Phase transition

Phase transitions separate qualitatively different equilibrium phases and are identified by nonanalytic behavior in a thermodynamic limit. A solid melting into a liquid, a liquid boiling into a gas, and a ferromagnet losing its magnetization are standard examples. The relevant control variable may be temperature, pressure, chemical potential, magnetic field, or a combination of them. A phase is not merely a chemical substance: water and ice have different phases, while distinct magnetic or superconducting states can occur in the same material.

In the classical Ehrenfest scheme, a first-order transition has a discontinuity in a first derivative of the free energy, such as entropy or volume, and therefore latent heat or a volume jump. A continuous, or second-order, transition has continuous first derivatives but a divergent or discontinuous second derivative, such as heat capacity or magnetic susceptibility. Modern treatments emphasize the order parameter and the singular part of the free energy rather than relying exclusively on this older classification.²

Phase diagrams show which phase is stable as control variables change, and their boundaries mark coexistence or critical behavior. Along a first-order boundary, two phases can coexist; the Clapeyron relation connects the slope of that boundary to entropy and volume changes. Such boundaries may terminate at a critical point, beyond which liquid and gas are continuously connected and no sharp interface separates them.

Near a continuous transition, long-range fluctuations dominate and ordinary microscopic details become less important. The correlation length grows, while observables follow power laws described by critical exponents. The renormalization group explains this behavior through successive changes of scale and groups systems into universality classes. The three-dimensional Ising model, for example, describes critical behavior in many fluids and uniaxial magnets despite their different microscopic compositions.³ Finite samples round off ideal singularities, so experiments infer critical behavior over a limited scaling range.

Symmetry breaking is a central mechanism in many continuous transitions: a high-temperature phase has greater symmetry, while the ordered phase selects one among several equivalent possibilities. In a ferromagnet, spontaneous magnetization appears below the Curie temperature; in a superconductor, a phase-coherent state forms below a transition temperature. The order parameter provides a compact description of this change, although some transitions have no conventional local order parameter.

First-order transitions commonly involve nucleation, metastability, hysteresis, and interfaces. Supercooling water below its freezing point and superheating a liquid are familiar examples. Quantum phase transitions are driven at effectively zero temperature by a nonthermal parameter, such as pressure, interaction strength, or magnetic field; their fluctuations extend through imaginary time and can influence finite-temperature behavior.⁴ In two-dimensional systems, the Berezinskii–Kosterlitz–Thouless transition is an unusual topological transition associated with the binding and unbinding of vortex pairs rather than ordinary symmetry breaking.

Not every sharp change is a conventional equilibrium phase transition, and several edge cases broaden the subject. A glass transition is generally a kinetic arrest into a nonequilibrium amorphous state, not a simple coexistence line between two equilibrium phases. Likewise, jamming can arise when particles lose their ability to flow as density, stress, or activity changes, but its thermodynamic status depends on the model. In finite systems, apparent transitions may be rounded or replaced by crossovers; in driven systems, equilibrium free energy may not exist at all.

Topological phases provide another important extension. Their defining properties can be global and robust rather than captured by a local magnetization, and transitions between them may require closing an energy gap. The two-dimensional BKT transition is notable because it has an essential, rather than ordinary power-law, singularity. Experimental phase-transition studies therefore combine calorimetry, scattering, susceptibility, transport, and finite-size scaling, using several signatures rather than a single visual change.⁵

Order parameter

A quantity that distinguishes phases, such as magnetization in a ferromagnet or density difference near a liquid-gas critical point.

Critical point

The endpoint of a coexistence line at which two phases become indistinguishable and critical fluctuations occur.

Universality class

A group of systems sharing critical exponents and large-scale behavior despite differing microscopic details.

Latent heat

Heat absorbed or released during a first-order transition while temperature remains constant at fixed pressure.

Quantum phase transition

A zero-temperature transition driven by a nonthermal control parameter and quantum fluctuations.