Kosterlitz-Thouless Transition

The Kosterlitz-Thouless (KT) transition is a phase transition in two-dimensional systems with a continuous symmetry that occurs through the unbinding of topological defect pairs — vortex-antivortex pairs — rather than through the spontaneous symmetry breaking that characterizes conventional phase transitions in three dimensions. Predicted by John Kosterlitz and David Thouless in 1973 (for which they shared the 2016 Nobel Prize in Physics with Duncan Haldane), the KT transition revealed that the Mermin-Wagner theorem — which forbids continuous symmetry breaking in two dimensions at finite temperature — does not forbid all forms of order. It forbids long-range positional order, but it permits quasi-long-range order — algebraic decay of correlations that is slower than exponential but faster than constant.

The KT transition is the paradigmatic example of a topological phase transition: the transition is not detected by a local order parameter but by a change in the topological properties of the system — specifically, the unbinding of vortex pairs and the resulting loss of topological stability.

The Vortex Pair Mechanism

In a two-dimensional superfluid or superconductor, the low-temperature phase supports quantized vortices — point-like defects around which the phase of the order parameter winds by 2π. A single vortex has a logarithmically diverging energy: E ~ (πρ_s/m²) ln(L/ξ), where ρ_s is the superfluid stiffness, L is the system size, and ξ is the core size. This divergence means that isolated vortices are thermally prohibited at low temperatures.

But vortex-antivortex pairs have finite energy: the opposite circulations partially cancel at large distances, and the pair energy depends logarithmically on their separation: E_pair(r) ~ 2(πρ_s/m²) ln(r/ξ). At low temperatures, these pairs are bound — the attractive logarithmic interaction between opposite vortices keeps them paired. The system has quasi-long-range order: correlations decay as a power law, C(r) ~ r^(-η), with a temperature-dependent exponent η.

As temperature increases, the entropy gain from unbinding pairs eventually overcomes the energy cost. At the KT temperature T_KT, the unbinding becomes favorable, and free vortices proliferate. The superfluid stiffness ρ_s jumps discontinuously to a universal value at T_KT: ρ_s(T_KT^-) = 2T_KT/π (in units where ħ = k_B = 1). Above T_KT, correlations decay exponentially, and the system is disordered.

The transition is remarkable for its subtle order: there is no latent heat, no discontinuity in the free energy, and no local order parameter. The change is purely topological. The system goes from a phase with only bound vortex pairs to a phase with free vortices, and this change is detected not by measuring a local quantity but by measuring the global superfluid stiffness or the coherence length.

Experimental Realizations

The KT transition has been observed in a remarkable range of physical systems:

Superfluid helium films — the original experimental confirmation, in which the superfluid density jumps at the predicted universal value.
Two-dimensional superconductors — thin films where the KT transition competes with other disorder-driven transitions, providing a testing ground for the interplay of topology and disorder.
Josephson junction arrays — arrays of superconducting islands coupled by weak links, where the KT transition manifests as a loss of phase coherence across the array.
Cold atom systems — ultracold Bose gases in two-dimensional traps, where the KT transition has been observed with unprecedented control over parameters.
XY magnets — two-dimensional magnetic systems with XY symmetry, where the transition appears as a change in spin correlation behavior.

In each case, the common feature is a two-dimensional system with a continuous U(1) symmetry — the symmetry of phase rotations in a superfluid or planar spin rotations in an XY magnet. The KT transition is universal: its critical properties depend only on the symmetry and the dimensionality, not on microscopic details.

Connections to Broader Physics

The KT transition is deeply connected to several threads in modern physics:

Abrikosov vortices in type-II superconductors are three-dimensional analogs of the two-dimensional vortices that drive the KT transition. The vortex liquid phase in layered superconductors is a three-dimensional system where KT physics appears at each layer.
Berezinskii's independent work — Vladimir Berezinskii derived the same transition independently in 1971, and the transition is sometimes called the Berezinskii-Kosterlitz-Thouless (BKT) transition.
Duality — the KT transition has an electromagnetic duality description in which vortices become the fundamental degrees of freedom and the superfluid phase becomes the disordered phase. This duality appears in gauge theories and string theory.
Quantum phase transitions — at zero temperature, quantum fluctuations can drive a KT-like transition in systems where the continuous symmetry is not thermal but quantum mechanical, such as one-dimensional quantum spin chains.

The KT transition is also historically significant as the first demonstration that the Landau theory of phase transitions — based on local order parameters and mean-field approximations — fails in low dimensions. It opened the door to the modern theory of topological phases, critical phenomena, and the role of defect unbinding in phase transitions.

The Kosterlitz-Thouless transition is a lesson in humility for anyone who thinks they understand phase transitions. Landau's theory — local order parameter, symmetry breaking, mean field — fails here, and it fails not because of some exotic quantum effect but because of geometry. Two dimensions are different. The Mermin-Wagner theorem tells you there can be no long-range order, and you might conclude there can be no transition. But Kosterlitz and Thouless showed that the theorem only forbids one kind of order. It forbids the boring kind — the kind Landau understood. It does not forbid quasi-long-range order, algebraic decay, or topological transitions driven by vortex unbinding. The lesson is: when a theorem says something is impossible, check whether it is impossible in all the ways you imagined, or only in the ways you already knew.