Visual walkthrough — Phase rule — Gibbs phase rule

---

Step 0 — What are we even counting?

Term by term, in the sketch:

• The left bucket = every intensive number we are allowed to name.
• The right bucket = every equilibrium equation that removes freedom.
• $F$ = left − right. That subtraction is the entire proof.

---

Step 1 — Draw one phase, list its knobs

WHAT. Take a single phase — say liquid water with a bit of salt dissolved. What intensive numbers describe its state?

WHY. Before counting many phases we must know the cost of one phase. We use mole fractions (not masses) because the rule only tracks intensive variables — proportions, not amounts.

PICTURE. In the figure, one phase is a coloured box. Inside it we write its composition slots $x_1,x_2,\dots ,x_C$, one per component.

The composition slots obey one equation — the fractions of a whole must sum to $1$:

$\underset{\text{fraction of species 1}}{\underbrace{x_1}}+\underset{\text{species 2}}{\underbrace{x_2}}+\cdots +\underset{\text{last species}}{\underbrace{x_C}}=\underset{\text{the whole}}{\underbrace{1}}$

Why this equation matters: it means the last fraction is not free — once you name $C-1$ of them, the final one is forced. So:

---

Step 2 — Stack up all the phases

WHAT. Now put $P$ phases side by side (ice, brine, vapour, ...). Each carries its own $C-1$ composition knobs.

WHY. Different phases can have different compositions (salt is concentrated in brine, absent in vapour). So their composition knobs are independent from phase to phase — we add them.

PICTURE. A row of $P$ coloured boxes, each showing $C-1$ live knobs.

$\underset{\text{how many phases}}{\underbrace{P}}\times \underset{\text{knobs each}}{\underbrace{(C-1)}}=P(C-1)$

But every phase shares the same two global dials — one temperature and one pressure clip the whole assembly together:

$\text{Total variables}=\underset{\text{all compositions}}{\underbrace{P(C-1)}}+\underset{T\text{ and }P,\text{shared by all}}{\underbrace{2}}$

---

Step 3 — The equilibrium constraint: chemical potentials must match

WHAT. Equilibrium isn't free. For each component $i$, its Chemical potential ${\mu }_i$ (the "escaping tendency" of that species) must be equal in every phase.

WHY. If ${\mu }_i$ were higher in phase $\alpha$ than in phase $\beta$, species $i$ would flow from $\alpha$ to $\beta$ — matter drifts downhill in chemical potential. Nothing is at rest until they're level. That equality is the source of every constraint.

PICTURE. For one component, draw water-tanks (one per phase) connected by pipes; the water levels are the ${\mu }_i$ values. Equilibrium = all levels flat.

For a single component across $P$ phases:

$\underset{\text{phase }\alpha }{\underbrace{{\mu }_i^{(\alpha )}}}=\underset{\text{phase }\beta }{\underbrace{{\mu }_i^{(\beta )}}}=\cdots =\underset{\text{phase }P}{\underbrace{{\mu }_i^{(P)}}}$

Count the equations, carefully. A chain of $P$ equal things is written with $P-1$ equals-signs, not $P$. If $a=b=c$ that's two equations ($a=b$ and $b=c$), not three. So:

---

Step 4 — All components at once

WHAT. Repeat Step 3 for every one of the $C$ components. Salt must be level everywhere; water must be level everywhere; each supplies its own chain.

WHY. Each component equilibrates independently — salt's escaping tendency and water's escaping tendency are separate demands.

PICTURE. $C$ stacked copies of the tank picture, one layer per component.

$\text{Total constraints}=\underset{\text{one chain per component}}{\underbrace{C}}\times \underset{\text{equations per chain}}{\underbrace{(P-1)}}=C(P-1)$

---

Step 5 — Subtract, and watch the miracle cancel

WHAT. Now do the one subtraction the whole page was building toward.

WHY. $F$ = knobs minus locks. We have both piles from Steps 2 and 4.

PICTURE. The two towers of blocks from Step 0, now filled with real numbers, and the cross-terms glowing where they will annihilate.

$F=\underset{\text{variables (Step 2)}}{\underbrace{[P(C-1)+2]}}-\underset{\text{constraints (Step 4)}}{\underbrace{[C(P-1)]}}$

Expand each bracket, keeping track of every term:

$F=\underbrace{PC}\underbrace{-P}\underbrace{+2}\underbrace{-CP}\underbrace{+C}$

The two big product terms are the same number written twice — $PC$ and $CP$ are equal, one positive, one negative:

$\underset{=0}{\underbrace{PC-CP}}-P+2+C$

They cancel exactly. What's left is beautifully small:

---

Step 6 — Edge case: the pressure is nailed down (condensed rule)

WHAT. Metallurgists often work at a fixed $1\text{atm}$. Then pressure is no longer a free dial.

WHY. If you remove one global variable from the left bucket, the constant $+2$ drops to $+1$. Nothing else changes.

PICTURE. The same two-dial hat as before, but the $P$-dial is now taped shut with a lock.

$F=C-P+\underset{\text{only }T\text{ free now}}{\underbrace{1}}(\text{the }[[Condensedphaserule]])$

---

Step 7 — Degenerate case: zero freedom, maximum phases

WHAT. What's the most phases a pure substance ($C=1$) can pile together?

WHY. Freedom can't go negative — you can't lock more dials than you have. The most-locked state is $F=0$. Set it and solve:

$0=1-P+2\Rightarrow P=3$

PICTURE. Water's $P$–$T$ diagram: the triple point where solid, liquid, vapour lines all meet. $F=0$ means it's a single dot — you cannot move without losing a phase.

Reading the diagram against the rule:

• Inside an area, one phase ⇒ $F=1-1+2=2$ ⇒ move freely in 2D (an area).
• On a boundary line, two phases ⇒ $F=1-2+2=1$ ⇒ 1D freedom (a line, e.g. the boiling curve — see Clausius–Clapeyron equation).
• At the triple point, three phases ⇒ $F=0$ ⇒ a fixed point.

Every extra phase eats one dimension: area → line → point. That's the phase rule made geometric. (More on the layout: Phase diagrams of pure substances.)

---

The one-picture summary

Everything above compressed onto one canvas: the two buckets fill from Steps 2 and 4, the cross-terms cancel in the middle, and out drops $F=C-P+2$, whose values $2/1/0$ map onto area/line/point.

Recall Feynman retelling — the whole walkthrough in plain words

Picture a control panel. First I ask: how many dials could I possibly turn? Each phase lets me choose its recipe — but the fractions must add to a full 100%, so the last one is decided for me: that's $C-1$ real dials per phase, times $P$ phases. On top sit exactly two master dials, temperature and pressure, shared by everybody, because heat and squeezing spread equally through touching phases. That's my "freedom" pile: $P(C-1)+2$.

Then nature fights back. For every ingredient, its "eagerness to escape" (chemical potential) must read the same in every phase — otherwise it sneaks from the crowded phase into the empty one until they're level. Lining up $P$ phases needs $P-1$ equals-signs, and I do that for all $C$ ingredients: $C(P-1)$ locks.

Subtract locks from freedom. The two big tangled products $PC$ and $CP$ are the same number, one plus, one minus — poof, gone. Left behind: $C-P+2$. And if I bolt the pressure dial shut, the $2$ becomes a $1$. If a pure substance ($C=1$) has zero freedom, it's pinned at three phases at once — the triple point, a single dot you cannot budge.

Recall Quick self-check

Why do $PC$ and $CP$ cancel? ::: They are the same product with opposite signs — one from expanding the variable count, one from the constraint count — so they sum to zero, leaving the simple linear terms. What does each "$-1$" and "$+2$" physically remove or add? ::: The $C-1$ per phase removes the redundant mole fraction (sum-to-one); the $+2$ adds the two shared global dials $T$ and $P$. If pressure is fixed, what is $F$ for two components in two phases? ::: $F=2-2+1=1$.

---

Connections

• Chemical potential — supplies the ${\mu }_i^{(\alpha )}={\mu }_i^{(\beta )}$ equalities counted in Steps 3–4.
• Gibbs free energy and equilibrium — where the equal-$\mu$ condition comes from.
• Components and independent reactions — how to get $C$ right before you start counting.
• Condensed phase rule — the $+1$ version from Step 6.
• Phase diagrams of pure substances · Clausius–Clapeyron equation — the area/line/point geometry of Step 7.