The Gibbs Phase Rule

Ice, liquid water, and water vapor can coexist at exactly one temperature and one pressure. This is the triple point, 273.16 K and 611.657 Pa. Push the temperature up by even a millikelvin at fixed pressure. One phase vanishes. This rigidity is not an accident of water. It is a consequence of a single counting rule. That rule governs every phase diagram you will ever draw. The Gibbs phase rule tells you, before doing any experiment, how many variables you can independently adjust. These variables include temperature, pressure, and composition. It tells you how many you can adjust while keeping a given set of phases in equilibrium. It converts a qualitative phase diagram into a quantitative bookkeeping exercise. This article derives the rule from the equality of chemical potentials. It applies the rule to one-, two-, and three-component systems. It closes by connecting the rule to the one-component and two-component diagrams you meet next.

Key takeaways

The phase rule is $F = C - P + 2$ , linking degrees of freedom $F$ , components $C$ , and phases $P$ .
It comes from counting variables (temperatures, pressures, mole fractions) minus the independent equilibrium constraints between phases.
A one-component system has at most three phases in mutual equilibrium (the triple point), where $F = 0$ .
Adding components adds freedom; adding coexisting phases removes it, one degree at a time.
The "2" in the rule stands for temperature and pressure, the two intensive variables common to every phase.
Key idea: at equilibrium, chemical potentials must match across phases, and each matching condition destroys exactly one degree of freedom.

The core insight

Because the chemical potential of every substance must be identical in every phase it occupies at equilibrium, each additional phase imposes new equality constraints on the same pool of intensive variables, so $F = C - P + 2$ simply counts how many variables survive after all those constraints are satisfied.

The physical picture

Picture a closed vessel holding water. At most temperatures and pressures you find a single phase: liquid, solid, or vapor, and you are free to dial $T$ and $p$ independently within some range without anything changing. That freedom is what "degrees of freedom" means here: the number of intensive variables you can vary independently without altering the number or identity of phases present. Now suppose you tune conditions so that liquid and vapor coexist. You cannot change $T$ without $p$ following along the vapor-pressure curve; the two phases fix a relationship between them. You have lost one degree of freedom. Push further, to the exact point where solid, liquid, and vapor all coexist, and both $T$ and $p$ are pinned. Zero degrees of freedom remain. This is the triple point, a single dot on the phase diagram. The phase rule formalizes this loss of freedom systematically. Every time a new phase joins the equilibrium, the chemical potential of each substance in that phase must equal its chemical potential in every other phase present. Each such equality is one algebraic constraint on the variables $T$ , $p$ , and the composition variables of each phase. Count the variables, count the constraints, subtract, and $F = C - P + 2$ falls out. The rule does not tell you the shape of a phase boundary, only its dimensionality. A one-component, two-phase equilibrium has $F=1$ : it is a curve in the $T$ – $p$ plane (the vaporization or fusion curve). A one-component, three-phase equilibrium has $F=0$ : it is a point (the triple point). Composition enters only when $C > 1$ , giving genuinely new dimensions such as the tie-lines and lever-rule constructions on two-component diagrams you will meet next.

Figure

The triple point of water shown as F = 0 in the p–T plane

The core relations

F = C - P + 2

where: $F$ is the number of independent intensive variables (degrees of freedom), $C$ is the number of components, $P$ is the number of phases in mutual equilibrium, and 2 accounts for temperature and pressure.

F = C - P + 2 - r

where: $r$ is the number of independent chemical reaction equilibria (or other extra restrictions such as a fixed stoichiometric ratio) reducing the count of independent composition variables further.

C = S - r - r'

where: $S$ is the number of chemical species present, $r$ is the number of independent reactions among them, and $r'$ is the number of additional independent stoichiometric or electroneutrality constraints.

Derivation ACounting variables and constraints to obtain F = C - P + 2

Goal. Show that the number of independent intensive variables needed to specify a multiphase, multicomponent equilibrium system is $F = C - P + 2$ .

Step 1: Count the intensive variables. Consider $C$ components distributed among $P$ phases. Each phase is described by its temperature, its pressure, and the mole fractions of the $C$ components within it. Because all phases in mutual thermal and mechanical equilibrium share a common $T$ and a common $p$ (thermal and mechanical equilibrium are assumed throughout), $T$ and $p$ contribute only 2 variables total, not $2P$ . Within each phase, the composition is described by $C$ mole fractions, but they sum to 1, so only $C-1$ are independent per phase. With $P$ phases, composition contributes $P(C-1)$ independent variables. The total variable count is therefore

N_{\text{variables}} = 2 + P(C-1).

Step 2: Count the equilibrium constraints. At equilibrium, the chemical potential of each component must be the same in every phase in which it appears. For one component distributed over $P$ phases, equality across all phases gives $P - 1$ independent equations, since fixing phase 1's potential as reference leaves $P-1$ equalities to phase 2, phase 3, and so on. With $C$ components, each contributing $P-1$ such equations, the total number of independent equilibrium constraints is

N_{\text{constraints}} = C(P-1).

Step 3: Subtract constraints from variables. The degrees of freedom are what remains after imposing every constraint:

F = N_{\text{variables}} - N_{\text{constraints}} = \left[2 + P(C-1)\right] - C(P-1).

Step 4: Simplify algebraically. Expand both bracketed terms:

F = 2 + PC - P - CP + C = 2 + C - P.

The $PC$ and $-CP$ terms cancel exactly, leaving the compact result

\boxed{F = C - P + 2}

Step 5 (extension): Reactive systems. If $r$ independent chemical reactions hold at equilibrium (each contributing one relation $\Delta_rG = 0$ , equivalently one constraint among chemical potentials within a phase), those constraints remove $r$ further degrees of freedom without changing the phase-equilibrium bookkeeping above, giving $F = C - P + 2 - r$ . Equivalently, one can absorb reactions into a reduced component count $C = S - r$ , recovering the same rule in its basic form.

Worked examples

Example 1 — Foundational: locating the triple point of water on the phase rule. Water is a one-component system, $C = 1$ . Consider three cases. Single phase (e.g., liquid only), $P = 1$ :

F = C - P + 2 = 1 - 1 + 2 = 2.

Both $T$ and $p$ can be varied independently within the liquid region: this matches everyday experience, since you can heat or pressurize liquid water over a wide range without anything happening. Two phases coexisting (liquid + vapor), $P = 2$ :

F = 1 - 2 + 2 = 1.

Only one variable is free; fixing $T$ automatically fixes $p$ along the vaporization curve. At $T = 373.15$ K (100 °C), this forces $p = 1.013 \times 10^5$ Pa (1 atm), the familiar normal boiling point. Three phases coexisting (solid + liquid + vapor), $P = 3$ :

F = 1 - 3 + 2 = 0.

No freedom remains: $T$ and $p$ are both pinned, at $T = 273.16$ K and $p = 611.657$ Pa exactly, the defining triple point of water (fixed by international agreement as a calorimetric reference until the 2019 SI redefinition, and still an extremely precisely reproducible point). Sanity check: could four phases of pure water coexist, e.g., two solid ice polymorphs plus liquid plus vapor? That would require $F = 1 - 4 + 2 = -1$ , a negative number of degrees of freedom, which is physically impossible. The phase rule correctly forbids four-phase coexistence for a one-component system; you would need to tune a second independent variable (impossible for a pure substance at fixed composition) to satisfy that many constraints simultaneously. This is why four-phase points do not appear on single-component $p$ – $T$ diagrams, though invariant points with three phases (including two solid polymorphs) do occur away from the classic triple point, e.g., ice I–ice III–liquid.

Example 2 — Applied: a reactive three-component gas mixture. Consider the equilibrium

\text{CaCO}_3(s) \rightleftharpoons \text{CaO}(s) + \text{CO}_2(g)

in a closed vessel, starting from pure CaCO₃ so no external CO₂ or CaO is added. Step 1: Count species and reactions. There are $S = 3$ species (CaCO₃, CaO, CO₂) and $r = 1$ independent reaction, so

C = S - r = 3 - 1 = 2.

Step 2: Count phases. Two solids (CaCO₃, CaO) and one gas (CO�2) are present, so $P = 3$ . Step 3: Apply the phase rule.

F = C - P + 2 = 2 - 3 + 2 = 1.

One degree of freedom remains. Physically, this means that at each temperature there is one, and only one, equilibrium CO₂ pressure — the well-known decomposition pressure curve of limestone. Fix $T = 1173$ K, for instance, and the equilibrium partial pressure of CO₂ is pinned at a specific value (experimentally roughly $p_{\text{CO}_2} \approx 1.0 \times 10^4$ Pa near that temperature); you cannot independently choose both $T$ and $p_{\text{CO}_2}$ once all three phases are present. Step 4: Compare with the naive (non-reactive) count. If you mistakenly treated CaCO₃, CaO, and CO₂ as three independent components ( $C=3$ ) with $P=3$ phases, you would get $F = 3-3+2 = 2$ , implying both $T$ and $p$ could be freely chosen — contradicted by the observed single-valued decomposition-pressure curve. Recognizing the stoichiometric constraint ( $r=1$ ) and reducing $C$ to 2 restores agreement with experiment. This illustrates why correctly counting independent components, not just species, is essential.

Limiting cases & common traps

Situation	What happens to F	Common error
$C=1$ , $P=1$	$F=2$ (an area on the diagram)	Forgetting that a single phase still has 2 degrees of freedom, not "no constraints yet" confusion
$C=1$ , $P=2$	$F=1$ (a curve, e.g. Clausius–Clapeyron line)	Trying to independently fix both $T$ and $p$ along a coexistence curve
$C=1$ , $P=3$	$F=0$ (an invariant point)	Assuming the triple point can be shifted by choosing $p$ or $T$
$C=1$ , $P=4$	Impossible, $F=-1$	Proposing four-phase coexistence for a pure substance
Reactive systems	Effective $C$ reduced by each independent reaction	Counting chemical species instead of independent components
Non-reacting mixtures with fixed overall ratio (e.g. congruent melting)	Extra constraint $r'$ reduces $C$ further	Forgetting a stoichiometric restriction that isn't a "reaction" in the usual sense

The chemist behind it

Josiah Willard Gibbs derived the phase rule as part of his monumental 1876–1878 memoir On the Equilibrium of Heterogeneous Substances, the paper that essentially founded chemical thermodynamics as a mathematical discipline. Working largely in isolation at Yale, Gibbs built the entire framework of chemical potential, free energy, and phase equilibrium from statistical and thermodynamic first principles, publishing in the relatively obscure Transactions of the Connecticut Academy of Sciences. His work went largely unnoticed in America for years until European scientists — including Maxwell, who built a plaster model of Gibbs's thermodynamic surface for water, and later Ostwald and van der Waals — recognized its power and championed it across the continent. The phase rule, though it occupies only a few pages of the memoir, became one of the most quoted and applied results in physical chemistry, underlying metallurgy, ceramics, geology, and every phase diagram drawn since.

Origins · 1 minJosiah Willard Gibbs and how this was discovered

Josiah Willard Gibbs

1839–1903 · American mathematical physicist who founded chemical thermodynamics and derived the phase rule

Ask the tutor: How does the phase rule change when an external electric or magnetic field is imposed on the system?

Why does the phase rule use "2" instead of listing every possible intensive variable?

The "2" specifically represents temperature and pressure because these are the two intensive variables shared identically by every phase in the system at thermal and mechanical equilibrium. Other intensive variables, like composition, are already counted separately through the $P(C-1)$ term. If additional external fields were relevant, such as an electric or magnetic field constraint, the rule would need extra terms, but the standard derivation assumes only $T$ and $p$ as the shared external variables.

Does the phase rule tell you where phase boundaries lie on a diagram?

No. The phase rule only tells you the dimensionality of the region where a given set of phases coexists, such as an area, a curve, or a point. It says nothing about the actual numerical location or shape of that boundary. Finding the real vaporization curve, for instance, requires the Clausius-Clapeyron equation or experimental data, not the phase rule alone. The phase rule is a counting tool, not a predictive equation for boundary shapes.

How do independent chemical reactions change the phase rule?

Each independent reaction at equilibrium imposes one additional constraint of the form $\Delta_r G = 0$ among the chemical potentials of the reacting species. This effectively reduces the number of independent components, since some species compositions become dependent on others. You can either subtract $r$ directly from the standard formula, giving $F = C - P + 2 - r$ , or redefine $C$ from the start as $C = S - r$ , where $S$ counts chemical species. Both approaches give the same numerical answer.

Why can a one-component system never show four phases in equilibrium?

Four phases in a one-component system would require $F = 1 - 4 + 2 = -1$ , a negative number of degrees of freedom. Since $F$ cannot be negative, this configuration is thermodynamically forbidden under the assumption of only $T$ and $p$ as shared variables. Physically, you would need more independent variables, such as an additional composition axis, to accommodate that many simultaneous equilibrium constraints. This is precisely why single-component phase diagrams show at most triple points, never quadruple points, on the standard $p$ – $T$ plane.

How does the phase rule connect to the two-component liquid-vapor diagrams coming next?

Once $C = 2$ , the phase rule permits richer behavior: a single phase gives $F = 3$ (three variables, often reduced to $T$ and composition at fixed $p$ ), and two coexisting phases give $F = 2$ . This extra freedom is exactly what creates the two-dimensional tie-line and lever-rule structure seen in binary liquid-vapor and liquid-liquid diagrams. The phase rule tells you in advance how many independent axes such a diagram legitimately needs.