that even an arbitrarily weak attraction is enough to bind two electrons at the Fermi surface into a state of lower energy than the unpaired Fermi sea. The pairing is strongest for electrons with equal and opposite momenta (k and minus k) and opposite spins, forming a spin-singlet, zero-net-momentum pair. The characteristic phonon energy that sets the attraction scale is the Debye energy, which is why the superconducting transition temperature correlates with the lattice Debye temperature in conventional superconductors.

Q: How large is a Cooper pair compared to the electron spacing?

A Cooper pair is enormous compared with the spacing between conduction electrons. Its size is set by the BCS coherence length, given approximately by xi = hbar times v_F divided by (pi times the energy gap). For niobium the coherence length is about 38 nm, for aluminum about 1600 nm, and for clean conventional metals it ranges from tens to thousands of nanometers. Since the electron spacing in a metal is only about 0.2 to 0.3 nm, a single Cooper pair physically overlaps with the centers of millions of other pairs. This heavy overlap is exactly why all the pairs lock into one coherent macroscopic quantum state, described by a single complex order parameter with a well-defined phase. That phase coherence is what produces the Josephson effect, flux quantization, and the lossless screening currents in superconducting RF cavities.

Q: Why is RF surface resistance not exactly zero even though Cooper pairs carry no DC loss?

Cooper pairs carry DC supercurrent with zero resistance, but at radio and microwave frequencies the alternating field accelerates and decelerates the pair condensate every cycle. Because the pairs have inertia (kinetic inductance), they cannot perfectly screen the time-varying field, so a small residual electric field penetrates the surface and acts on the thermally excited unpaired electrons (quasiparticles). Those quasiparticles dissipate power, producing a nonzero RF surface resistance that scales roughly as Rs proportional to (omega squared over T) times exp(minus delta over kT), following the Mattis-Bardeen theory. This is why superconducting RF cavities are cooled to 2 to 4 K, well below Tc, to freeze out quasiparticles. A niobium cavity at 2 K and 1.3 GHz can reach surface resistance near 10 nanoohms, roughly five orders of magnitude below copper, but a residual resistance floor from trapped flux and surface defects keeps it from being truly zero.

/KOO-per pair/

Two electrons of opposite momentum and spin, bound together by a phonon-mediated attraction below the critical temperature, form the charge carrier behind superconductivity. Rather than acting as independent particles, these pairs condense into a single coherent quantum state described by BCS theory, separated from unpaired electron states by an energy gap of typically 1 to 3 meV. The condensate carries supercurrent with zero DC resistance and gives superconducting films their kinetic inductance. In RF and microwave hardware, this pairing is the physical basis for ultra-high-Q superconducting resonators, near-quantum-limited cryogenic receivers, and Josephson-based devices operating at 2 to 4 K.

Category: Cryogenic Systems

Energy Gap (2Δ): ~1 to 3 meV

Nb Coherence Length: ≈ 38 nm

How Paired Electrons Enable Lossless Supercurrent

In a normal metal, conduction electrons scatter off lattice defects and thermal vibrations, dissipating energy as heat and producing ordinary resistance. Below the critical temperature, a fundamentally different mechanism takes over. Leon Cooper demonstrated in 1956 that the Fermi sea is unstable against an arbitrarily weak attractive interaction, and that two electrons near the Fermi surface will bind into a pair with lower total energy. The attraction is not direct; it is mediated by the lattice. One electron polarizes the positively charged ion cores as it passes, and a second electron with opposite momentum is drawn into the trailing region of enhanced positive charge. Because the ions respond on the slow timescale of the Debye frequency, the two electrons need not be close in space, which is the origin of the very large pair size.

The full many-body extension by Bardeen, Cooper, and Schrieffer (BCS) showed that all such pairs condense into one macroscopic wavefunction with a common phase. This condensate is separated from single-particle excitations by the superconducting energy gap, 2Δ, which at zero temperature equals about 3.52 k_BT_c for a conventional weak-coupling superconductor. For niobium, the workhorse material of superconducting RF cavities, T_c ≈ 9.2 K gives a gap of roughly 2.8 meV. Because no low-energy states exist within the gap, the pairs cannot scatter individually, and a DC supercurrent flows indefinitely without loss.

At microwave frequencies the picture is more subtle. The oscillating field repeatedly accelerates the inertial pair condensate, and the resulting kinetic inductance means screening is imperfect. A residual field reaches the thermally excited quasiparticles, producing a small but nonzero surface resistance described by the Mattis-Bardeen theory. This is why a 1.3 GHz niobium accelerator cavity is operated near 2 K to suppress quasiparticles and reach surface resistance on the order of 10 nΩ, far below the few mΩ of copper at the same frequency.

Governing Relations for Cooper Pairing

BCS Energy Gap at T = 0:
2Δ(0) ≈ 3.52 × k_BT_c

Pippard / BCS Coherence Length:
ξ₀ = (ℏ × v_F) / (π × Δ)

Mattis-Bardeen RF Surface Resistance:
R_s ≈ A × (ω² / T) × exp(−Δ / k_BT) + R_res

Where k_B = Boltzmann constant, T_c = critical temperature, ℏ = reduced Planck constant, v_F = Fermi velocity, Δ = half the energy gap, ω = angular frequency, R_res = residual resistance floor. Example (niobium): T_c ≈ 9.2 K → 2Δ(0) ≈ 2.8 meV, ξ₀ ≈ 38 nm.

Superconductor Pairing Parameters

Material	T_c (K)	Gap 2Δ(0) (meV)	Coherence Length ξ₀	RF Use Case
Aluminum	1.2	0.36	≈ 1600 nm	Qubits, kinetic-inductance detectors
Niobium	9.2	2.8	≈ 38 nm	SRF accelerator cavities
NbTiN	15 to 16	~4.5	≈ 4 to 5 nm	High-kinetic-inductance resonators
Nb₃Sn	18	~6.5	≈ 3 nm	Higher-temperature SRF coatings
YBCO (high-T_c)	92	~30 to 40	≈ 1.5 nm (ab-plane)	Cryo-cooled microwave filters

Common Questions

Frequently Asked Questions

Why do two electrons attract each other to form a Cooper pair?

Two electrons repel via Coulomb force, but in a metal below T_c a phonon-mediated attraction wins out near the Fermi surface. A moving electron distorts the positive ion lattice, and a second electron of opposite momentum and spin is drawn toward that transient positive charge. Cooper showed in 1956 that even a weak attraction binds the pair into a lower-energy spin-singlet state with zero net momentum, with the Debye energy setting the attraction scale.

How large is a Cooper pair compared to the electron spacing?

The pair size is set by the coherence length, ξ₀ = ℏv_F / (πΔ): about 38 nm in niobium and roughly 1600 nm in aluminum. Since conduction electrons sit only ~0.2 to 0.3 nm apart, one pair overlaps the centers of millions of others. That overlap forces all pairs into a single coherent quantum state with one phase, which produces flux quantization, the Josephson effect, and lossless RF screening currents.

Why is RF surface resistance not exactly zero even though Cooper pairs carry no DC loss?

At microwave frequencies the AC field keeps accelerating the inertial pair condensate, so its kinetic inductance prevents perfect screening. A residual field then acts on thermally excited quasiparticles, dissipating power. The Mattis-Bardeen result gives R_s ∝ (ω²/T) exp(−Δ/k_BT) plus a residual floor, so a 1.3 GHz niobium cavity cooled to 2 K reaches ~10 nΩ but never a true zero.

Related Terms

← Cooling Power Cooperative Awareness →