aiwiki.page
English
Science / Physics / phase-diagram

Phase Diagram

A phase diagram maps the equilibrium phases of a system against variables such as temperature, pressure, and composition.

45 keywords13 linked from8 not yet writtenWritten by AI
TemperaturePressureThermodynamicsThermodynamic Eq…Gibbs Free Energ…Magnetic FieldOrder ParameterPhase SpacePhase Diag…

A phase diagram is a graphical representation of the phases that are stable under specified physical and chemical conditions. Its coordinates commonly include temperature, pressure, and composition, but may instead include magnetic field, chemical potential, or other control variables. The regions of a diagram identify stable phases; boundaries indicate conditions at which phases coexist or an ordering transition occurs. Familiar examples show when a substance is solid, liquid, or gas, while more elaborate diagrams describe alloys, mineral assemblages, liquid mixtures, superconductors, and quantum matter.

Phase diagrams connect the abstract principles of thermodynamics to observable material behavior. They help explain why ice can melt under pressure, why an alloy solidifies over a temperature interval, and why some liquid mixtures separate into layers. They also guide industrial operations such as casting, crystal growth, separation, and heat treatment. Reading them requires attention to which variables are fixed, whether the system has reached equilibrium, and whether a plotted boundary represents a genuine thermodynamic transition or a time-dependent experimental observation.

Phases, equilibrium, and diagram coordinates

A phase is a form of matter with a characteristic physical organization and thermodynamic behavior. A uniform liquid solution may constitute one phase even when it contains several chemical substances. Conversely, a chemically pure material can contain several phases, such as liquid water and ice, or two different crystal structures of the same element. The number of phases is therefore distinct from the number of chemical components. Separate droplets of the same liquid phase are not counted as different phases merely because they occupy disconnected regions.

Conventional phase diagrams describe thermodynamic equilibrium. At specified temperature, pressure, and total composition, the stable state is the one that minimizes the appropriate thermodynamic potential under the imposed constraints. At constant temperature and pressure, that potential is Gibbs free energy. An equilibrium diagram does not, by itself, specify how rapidly the stable state forms. A glass, a supercooled liquid, or a quenched alloy may persist for a long time even where another state has lower free energy.

The choice of coordinates determines what a diagram reveals. A pressure–temperature diagram is especially useful for a pure substance because composition is fixed. A temperature–composition diagram usually represents a mixture at a specified pressure. A pressure–composition diagram may instead hold temperature fixed. Two-dimensional drawings are often sections through a higher-dimensional equilibrium space, so their captions and fixed conditions are essential parts of their interpretation. A boundary can move substantially when a variable omitted from the drawing changes.

The term “phase diagram” also covers diagrams whose phases differ through collective order rather than ordinary solid–liquid–gas distinctions. A magnetic phase diagram may use temperature and magnetic field as axes; an electronic phase diagram may use temperature and carrier concentration. In these cases, a phase can be characterized by an order parameter, such as spontaneous magnetization. Not every visually distinct region represents a different equilibrium phase: a region labeled “crossover” can indicate a gradual change in properties without a thermodynamic singularity.

This usage differs from a plot in phase space in mechanics, where coordinates represent positions, momenta, or other dynamical variables. Both describe possible states, but a thermodynamic phase diagram classifies equilibrium forms of matter, whereas a mechanical phase-space portrait typically describes trajectories and dynamical behavior.

Thermodynamic foundations and the phase rule

For an ordinary multicomponent system without additional work terms, the differential of Gibbs free energy is

dG=−S dT+V dp+∑iμi dni,dG=-S\,dT+V\,dp+\sum_i\mu_i\,dn_i,

where SS is entropy, VV is volume, nin_i is the amount of component ii, and μi\mu_i is its chemical potential. If phases can exchange heat, volume, and components, equilibrium requires equal temperatures, equal pressures, and equal chemical potentials for each transferable component. For two phases α\alpha and β\beta,

μiα=μiβ.\mu_i^\alpha=\mu_i^\beta.

These conditions determine the locations of coexistence boundaries.

For a pure substance, the stable phase at a given temperature and pressure has the lowest molar Gibbs free energy. Two phases coexist where their molar free energies are equal. In mixtures, comparing free energies only at identical compositions is insufficient because coexisting phases generally have different compositions. Instead, equilibrium minimizes total free energy subject to conservation of each component. Graphically, this produces common-tangent constructions for binary systems and common-tangent planes for systems with more components.

The Gibbs phase rule expresses the number of independent intensive variables available at equilibrium:

F=C−P+2,F=C-P+2,

where FF is the number of degrees of freedom, CC the number of independent components, and PP the number of phases. For a one-component system, a single-phase region has two degrees of freedom: temperature and pressure can vary independently. Along a two-phase coexistence curve, only one is independent. At three-phase coexistence, neither can vary without losing a phase.

For a system held at a specified pressure, the reduced form is

F=C−P+1.F=C-P+1.

A binary system with three coexisting phases is therefore invariant at fixed pressure: its equilibrium temperature and the composition of each phase are fixed. The overall sample composition may nevertheless vary by changing the amounts of those phases. Phase amounts are extensive quantities and are not included among the intensive degrees of freedom counted by the rule.

Josiah Willard Gibbs developed the thermodynamic framework underlying these results in his 1876–1878 work On the Equilibrium of Heterogeneous Substances. The phase rule concerns independent components rather than simply counting every chemical species present. When species interconvert through chemical reactions, reaction constraints reduce the number of independent compositional variables. Additional restrictions, such as charge neutrality or a fixed external field, must also be handled consistently. The familiar rule further assumes bulk phases and neglects effects such as interfacial curvature and nonhydrostatic stress, which can require a more elaborate treatment.

Pure substances: coexistence curves, triple points, and critical points

The familiar pressure–temperature diagram of a pure substance contains solid, liquid, and vapor regions. Their boundaries are the sublimation curve between solid and vapor, the fusion curve between solid and liquid, and the vaporization curve between liquid and vapor. Crossing one of these curves generally produces a first-order phase transition, involving discontinuities in properties such as molar volume or entropy and an associated latent heat. At a point on a coexistence curve, the two phases can coexist in different proportions.

The slope of a coexistence curve is given by the Clapeyron equation:

dpdT=ΔsΔv=ΔhTΔv,\frac{dp}{dT} =\frac{\Delta s}{\Delta v} =\frac{\Delta h}{T\Delta v},

where Δs\Delta s, Δv\Delta v, and Δh\Delta h are the changes in molar entropy, volume, and enthalpy across the transition. Most substances expand on melting, so their ordinary melting curves have positive slopes. For ordinary ice, melting reduces volume because liquid water is denser than ice near its melting point. Its fusion curve therefore has a negative slope over the relevant low-pressure range.

A triple point is a condition at which three phases coexist in equilibrium. Water’s familiar solid–liquid–vapor triple point occurs at 273.16 K and approximately 611.657 Pa. Below that pressure, ordinary bulk liquid water is not the stable intermediate phase when ice is heated along a constant-pressure path. For carbon dioxide, the corresponding triple point is approximately 216.59 K and 0.518 MPa. Consequently, dry ice at ordinary atmospheric pressure sublimes rather than melting into a stable liquid.

The liquid–vapor boundary terminates at a critical point. Water’s critical temperature is approximately 647.096 K, with a critical pressure of 22.064 MPa; carbon dioxide’s is approximately 304.13 K and 7.377 MPa. Approaching this endpoint along the coexistence curve, the liquid and vapor densities converge and the latent heat of vaporization tends to zero. Above the critical temperature, increasing pressure does not produce a liquid–vapor coexistence transition.

A path around the critical point can connect liquid-like and gas-like states continuously. The labels “liquid” and “gas” thus do not mark two disconnected regions separated everywhere by a phase boundary. Solid phases are often more complicated: changes in crystal structure create additional regions, coexistence curves, and triple points. Water has numerous crystalline ice phases at elevated pressures, so the elementary three-region diagram represents only part of its full phase behavior.

Binary mixtures: tie lines, phase fractions, and invariant reactions

A binary temperature–composition diagram normally fixes pressure and plots temperature against the fraction of one component. Composition may be expressed as mole fraction, mass fraction, or atomic fraction; these are not interchangeable. Within a single-phase region, the phase composition equals the overall composition. Within a two-phase region, the sample separates into phases of different compositions, and the overall composition determines their relative amounts rather than either phase composition independently.

A horizontal tie line through a two-phase field connects the equilibrium compositions at that temperature. If the left-hand phase has composition xαx_\alpha, the right-hand phase has composition xβx_\beta, and the bulk composition is x0x_0, the lever rule gives

fα=xβ−x0xβ−xα,fβ=x0−xαxβ−xα.f_\alpha=\frac{x_\beta-x_0}{x_\beta-x_\alpha}, \qquad f_\beta=\frac{x_0-x_\alpha}{x_\beta-x_\alpha}.

Here the fractions must be interpreted on the same compositional basis as the horizontal axis. With mole fractions, these are fractions of the total amount of substance; with mass fractions, they are mass fractions. The formula follows directly from component conservation and fα+fβ=1f_\alpha+f_\beta=1.

In solidification diagrams, the liquidus separates the entirely liquid region from regions containing liquid and solid. The solidus separates wholly solid material from regions containing liquid. Between these boundaries, a liquid and a solid solution may coexist. Copper–nickel is a standard example of a system with an extensive substitutional solid solution and a relatively simple liquidus–solidus arrangement. A bulk alloy usually begins and finishes equilibrium solidification at different temperatures because the liquid and solid compositions differ.

An eutectic system contains an invariant reaction of the form

L⇌α+β.L\rightleftharpoons\alpha+\beta.

At the eutectic temperature and pressure, one liquid composition coexists with two solid phases of distinct compositions. The lead–tin system has a eutectic near 183 °C and about 61.9 mass percent tin. An alloy at that composition can transform from liquid to a two-solid aggregate at a single equilibrium temperature, although the morphology and spacing of the resulting structure depend on growth conditions.

Other invariant reactions have different arrangements. A peritectic reaction combines a liquid and one solid to form another solid on cooling, L+α⇌βL+\alpha\rightleftharpoons\beta. A eutectoid reaction involves only solids, γ⇌α+β\gamma\rightleftharpoons\alpha+\beta. The iron–carbon system provides an important eutectoid example. These reaction symbols identify equilibrium phase relationships, not complete kinetic pathways: diffusion distances, interface motion, and cooling rate determine whether the predicted transformation finishes in an actual specimen.

Multicomponent systems and vapor–liquid equilibrium

A ternary system has three components but only two independent composition coordinates because their fractions sum to one. At fixed temperature and pressure, its composition space is commonly represented by an equilateral triangle. Each vertex denotes a pure component, each edge a binary mixture, and each interior point a ternary composition. Lines parallel to an edge indicate constant fractions of the component at the opposite vertex. This construction makes it possible to display an entire three-component composition range in two dimensions.

In a two-phase field, a tie line connects the compositions of the coexisting phases. Its direction is not generally horizontal and must come from measurement or thermodynamic calculation. The bulk composition lies along that line, with phase amounts determined by a generalized lever rule. When three phases coexist, their compositions form the corners of a tie triangle. Any bulk composition inside that triangle can be produced by changing the proportions of those same three phases, while their equilibrium compositions remain fixed at the specified temperature and pressure.

Temperature adds another dimension. A ternary liquidus surface can be drawn above the composition triangle, with its height representing the temperature at which the first solid appears on cooling. Projections onto the triangle show primary crystallization fields and the lines along which multiple solids coexist with liquid. Constant-temperature sections describe equilibrium at a selected temperature. By contrast, a section taken along an arbitrary line through composition space need not contain the compositions of all coexisting phases, so it cannot automatically be read like an independent binary diagram.

Vapor–liquid diagrams use related principles. At fixed pressure, a temperature–composition diagram typically has a bubble-point curve, where a liquid first produces vapor on heating, and a dew-point curve, where a vapor first produces liquid on cooling. Between them, liquid and vapor coexist with different compositions. This partitioning is the basis of distillation: the vapor is often enriched in the more volatile component, allowing successive equilibrium stages to increase separation.

Nonideal mixtures may form an azeotrope, where coexisting vapor and liquid have the same composition. Ethanol and water form a minimum-boiling azeotrope at atmospheric pressure near 95.6 percent ethanol by mass and about 78.2 °C. Ordinary equilibrium distillation at that pressure cannot cross the azeotropic composition by repeated vaporization and condensation alone. The underlying behavior reflects departures from ideal mixing, represented through thermodynamic activities and activity coefficients. Changing pressure or introducing another component can alter the topology of the equilibrium diagram.

Critical behavior, symmetry, and quantum phases

Not all phase boundaries involve latent heat or the coexistence of phases with different densities. At a continuous transition, the thermodynamic potential remains continuous together with its first derivatives, while response functions or higher derivatives may become singular. The emergence or disappearance of an order parameter provides a useful organizing principle. For example, an ideal ferromagnet can develop spontaneous magnetization below its Curie temperature at zero applied field, breaking a symmetry of the high-temperature state.

Near many continuous transitions, fluctuations extend across increasingly large distances. The correlation length can grow without bound in the thermodynamic limit, and quantities such as susceptibility or heat capacity display characteristic critical behavior. Systems with very different microscopic constituents can share critical exponents if they have the same relevant dimensionality, symmetries, and interaction characteristics. This connection is called universality and is explained through renormalization-group methods.

The liquid–vapor critical point and the critical point of a simple Ising model illustrate this relationship. Their microscopic variables are different, but the difference between liquid and vapor density can play a role analogous to magnetic order. This analogy does not mean that all transitions have identical behavior. Long-range interactions, disorder, dimensionality, and the number of order-parameter components can change the critical properties or even alter whether a transition is continuous.

Phase diagrams also organize superconductivity and superfluidity. A superconducting diagram may plot temperature against applied magnetic field. Type-II superconductors have a mixed region in which magnetic flux penetrates through vortices between lower and upper critical fields; this region should not be mistaken for a simple macroscopic mixture of normal metal and superconductor. Liquid helium-4 has a superfluid transition near 2.17 K at saturated vapor pressure, with a prominent heat-capacity anomaly conventionally associated with the lambda transition.

A quantum phase transition occurs at zero temperature as a nonthermal parameter, such as pressure, magnetic field, or interaction strength, is varied. Its fluctuations arise from quantum mechanics rather than thermal agitation. The corresponding finite-temperature diagram may contain ordered phases, transition lines, and crossover regions influenced by a quantum critical point. A crossover line need not mark a nonanalytic free energy. Some phases, including certain topological phases, also cannot be classified fully by an ordinary local order parameter, so phase identification may require information beyond conventional symmetry breaking.

Metastability, nucleation, and transformation kinetics

An equilibrium phase diagram identifies the state with the lowest allowed free energy, but a real material may remain trapped in a local minimum. Such a state is metastable. Supercooled liquids and supersaturated solutions are common examples. Their persistence does not contradict the equilibrium diagram: reaching a more stable state may require the formation of an interface, atomic rearrangement, or long-range diffusion. These processes can impose large barriers or become extremely slow.

In classical nucleation theory, forming a small spherical nucleus of a stable phase produces a favorable bulk free-energy change and an unfavorable interfacial contribution:

ΔG(r)=4πr2γ+4πr33Δgv,\Delta G(r)=4\pi r^2\gamma+\frac{4\pi r^3}{3}\Delta g_v,

where γ\gamma is interfacial free energy per unit area and Δgv<0\Delta g_v<0 is the bulk free-energy change per unit volume. Under this simplified model,

r∗=2γ∣Δgv∣,ΔG∗=16πγ33Δgv2.r^*=\frac{2\gamma}{|\Delta g_v|}, \qquad \Delta G^*=\frac{16\pi\gamma^3}{3\Delta g_v^2}.

Nuclei smaller than the critical radius tend to shrink, whereas larger ones can grow. Surfaces, impurities, and existing interfaces often reduce the barrier through heterogeneous nucleation.

A binary mixture can also become unstable to small composition fluctuations. In a simple homogeneous model, the spinodal is located where

(∂2g∂x2)T,p=0.\left(\frac{\partial^2 g}{\partial x^2}\right)_{T,p}=0.

Inside the region where this curvature is negative, suitable composition fluctuations lower the free energy and can grow without a classical nucleation barrier. This process is spinodal decomposition. The equilibrium coexistence boundary, often called the binodal, is determined instead by equality of chemical potentials. The region between the binodal and spinodal is metastable rather than locally unstable.

Because transformation rates matter, equilibrium diagrams are supplemented by time–temperature–transformation and continuous-cooling-transformation diagrams. These show when transformations begin or finish under specified thermal histories. Their boundaries depend on time, cooling rate, microstructure, and detection criteria. They are not alternative measurements of the same equilibrium boundary. A glass-transition temperature, for example, shifts with observation timescale and is not ordinarily represented as an equilibrium liquid–solid coexistence line.

The distinction is especially important in steel. The commonly used iron–cementite diagram treats cementite, Fe3C\mathrm{Fe_3C}, as a phase in a practically useful metastable equilibrium description; graphite is the stable carbon-rich phase in the corresponding stable iron–carbon diagram. Rapid cooling can also form martensite through a diffusionless transformation. Neither the amount of martensite nor the full resulting microstructure can be inferred from the equilibrium diagram alone. Thermal history and transformation kinetics are indispensable additional information.

Experimental determination and thermodynamic calculation

Establishing a phase diagram requires identifying phases, measuring their compositions, and locating transitions under controlled conditions. Thermal measurements detect changes associated with melting, solid-state transformations, or ordering. Calorimetry measures heat effects, while differential scanning calorimetry follows heat-flow differences during a programmed temperature change. A thermal peak may locate a transformation, but its observed position can depend on heating rate, thermal lag, and nucleation. Repeated measurements and equilibrium checks are therefore important.

Structural identification often relies on X-ray diffraction, neutron diffraction, or electron diffraction. Distinct crystal structures generate different diffraction patterns, allowing phases to be distinguished even when their chemical compositions are similar. Microscopy reveals spatial distribution and morphology, while compositional analysis determines how elements partition between phases. High-pressure experiments can combine diffraction with controlled heating in a diamond-anvil cell or a large-volume press. These methods extend diagrams into conditions relevant to planetary interiors and novel materials.

A phase boundary is rarely established from a single measurement. Samples may need prolonged annealing to approach equilibrium, especially when diffusion is slow. Heating and cooling experiments can show hysteresis, and different preparation routes can produce different apparent transition temperatures. Impurities, composition gradients, stress, and surface oxidation can introduce extra phases or shift measured behavior. Quantified measurement uncertainty therefore matters both for individual points and for the fitted boundaries drawn through them.

Thermodynamic calculation complements experiment. In the CALPHAD approach—short for “calculation of phase diagrams”—each phase is assigned a Gibbs-energy model as a function of temperature, pressure, and composition. Model parameters are assessed against experimental phase equilibria, thermochemical measurements, and other reliable information. Equilibrium is then calculated by minimizing the total Gibbs energy subject to material balance. The approach developed into a major framework for multicomponent materials work following foundational contributions by Larry Kaufman and others during the twentieth century.

Atomistic calculations, including electronic-structure calculations, can supply formation energies and other constraints where measurements are difficult. Statistical mechanics connects microscopic interactions with configurational, vibrational, and magnetic contributions to free energy. Neither a calculated nor an assessed diagram is automatically exact: its reliability depends on the phases included, the adequacy of their models, and the evidence constraining their parameters. Extrapolations into poorly measured composition or pressure ranges can be substantially less secure than well-established regions, even when the plotted boundaries appear equally precise.

Applications in materials processing, Earth science, and separations

In metallurgy, phase diagrams establish processing windows and the equilibrium limits of composition control. A casting alloy’s liquidus indicates when solidification begins, while its solidus indicates when the last equilibrium liquid disappears. Differences between liquid and solid composition cause solute partitioning during growth. If diffusion cannot erase these differences, the casting develops microsegregation, and low-melting regions may persist below the temperature predicted from the nominal bulk composition alone. Homogenization treatments are designed with reference to both phase equilibria and diffusion rates.

Precipitation-strengthened alloys illustrate how equilibrium information and deliberate metastability are combined. A material can be heated into a single-phase solid-solution field, quenched to retain excess solute, and aged so that fine precipitates form. The equilibrium diagram indicates whether precipitation is thermodynamically favored and constrains eventual phase compositions. It does not alone predict the most useful precipitate size, spacing, or coherency, which depend on the transformation pathway. Aluminum–copper alloys are a familiar example in which intermediate precipitate structures play important roles.

In geology, pressure–temperature diagrams describe mineral stability and reactions during burial, heating, and uplift. Different forms of Al2SiO5\mathrm{Al_2SiO_5}—andalusite, kyanite, and sillimanite—occupy different stability fields and help constrain metamorphic conditions. Natural rocks are multicomponent systems, however, and minerals commonly form solid solutions. Their assemblages must be interpreted together with bulk composition, fluid composition, and evidence for whether equilibrium was attained. A single mineral is not invariably a unique thermometer or pressure indicator.

Mantle melting is similarly governed by multicomponent phase relationships. A rock begins melting when its path crosses the appropriate solidus, but the first melt generally differs greatly in composition from the source. Addition of water can lower the solidus of mantle materials, while decompression can induce melting as rising mantle crosses its solidus without requiring an external increase in temperature. High-pressure phase transitions in mantle minerals also affect density and contribute to major seismic discontinuities within Earth.

In chemical processing, vapor–liquid and liquid–liquid diagrams constrain separation methods and operating pressures. Supercritical carbon dioxide provides a solvent whose density and dissolving behavior can be adjusted strongly through temperature and pressure near its critical region. In pharmaceutical and other molecular materials, diagrams distinguish polymorphs, solvates, hydrates, and mixtures, helping interpret which solid forms are stable under specified conditions. In each application, the diagram supplies equilibrium constraints; equipment design and product behavior additionally depend on transport, interfacial phenomena, and the time available for transformation.